Elektor Electronics 2014-12

Elektor Electronics issue

Uploaded by

Adrian_Andrei_4433

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF or read online on Scribd

100% found this document useful (1 vote)

332 views76 pages

Elektor Electronics 2014-12

Elektor Electronics issue

Uploaded by

Adrian_Andrei_4433

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF or read online on Scribd

You are on page 1/ 76

www.el magazine.com emagazine December 2014 | No. 456 CoCo-ri-Co: Cool Controller Concept mi Wier lee) Human Inte: ig CoCo-ri-Co: The Cool Controller Concept | T Board 28 Lowest-Power Exercising Programmable Christmas Tree | varitab 402 (2) | ADS1115-eBoB | Arduino Software Development with Atmel Studio | Infinite RGB LED Cube with Flowcode | Distance and Level Gauge Improved Current Transformer | Red Pitaya | DesignSpark Tips & Tricks Tantalum Bead Capacitors | Bulging Caps | Elektor.Labs goes wearable | GestIC & 3D TouchPad Workbook (1) | GM2308 Audio Signal Generator (1950/64) | Hexadokutor © Projects 10 CoCo-ri-Co: The Cool Controller Concept A truly flexible, modular interface for microcontroller projects. Forget ‘about switches, dials, displays, LCDs—with the Elektor CoCo-ri-Co > This isthe feet fourhundred and fifty-icth ‘edition of Elektor, a magatine that introduces a new ‘way of presenting electronics. ‘The Dutch edition of Elektor has been published for over #453 years and the (German for over ¢43. very month #29:00¢ copies several terabytes find their way to rexier= members ranging from enthusiastic amateurs to professional electronic engineers ‘Hlekaor’s dynamic and practical application of new electonic techniques has stimulated the ‘ever-present curiosity and imagination of engineers. Modern components, active and passive and especially cheap digital and linear circuits, are used in practical designs. Many ofthe circuits are developed in cur own laboratories, and cireit building is greatly facilitated by ‘using ready-made printed circuit boards we produce for the more important projects. [.] “Elektor haz always tried to be dmamic and informative: but it can oocasionally irritate, a2 ‘when it deflates technical imperiousness or indulges in humorous self-criticism that has given ‘ta British’ image on the continent. ‘We shall be working on the frst copies for 3975 2015 even as you ead this. Articles already accepted describe an electronicelly-compensated:loudopenker system ARM controlled “synthesizer stuglrquainy preampbifer an AVR controlled function generator, xramatogue ‘digital converter a LED replacement for 6-V moped lamps. gyraters Weird Components, and ‘farther developments of the mer-elock Platino,electronse-irem TBoard and SAPELPP. << Welcome back in 2014, Happy Reading, Jan Bulting Editor-in-Chief Elektor International Media The Team Editor-in-Chief Jan Bulting Publisher / President: Don Akkermans Membership Menager: Raoul Morreau (all areas) International Editorial Sta: Harry Baggen, Jaime Gonzalez-Arintero, Denis Meyer, Jens Nickel Laboratory Staff ‘Thijs Beckers, Ton Glesberts, Wisse Hettings, ‘Luc Lemmens, Mart Schroijen, Clemens Valens, Jan Visser, Patrick Wielders Graphic Design & Prepress: Glel Dols, (Online Manager: Danielle Mertens Managing Director: Don AkkermansGestIC & 3D TouchPad WorkBook (1) Raspberry Pi gets gesture control added By Thomas Lindner Last month we introduced the Microchip/Elektor 3D Gesture Control dev kit and & Roland Aubauer 3 Touchpad product bundle. Here we showcase the potential of the kit by con- (Wicrochip GestIC® Team, Germany) and necting the key component in the system, the MGC3130 GestIC® controller, to an Jan Buiting (Elektor) RPI computer. In response to introducing 1. What hardware Microchip & Elektor’s Joint To be able to follow the project in an educational exclusive offering of a prod- manner you need this hardware: uct bundle consisting of the # Raspberry Pl Model B v.2, RBCAOOO Hillstar MGC3130 3D Ges- 2x USB 2.0 3.5 watts ture Control Development Kit Power supply: as ol (dev kit’) and a ready to use micro USB 1200 mA SV for RPI 3D Touchpad, the most fre- # MGC3130 Hillstar Development Kit MGC3130 Hillstar Single Zon« Development kit | quently heard sighs and wishes (Part # DM160218) were Iwant to connectit to my The final sensor system works independent from ‘micro! Followed by I want to the PC and just needs to be powered by @ USB ‘connect t to my PC! The offer charger or directly from the RPI. Communication was announced last month [1] between the chips is established via the Ges- and the hardware ison sale by _turePort pins provided on the little MGC3130 the time this article appears inboard in your Hillstar dev kit. A PC Is required print (2). for parameterization and flashing the firmware In this short WorkBook series to the MGC3130. A block diagram of the hard- welll step-demo the case of aware configuration Is shown in Figure 1, and a dedicated, mixed touch & 3D picture of the setup in Figure 2. touchfree, controller for connection to the Raspberry Pl 2. Connecting it up ‘comput, aiming eventually to The MGC3130 IC takes all the hassle out of capac- play the “2048" game. It should itive sensing for 3 gesture control applications. tenable the first wish to come The E101, E102, £103, £106, £107 signals on the true—more to follow. hillstar’s MGC3130 board are connected to the P| GPIO connector as pictured In Figure 3. Not ‘This being @ WerkBook, the idea forgetting Ground (GND) of course. Is for You to get active with the technology involved, so we'll 3. Parameterization with Aurea use up-tempo notes and tech The MGC3130 allows East-West and North-South scribbling. The information sup- hand flcks to be assigned to and flagged by its plied here Is intended to spur EIOx pins. Aurea, the free graphic shell around Ideas for hardware and soft- the GestIC controller, allows you to parameter- ware development with the ize many sizes and configurations of planes that MGC3130 chip for 3D gesture make up the capactive sensing pad to be call control on any microcontroller brated and configured with high precision. The system really. parameters for the GesturePort are pictured inFigure 4. Note the unique mix of 3D Gesture (Flick EW; Flick NS) and Touch (Center). 4. Programming stuff In terms of software, we will be using: © OS: Raspbian (Debian Wheezy), version: January 2014, release date: 2014-01-07 ‘* Python version: 2.7.3 (default, Mar 18 2014, 05:13:23), (already installed on Raspbian) ‘© RPI.GPIO library version: 0.5.4 (already Installed on Raspbian) ‘¢ Tkinter (already installed on Raspbian) Leafpad text editor (already installed on Raspbian) ‘© The code for the game ‘2048_with_Gesture_ Port_Demo.py’ The game proper will be discussed and url'ed to you next month. 5. Quick Start 1. Install Raspbian on your Raspberry Pl. 2. Connect the Hillstar dev kit via USB to your PC and start Aurea. There should be a small popup window due to synchronization. 3. At the top side go to ‘Setup’ and then choose ‘Parameterization’. 4. Now click at the left side on ‘Extended’, browse in ‘Parameters’ for the pre-configured file ‘Hill- star GesturePort to Raspberry Pi Demo 2048. enz’ and start the parameterization, which takes a while. 5. After finishing, at the left side click on ‘Ges- turePort’ and you see the configuration for the events. If you click on ‘Flash’ in the upper right comer the configuration shown in Figure 4 is, saved in the MGC3130 chip. 6. Once flashed, the Hillstar dev kit runs in standalone mode and you can connect the demo with your RPi as pictured in Figure 1. It can be powered either by an external USB charger via the mini USB connector or you steal 5 V from the Raspberry Pi. Web Links Hillstar MGC3130 Unit Raspberry Pi Next month GestIC & 3D TouchPad WorkBook delves deeper in the demo and the 2048 game. Meanwhile Elektor Lab notes on the GestiC devel- ‘opment kit and 3D Touchpad may also appear in the Elektor. POST newsletter. (140423) [1] Add 3D Sensing to your Micro or PC. Elektor November 2014, www.elektor-magazine.com/ 140408. [2] Microchip Hillstar GestIC dev kit and 3D Touchpad product bundle: www.elektor.com/microchip-dm160218-1 [3] www.raspberrypi.org/downloads/ istar-development-kit-and-dm160225-3d-touchpadBy Clemens Valens (GiektorLabs) CoCo-ri-Co: The €@ol Controller GOncept CS at solving the perennial problem of adding knobs, buttons and displays to microcontroller designs without reinventing the wheel. I have tried this before and faithful Elektor readers may remember the J2B ARM Cortex-M3 board [1], or ‘Platino’, the Arduino compatible, universal Seite tee ‘© NXP LPCB12 32-bit ARM Cortex-MO+ ‘All 48 MCU /0 pins accessible through ‘extension connectors ‘Up to 17 bicolor LEDs '* Rotary encoder and/or pushbutton Buzzer ‘© Supports TC, SPI/synchronous and ‘asynchronous serial communication ‘* ISP port compatible with 3.3-V FTDI USB- serial cable (except 5-V supply) Here is a new initiative in the quest for the ultimate human interface in microcontroller designs. Instead of the 210%” all-in-one solution this time we resort to the modular approach. The result is truly flexible-it can even be used as a (wearable) Christmas decoration or for 3D modeling. ‘Atmega board [2]. These two boards offer a variable number of rotary encoders, ‘and pushbuttons together with a choice of LCD sizes, the idea being that one of these boards ‘would satisfy the designer's appetite for controls. ‘This time I went for a different approach. Instead ‘of many controls that can be positioned in different ways, this project only has one. Instead of one large versatile board that’s supposed to contain everything, this design Is for a small ‘module that you sprinkle on or into your project like salt on an egg. No more bulky hard-to-read LCDs displaying cryptic messages in ugly fonts, this time we stick to LEDs, Bright lights visiblefrom a great distance and blinking fast, slowly or not at all is what we want; LCDs are uncool. Compromises from the past Many electronic devices are equipped with knobs and buttons that allow its user to adjust one or more parameters. Some of these controls do not need a precise indication of the value of the parameter they adjust. The volume control of an amplifier is an example. When the sound is loud (or soft) enough you simply stop turning the knob. Other controls—function selectors for instance—do need some indication to show their position but the resolution is not too important. A few LEDs are just fine in these cases. For controls that allow precise adjustments a numeric display is probably best. Many applications use up/down pushbuttons to replace rotary controls even though the latter are faster and easier to operate. While it is true that pushbuttons are small, cheap and easily handled by microcontrollers, they remain a compromise from the days that microcontroller resources (memory, 1/0, processing power, etc.) were expensive. Today this Is no longer the case and we can throw lots of them at simple problems without adding much beyond one dollar to the bill of materials. So why stick Reinvent the wheel The design presented in this article is revolution- ary (pun intended) in offering ‘only one rotary encoder, up to 17 bicolor LEDs (i.e. 34 LEDs in total) and a buzzer controlled by nothing less than a 32-bit ARM Cortex-M0+ based microcontroller unit (MCU from now on). Overkill? Maybe, if you only look at processing power, but clearly no, also taking costs and interconnection potential into account. The MCU I chose for this project Is the LPC812M101JDH20FP, LPC812 for short, featuring 18 1/0 ports, 16 KB flash memory, 4 KB RAM, a good number of timers, an analog comparator, one 12C interface, two SPI controllers and three asynchronous serial communication Interfaces. What's more, thanks to an integrated ‘switch matrix, the pins of these communication peripherals can be connected to any one of the 18 available 1/0 ports. This cool feature means that it is possible to build a multi-protocol low-on- wires communications port that can be configured on the fly in software to talk either I2C, synchronous (SPI or similar) or asynchronous serial. Add pin and full duplex communication becomes a reality. At a high-margin retailer's price of about 2 dollars apiece this clearly is a steal. All components are mounted on a small circular printed circuit board (PCB) with the MCU, one LED and the rotary encoder in its center and the remaining 16 LEDs arranged in a circle around the encoder. At the edge of the PCB there Is a ring of pinheaders for connecting the module to other modules to create a control surface or to wire it up to a host board. Since there are only cool components on the board (ARM Cortex-M0+ MCU, bicolor LEDs, rotary encoder) it was baptized Coo! Controller Concept or CoCo-ri-Co (the “ri” was added as a playful hint at the French way of writing the rooster’s call). BTW, if you look hard at the board you will notice that it is hexagonal rather than circular. Round Is cool, hexagonal Is cooler; this is how (CoCo-ri-Co reinvents the wheel. Who needs this? The main application of the CoCo-ri-Co board should be a digital potentiometer with instant position indication. Any device or appliance that could benefit from one or more rotary controls to adjust its parameters Is a potential host to ‘CoCo-ri-Co. Think of volume and tone controls in audio amplifiers, function selectors and speed controls on household appliances, temperature controls of heating systems or brightness and color controls of lamps—all of these applications make excellent hosts to one or more CoCo-ri-Cos. Because it can communicate using serial, I2C, SPI or custom protocols all through a single 3-pin port, CoCo-ri-Co can easily be connected to host applications that have a microcontroller inside. For those applications that do not have their own MCU, CoCo-ri-Co can be the main controller. CoCo-ri-Co is small enough to fit in most enclo- sures, even In wall-mounted junction boxes and power outlets, and because of Its hexagonal shape multiple boards can be tiled together to create a control surface. As an example, think of a speedcontro! with dashboard for an electric vehicle (Figure 1). One CoCo-ri-Co with encoder controls as a speedometer, another—encoderless— for displaying the battery level and a third might show the number of revolutions per minute (RPM) the motor is doing, or its temperature (or both). But there is more. ‘Thanks to its size and low-power requirements, (oh, did I forget to mention that?) CoCo-ri-Co can be battery powered, from a 3 V button cell for instance, making it portable and even wearable (a geeky smart watch, anyone?). The bicolor LEDS allow for decorative applications and it excels at Christmas with its red and green looks and its buzzer happily beeping away at Christmas carols (see inset). But that is not all. I already mentioned tiling boards together to create control surfaces, but these surfaces don’t have to be flat. Indeed, you can create 3D surfaces from CoCo-ri-Co boards. The Cool Controller Concept may be extended to include pentagonal, ‘square and triangular boards. With these shapes it 's possible to create so-called Archimedean solids, you know, those highly symmetric, semi-regular convex polyhedrons composed of two or more types of regular polygons meeting in identical vertices. Like the stitched soccer ball. ‘Such a ball, mathematically known as a truncated Icosahedron, is composed of 20 hexagons and 12 pentagons. A soccer or Bucky ball made out of CoCo-ri-Co boards (Figure 2) would be about 12 cms in diameter (25-mm vertex). Add a 3D accelerometer and it can show LED animations ‘and play sounds depending on its position, speed ‘and movement. Now how cool is that? CoCo-ri- Cool! I mean, totally cool. What does it do? Every time when the position of the rotary encoder changes CoCo-ri-Co will send a value (increasing or decreasing, depending on the direction of rotation) on the communications port (default: UART, 115200n81). This value lies between two boundaries (0 to 100 by default). ‘The LED ring indicates the value as a bargraph (default) or a dot, in red (default) or green. The center LED can be controlled separately. Pressing the pushbutton will produce a Key-Down message and releasing it generates a Key-Up message. ‘The messages can have the following formats: ASCII mode © Rotary encoder: Bxxocxx, where 20000 Is ‘a S-digit field transporting values from 00000" to “65535”. + Pushbutton: Bx, where x = ‘0° (button down) oF ‘1” (button up). Binary mode © Rotary encoder: Eyz, where the value is calculated as 256 y +2. Pushbutton: By, where y = 0 (button down) or 1 (button up). You can modify the behavior of the program on the fly by sending commands to CoCo-r1-Co. This allows you to change things like the color of the LEDs, the boundaries of the encoder value or thedefault value). Every command req ‘Command —_|Data Deseri LEDs (0x00 (0) LED ring color: off | red* | green LED ring mode: bargraph* | dot [0x20 (32) (0x02 (2) LED ring first LED (0*) (0x03 (3) LED ring last LED (15*) ‘0x04 (4) LED ring direction: clockwise* | anti-clockwise (0x05 (5) LED ring rotation (0*) ‘0x06 (6) LED ring LEDx on (0x07 (7) LED ring LEDx off (0x08 (8) O1t*12 (Centre LED: off | red* | green. Rotary Encoder (Oxi0 (16) [O*l Encoder mode: normal* | accelerated Oxii (17) | 0-255 (0) Encoder minimum, low byte (0*) (Oxi2 (18) | 0-255 (0) Encoder minimum, high byte (0*) (0x13 (19) (0-255 (64) Encoder maximum, low byte (64; 0x14 (20) [0-255 (0) Encoder maximum, high byte (0*) 0x15 (21) [0-255 (0) Encoder value, low byte (0*) x16 (22) [0-255 (0) Encoder value, high byte (0*) Buzzer Buzzer disabled | enabled* (0x21 (33) 1 Buzzer beep now (0x22 (34) [0-255 (232) [Buzzer beep frequency [Hz], low byte (232 (0x23 (35) | 0-255 (3) Buzzer beep frequency [Hz], high byte (3*) [0x24 (36) 0-55 (64) Buzzer beep duration [ms], low byte (64*) [0x25 (37) 0-255 (0) Buzzer beep duration [ms], high byte (0*) Other Oxfe (254) TO*]1. ‘Output mode: ASCIT | binary OxfF (255) 1 Restore default values encoder value Itself, the mode of the encoder, the range of LEDs to use (default is 0 to 15). You can also rotate the bargraph, produce a beep, etc. At the time of writing this article I am still adding useful commands, so please refer to the free downloads for the complete list. Updates for this project will be posted on Elektor.Labs [3]. Table 1 lists the currently implemented commands. Configuration One-time configuration of CoCo-ri-Co Is by means of trimpot P1. Cofiguration consists mainly of choosing the communications port type to use (UART, SPI/synchronous or IC). With P1 you can select up to 32 values in theory, but for practical reasons—the precision of P1 being the most important—this has been limited to eight ranges. In Table 2 you can find which functions Presenter eo oro err cere Pe a enone econo ake rea High (3.3 V), otherwise CoCo-ri-Co will boot into ISP mode. UART, no parity, 8 data bits, 1 stop bit (n81) /SPI/synchronous, Slave address Oxib (29) address Ox3a (58) ‘address 0x74 (116) BC, Slave P4=SDA, PO=SCK (P12=IRQ) ea] 5)>)9] >]are activated by a range. you want. Stay In the middle of a range, L.e. ‘The ranges are defined such that turning P1 to _away from range boundaries or you may expe- Its minimum value selects UART mode, setting _rience random mode changes in the (near) it in the middle gives SPI/synchronous and at its _ future. maximum CoCo-ri-Co will talk I2C. A more precise 5. Switch off the board. Figure 3. procedure Is available also and goes like this: 6. Remove the jumper between P6 and GND (con- penance ae 1. Switch off the board. nector K5, pins 1 and 2} (Co. The LPCBIZ ARM ; Bl ) CortexMO+ microcontroller 2° Fit @ jumper between P6 and GND (connector 7. Switch on the board and test. has more than enough KS, pins 1 and 2). braine ond muscle tetake 3+ SWitch on the board, Do not fiddle too much with P1 because It Is a care of everything. 4, Adjust Pi to make the LED ring ight the range tiny and fragile device. oe § e E e ce me i es 3 a wv e : :Note that the SPI/synchronous and I?C Slave modes require the host application to poll CoCo- 11-Co on a regular basis. You could use a third port to interrupt the host when new data is available. Also note that real SPI Slave mode using the MCU's SPI peripheral requires the SSEL signal. It Is up to you to decide which port to use for this. The “SPI” implemented uses the USART in synchronous mode that can work without a SSEL signal. If Slave mode doesn't suit you, be aware that Master mode is available too (after some programming). Into the schematic After introducing the board, its functions and capabilities let's have a look at Its schematics In Figure 3. Connecting 34 LEDs (bicolor, remember?) directly to 18 1/0 ports is hard if not impossible, but when connected as two anti-parallel 4x 4 matrices (16 LEDs per matrix) they only eat up 8 pins of our budget. These 16 LEDs form the circle. This leaves one bicolor LED—positioned in the center of the board—that must be connected in another way. Little known, the LPC812 MCU I2C module supports two types of interfaces: ‘normal’ capable of using any 1/0 pin, and ‘dedicated’ insisting on using ports 10 and 11. ‘These two ports have high current sink capabilities and are therefore great for controlling LEDs, so I wired the remaining bicolor LED to these ports. We will content ourselves with normal 12 for communication. To limit the current through the LEDs resistors are connected in series with them. Normally the red and green LEDs would require different valued resistors to obtain the same perceived brightness (red usually being brighter than green) but for the LED matrix this is not possible. It would have been possible to do this for LED17, but for convenience’s sake I only used one value resistor on the board: 180 2. For the rotary encoder I picked a model with integrated pushbutton. Encoders without pushbutton are about five times cheaper, but I like the tum ‘n’ push way of setting parameters. The pushbutton is hooked up to the MCU's Reset pin. That may seem strange, but it makes sense once you know you can disable the reset function on this pin in software. During software development you would normally keep the reset function available, but once the application Is finished you can remove it. Or mount an encoder without an integrated pushbutton. For those applications requiring a pushbutton only, the PCB has a special pushbutton footprint too, wired in parallel with the encoder’s pushbutton. ‘The encoder is in the center of the board, on top of LED17. Why? First, your application may need ‘only one of the two components, so you mount only what you need. Second, we can have shaft encoder lighting. LED17 allows you to use the encoder shaft as a light guide (of course, you will need an encoder with a transparent shaft for this). The pushbutton footprint is also in the center of the board but as far as I know the model I chose does not come in a transparent package. In any case, LED17 can produce a sort of cool background glow of the control. Many devices beep, or, as serious engineers would say, they provide audible feedback. A module presented as a universal control module must have the possibility to do the same. Therefore a footprint for a buzzer wasn’t forgotten. Due to space constraints I had to use an SMD buzzer, but unfortunately they can be pretty expensive. A variable resistor P1 is attached to port PIOO_1, which Is connected inside the MCU to an analog comparator. This comparator can distinguish between 32 levels that make it useful, for example, as a 32-position switch to select different software functions. It is mounted at the LED side of the board for easy access (though it’s pretty small). Do not mount P1 if you have no need for it. ‘The MCU does not require an external crystal or oscillator because It has a good one built in. If, you need a better clock you should wire it up to PIO0_8 and PIO_9 but the board Is not equipped with footprints to make this an easy job. All resistors on the board have the same purpose (current limiting) and value (180 2). They protect most of the 18 1/0 ports, but not all. The two ‘SWD pins PIOO_2 and PIO0_3 are unprotected 50 as not to interfere with any debugging pods. ‘The LED matrix row pins are not protected either because it Is unlikely for them to see any use outside the board, and board space is limited. Capacitors C1 and C2 provide power supply filtering.‘Component List Resistors R1-R12 = 1800, SMD 0603 Pi» 100k0 trimpot, 3mm, e.. Bourns TC33K-2-104E Capacitors C1 = 100nF, SMD 0603 C2 = 10uF 6.3v, SMD 0805 Semiconductors ICA = LPCB12M101IDH20FP (TSSOP20) LED1-LED17 = LED, bicolor red/green, Dial ‘ight 5988610207 Miscellaneous BUZA = buzzer, Kingbright KMTG1102-AL ‘S2 = pushbutton, Multimec 3FTL6 ‘al (optional) Extension connectors The MCU also Is located in the center of the board, but at the solder side with the resistors, capacitors and buzzer. All of its 18 1/0 pins have been routed to six 5-pin extension connectors, each giving access to three I/O ports (6x 3 = 18) plus, 3.3 Vand GND. These connectors are not wired up in a random way, but so that they have a useful function. K1 and K4 for instance provide in-system programing (ISP) capabilities. K4 is the real 15P connector (almost!) compatible with a 3.3V FTDI USB-to-serial converter cable and K1 adds the Reset pin to It. They are on opposite edges of the board, 1.5 inch apart, so itis easy to con- struct a more elaborate programming adapter on. prototyping board if you would like to. Such an adapter would probably have a 3.3-V regulator on it to transform the 5 V from the 3.3-V FTDI cable (yes, beware!) to 3.3 V. Actually, the Elek- tor 110553 BoB-FT232R USB-to-serial adapter [4]|s preferable to the FTDI cable because it can provide the 3.3 V too. K2 Is considered to be an analog port, because It 1s connected to the MCU’s analog comparator’s ‘S1 = rotary encoder with/without integrated Ready assembled module: Elektor Store # pushbutton, e.g. Alps EC12E2424407 Ki-K6 = 5-pin pinheader, horizontal or verti-- Optional: PCB only, Elektor Store # 140183- 4140183-91, see wnw.elektor.com 4002.0) Input 2 (-----ACMP_I2, PIOO_0 can be used as Input 1). Together with PIO0_8 or PIO0_9 you can Use it to create for instance a capacitive touch sensor. If P1 is mounted you will have a variable voltage on pin 2 of this connector. 3 provides access to the serial wire debug (SWD) port of the MCU. It shares this port with the rotary encoder but when in rest its contacts should be open (if you use a model with detents and it is, positioned between the detents). The pinout of K3 Is not conforming to any SWD pod standard that T know of so you will probably need an adapter of some sort. KS Is intended for digital 1/0. Its pin 2 (PIO0_6) has a current limiting resistor because this pin is not SV tolerant (all the others are). The other current limiting resistors are on ports that may be frequently connected and disconnected. 6 Is the full-speed 12C interface, connected to the pins 10 and 11 (and the center LED). (Of course most ports on the extension connectors are shared with the LED matrix and the other Table 3. I/O ports In relation to extension connectors and functions. PIO, K | Function PIO] K [Function PIO) K [Function 0 | 4 |RxD/Free 6 [5 [Buzzer 12 | 4 |ISP/Free 1 | 2 [Pi/Analog 7 [6 [Row 3 13 | 3 [Row 1 2_|-3 [SWOIO/sia & | 2 [Column i 14 | 5 [Column 3 3-3 [SWwcLWsiB ‘| 2 [Column 0 15 [1 [Row 0 4 [4 [txd/Free 40 |6_[tepi7R7FC 16 | 5 [Row 2 5 | 4 [Reset/Pushbutton a1 [6 [LEDI7G/PC 37 [4 [Column 2on-board peripherals so beware when you start wiring the extension ports to all kinds of other components. Table 3 summarizes the relations between the I/O ports, connectors and functions. From this table it Is clear that connector K4 Is meant for communication with the host appli cation. In ISP mode (i.e. when the MCU is reset while PIO0_12 Is being held Low) PIOO_o and PIO0_4 become a serfal port. In normal operating mode these pins are free and can be assigned to for instance the SPI or I2C module or to something else altogether. Once out of reset PIOO_12 becomes a normal 1/0 port. Note that PIOO_12 does not need a pull-up resistor because It already has one inside the MCU (as do all the 1/0 ports, except PIO0_10 and PIO0_11) so ISP mode will not be activated accidentally when this pin is left unconnected. The project software NXP has a website dedicated to the LPC microcontroller products [5] where you can find libraries for all of them and which saves you the work of creating the low-level hardware abstraction layer. These libraries come in pairs, one for the chip family itself and another for the evaluation and/ or demonstration board. Since this project uses the LPC812 we need to look in the LPC8xx packages, and since we use the LPCXpresso compiler tools (Figure 4) we have to pick the package for this tool. The libraries are included (version 2.01) in the download for this article (6). I urge to not replace them with freshly downloaded versions without doing an extensive file compare first, because I have corrected some bugs in my libraries that I have come across. The LPCXpresso workspace for CoCo-ri-Co contains two projects: Ipc_chip_8xx_Lib (the NXP Figure 4, The LPO(presso IDE used to:develop the CoCo-ri- Co software. Although you cannat see it inthis screenshot, the project named Ip. chip_8xx fib is also part of the workspace.Things to know about LPC812 1/0 ports ‘© In Table 3 the ports 10 and 11 are listed as IC without specifying a signal (SDA or SCK). ‘That's because it is up to you to decide which port carries what signal. See the Switch Matrix section in the LPCB1x User manuel. These ports have open-drain outputs and should not be used for fest SPI or UART modes. ‘© Ports 2, 3 end 5 default to their SWD and Reset function on power-up and must be reconfigured in software before they can be used for other purposes. Pin Enable Register 0 controls these pins. ‘In digital mode all ports are 5-V tolerant except for PIO_6. Ports that can have analog functions (PIO0_0, PIO0_1 end PIO0_6) become 5-V Intolerant when their anelog function is activated. «If the software disables the Reset input i is still possible to enter ISP mode simply by holding PIO0_12 low while power-cycing the MCU (off-on).. ‘The LPCBIx family comes in several packages. On early versions of the 16- and 20-pin packages the ISP mode entry pin was PIO0_1 instead of PIOO_12. If your MCU refuses to go into ISP mode, check the revision code. It must be “4C* or higher for PIOO_12 to work. ‘© A virgin LPCB1x MCU boots directly into ISP mode so you should be able to program the MCU at least once. library) and 140183_cocorico_digital_poten- ‘tioneter (the real stuff). Both are included in the downloadable archive that you can import in LPCXpresso without unpacking It first. The file nain.c contains the main engine that reads the encoder and sets the LEDs in a virtual matrix. In the back- ‘ground a 1-kHz systick timer takes care of actually lighting the LEDs and debouncing the rotary rs _ encoder. Buzzer beeps PNAS, are produced by yet Bete ess another timer (the aes Ae Multi-Rate Timer). et To save program = memory and because they are available the communi cations are handled as much as possible by the MCU's built-in drivers. T encourage you to have a look at the source code (in plain C) and play around with It. Copy the digital potentiometer project to a new name in the same workspace and modify it. It is not too difficult I hope and comments are plentiful. ‘When you build a project it will create a HEX file that you can program into the MCU using a serial port and the Flash Magic tool [7]. CoCo-ri-Co Is available as a ready assembled module from the Elektor Store. Bare PCBs are also supplied for those who insist on an all-DIY build at home or in the lab. Questions and con- tributions can be posted on the Elektor.Labs page of the project [3]. (140183:1) Web Links [1] J°B: Elektor September 2011, ‘www elektor-magazine.com/ 110274; www elektor-labs.com/node/3832 [2] Platino: Elektor October 2011, wwwelektor-magazine.com/ 100892; wwwelektor-labs.com/node/2288 [3] CoCo-ri-Co on Elektor. Labs: www.elektor-labs.com/node/4257 [4] Elektor 110553 Bo8-FT232R: www .elektor.com/110553 [5] LPCOpen: www.lpoware.com [6] Project downloads: www.elektor. comy140183, [7] Flash Magic programming tool: www flashmagictool.comby Andrew Retallack (Gouth Africa) T-Board 28 Lowest-Power Exercising Temperature Logger beats Arduino hands down With the Internet of Things increasingly driving MCU-based projects, power consumption is becoming a critical element of any design. Every milliamp counts when remote sensors need to run for months or years on a single battery. Elektor’s latest T-Board 28 is perfectly placed to help you prototype your next low-power project, as illustrated in this build of a power-efficient standalone temperature logger. Put on your t-shirt! ‘The T-Board 28 [1] has a number of features that give a designer the ability to optimize the power utilization by tweaking the design and then measuring the impact of these. This is one of the key benefits to the T-Board over other fixed platforms such as the Arduino. As many first-time users of the T-Board will have made the jump from an Arduino platform, I thought It would be relevant to show my own experimental approach using the T-Board 28 to pro- gressively reduce the power consumption of a temperature logger. The T-Board 28 comes with an ATMEGA328 micro fitted, and is the largest of the T-Boards series. The series and the t-shirt are in the Elektor Store. In order to become more familiar with ways of reducing power consumption, I wanted to work on a relatively simple project that would have a practical use without distracting me from my power-optimization goal. A temperature logger was a perfect fit. The project was bullt to measure temperature at regular intervals using 2low-cost temperature sensor, recording these ‘on an EEPROM chip. When connected to a PC (or a laptop in the field), the readings can be downloaded over a serial connection. SD Cards were considered in the initial design, but with significantly higher power consumption (up to 3.5 times) and lack of tolerance for 5-V supply, It was decided to use EEPROM instead. At a later stage the project could be enhanced to include RF transmission of the readings; power-optimization of RF Is of course a whole ‘other ball-game! A quick look at the project design {As can be seen from the schematic in Figure 1, the design is very simple (a major advantage of using a T-Board). A Microchip 24LC128P EEPROM was connected via °C to a T-Board 28, with the required pullup resistors on the I2¢ communi cation lines. An LM6O temperature sensor was selected due to its low-current drain, ability to easily record negative temperatures (for those readers and editors not lucky enough to live in sunny South Africa!), and importantly its support for voltage inputs starting at 2.7 V. Finally, an LED was connected to give a visual indication of the tasks the logger was performing. This helped when measuring the power consumption, and would of course be removed in the final version of the logger. The 241C128, as with all PC devices, has a device address. This address can be set by tying pins 1, 2, 3 (AO, Al, A2) either High or Low. The 7-bit address of the device starts with a fixed ‘1010 followed by the logic levels on pins A2 to AO. For this design (with AO tied High) the address is therefore 1010001. When you wire the chip ‘onto the breadboard, pay careful attention to all pins—I spent hours debugging my code, only to find I had tied pin 7 (the Write Protect pin) to Vec instead of V;; (GND) effectively write-pro- tecting the chip! Datasheets are easily misread. Serial 1/0 (for debugging and producing a dump of stored readings) Is carried out over the FTDI pins of the T-Board, using the ATmega328’s, built-in UART. An FTDI cable or module, connected to the T-Board's FTDI connector, allows the ‘T-Board to be connected to a PC's USB port. Fur- ther iterations of the design could include porting the code to the T-Board 8, which would require a software-Implementation of the Serial protocol. The program code The code was written using Atmel Studio 6.2, the latest version of Atmel's in-house IDE. Sim- ple modules were written to handle the ADC (Analog-to-Digital Converter), TWI (Two-Wire Interface), and UART functions. Here we need to digress briefly to discuss terminology. Due to trademarking around the [°C protocol, Atmel have called their 12C-compatible protocol TWI. It is 100% compatible with I2C, but Is just not called I2C. Therefore in the datasheets and reg- Ister naming you'll be looking for TW1, not I°C. The code Is available for downloading at (7) It should be straightforward to follow, however a few notes may be needed for the compiler symbols (accessed through the Symbols option of the AVR/GNU C Compiler section of the project properties). In order to calculate communication speeds and interval timing accurately, F_CPU defines the CPU speed in hertz. System voltage 's defined in millivolts by SYSTEM_MILLIVOLTAGE, to allow ADC readings to be correctly converted into voltages. A DEBUG symbol exists that will ‘output debugging information on the serial port if set to When the MCU powers on, it first displays a simple menu over the Serial port to allow the user Figures. *Crcugran” of our exercise aimed at geting T-Board 28 to draw as ite current 25 possible while acting as 2 fll blown Temperature Logger.Pete iM Fuse words are used by AVR microcontrollers to configure 2 range of functions / options. They need to be flashed by ‘an external ISP programmer, and once set they generally ‘cannot be modified by the program running on the AVR (with a few exceptions). They are often approached with great nervousness—and sometimes rightly so as setting invalid combinations can result in @ “bricked” (no longer functioning) MCU. To “un- brick” the MCU, a special programmer is needed that operates at higher voltages. ‘There are three fuses on the T-Board 28's ATmega328—high fuse word (hfuse), low fuse word (fuse) and extended fuse word (efuse). For our experiments, we will primarily alter the fuses that determine the clock source for the MCU—the Muse. AVR microcontrollers can run off a range of internal and external clock sources, and we need to use the fuses to configure the MCU to use the ones that we want. In order to calculate the values of the fuse words, you can wade through ‘some fairly cryptic datasheet text— or use a fuse calculator. There are ‘a number of online ones (I like the Engbedded one [4]) as well as ‘apps you can install on your tablet/ ‘smartphone. Setting fuses using a programmer ‘that is supported by Atmel Studio is very straightforward ~ you simply center the fuse values in the Device ‘Setting Fuse Words in Atmel Studio Programming Dialog like so: If your programmer is not supported natively by Atmel Studio, then you will need to use AVRDude [5] to set the fuses: 1. Open a command prompt and navigate to the folder that AVRDude.exe is stored in 2. Enter the following to read the Fuse settings: avrdude -c -p -U Lfuse: 3. Enter the following to program a Fuse: avrdude ~c ~p -v -U Lfuse:w:0xFF:m where: is AVRDude’s code for the specific programmer you're using (e.g. usbtiny for the USBTinyISP}— refer to the file avrdude.conf for a list of these; is AVRDude’s code for the specific MCU being programmed (e.g. 'm328 for the ATmega328); “Lfuse is the fuse being programmed—use fuse, hfuse or fuse as appropriate; OXFF is the value of the fuse being ‘set—change this as necessary. Finally, if you want to experiment ‘more extensively with fuses, have @ ‘search online, there are a number of {good resources. Timer2 is an 8-bit timer, so can only count up to a value of 255. Each clock cycle adds to the timer until it reaches 255 where it can be configured to generate an interrupt. As you can see, if the Timer is running off a fast clock it won't take long to reach 255 and generate an interrupt: a 1-MHz clock will reach 255 after only 256 microseconds! Fortunately there are @ couple of ways to slow the timer down. The first is to use a prescaler that divides the clock speed down further. The maximum prescaler available is 1024, which will result in @ count of 255 being reached after 0.26 seconds. ‘An improvement, but still too quick. Li: Bei ‘The second is to slow the clock speed down. We can’t slow the main clock down, as we then aren't able to ‘communicate over the serial port reliably. Timer2 however can work off an external crystal oscillator as well—in asynchronous mode (asynchronous because the external oscillator is not synchronized to the main CPU clock). The Useful thing here is that very slow crystals can be used—I used a watch crystal at 32.768 kHz. With this slower crystal, the timer reaches 255 in 7.97 seconds, much more useful! Additionally, a slower crystal of course uses less power.to print the log in comma-delimited format, or to clear the log. The program waits for 10 seconds to allow the user to make a selection, before entering the normal logging loop. ‘There are two versions of the code: Version 1 (Figure 2) shows the completely unoptimized code, while Version 2 (Figure 3) contains the far more power-efficient version. The optimization in the final version has been reasonably limited in order to ensure the code remains readable and easy to follow; there are certainly small additional power-savings that can be realized, and Convert Reading to Temperature Write time & temperature to EEPROM Flash LED 10 times Initialization: ART ™ EEPROM yo Pins Figure 2. Flowchart of “version 1”, the entirely palatable ‘but unoptimized firmware Figure 3. Flowchart of ‘version 2", the F-Board 28 firmware optimized for current draw.The O; imization Experiments L, Estabteh 2 Basetine Once the project is connected on a breadboard as set ‘out in in Figure 1, the T-Board Is initially configured in a similar way to an Arduino Uno. The input voltage should be set at 5 V using the voltage-selection jumper, and a 16 MHz external crystal oscillator connected. As we need to measure the current we'll remove the jumper completely and connect the DMM to. the voltage-selection pins (black to the center pin, and red to the 5-V pin). In this configuration the DMM needs to be tumed to the Amps setting in order to complete the circuit. Next, the fuses need to be flashed to the values in Table 1. Finally, the “version-1" code (software download at {7]) needs to be loaded Into Atmel Studio, complled and then flashed to the microcontroller. This version uses a simple delay loop (implemented in the _delay_ms() function) to time the delays between readings. The flowchart in Figure 2 gives ‘an overview of the code. ‘The initial set of measurements show a consumption of ‘approximately 12.4 mA while idle. From the code it can bbe seen that the LED flashes 10 times before entering idle state— measurements are taken once the DMM reading has stabilized after these 10 flashes. AA further single flash of the LED indicates that the temperature reading is about to be taken and written to the EEPROM— however the time taken to read the temperature turned out to bbe too short to register a meaningful reading on the OMM. Z__ Reduce the Input Voltage The simplest optimization yielded the greatest savings. ‘Simply switch the voltage-selection jumper from the 5 V position to the 3V3 setting, and the T-Board now operates off 3.3 V. In ‘our experiments of course we won't use the jumper, so switch the red lead of the DMM from the 5 V pin to the 3V3 pin. ‘The code needs to be changed so that the temperature calculations work off a reference voltage of 3.3 V; we need to change the compiler symbol SYSTEM_MILLIVOLTAGE. To do this, right-click on the project in the project browser and sselect properties. Then, on the Toolchain tab, click on Symbols under the AVR/GNU C Compiler section. Change the SYSTEM_ MILLIVOLTAGE from 5000 UL to 3300 UL (the UL ensures the value is cast as an unsigned long). We then compile and flash the new code to the MCU. ‘The measured current at 3.3 V reads 5.76 mA, a reduction of 6.6 mA: that simple change has more than halved the power ‘consumption of the logger! Slow that CPU 3, 4, 5+ neem re eens ‘examine the impact of slowing the CPU down, using a ‘combination of swapping the external crystal oscillator ‘and dividing the clock speed. With each change made, the ‘compiler symbol F_CPU needs to be updated to reflect the effective clock speed in Hz as listed in Table 2 (Note the ‘spaces in the F_CPU column are there for ease of reading here, and should be removed when entered in Atmel Studio). For each of the rows in Table 2, follow the steps highlighted earlier: connect the programmer, update the fuses, change the F_CPU in the code, compile and flash the code to the MCU, disconnect the programmer and then switch the crystal if needed. ‘As can be seen there is a clear correlation between the CPU ‘speed and the current usage. Looking at the current usage at ‘500 kHz it looks like this is the path to choose. Unfortunately, the ATmega’s USART is not able to reliably function at CPU ‘speeds as low 2s S00 kHz (refer to the datasheet as well as [6] for useful BAUD rate accuracy calculator). So we need to go back a step. 6 Use the Internal Oscillator Next I looked at the impact of using the internal oscillator at 8 MHz, and dividing by 8 to get an effective CPU speed of 1 MHz. This speed is sufficient to achieve reliable serial communications at low BAUD rates up to 4800—good ‘enough for the simple menu at startup. Once the LFuse bits have been set to 0x62, the F_CPU symbol amended to 1000000UL, and the compiled program flashed to the MCU, then the external crystal can be removed. Current draw was now slightly higher than at the clock speed of S00 kHz, but at 748 WA a great improvement on the original 12.36 mA! 7 Disable BOD ‘The ATmega328 has a Brown-out Detector (BOD), which allows a developer to guarantee consistent operation of ‘a project by putting the MCU into a “reset” state if the input voltage drops below a specified level. However, this feature ‘comes at a power cost. By setting the EFuse we can disable the BOD and squeeze those extra microamps out of the project. Set the EFuse to 0x07 to disable the BOD, and you'll note a small ‘saving of microamps—not a large saving, but every bit counts. 8 Sleep It’s now time to take a look at optimizing the code. In the version 1 (Figure 2), we used a delay loop to time the temperature readings. While it may seem like a delay loop is not doing anything, the CPU is in fact spinning its ‘cogs counting the delay out. This of course uses power. Most microcontrollers allow you to put them to sleep (effectively turning the CPU and certain peripherals off) in order to save Power, and then wake them up again when needed. Interrupts are used to wake the CPU up—elther based on changes on particular pins, incoming communications, or specific elapsed time. The AVR is no exception, and has 6 different sieep modes with differing levels of power saving. I won't go into these in detail here, other than to say that I'll use the mode called (imaginatively!) “Power-save". This mode allows us to use Timer2 to wake the CPU up after a specific interval— ‘Timer2 being the only timer that can in fact wake the CPU. In order to save on power and work with more manageabletimings, Timer2 Is set to run off an external watch crystal. Refer to the text box “Timing the Timer". ‘To put this experiment in action, no Fuse bits need to be set. ‘A 32.768 kHz watch crystal needs to be inserted carefully in the T-Board 28—these leads are very thin so make sure that you don’t damage them, while at the same time ensuring a ‘900d contact with the header on the T-Board. ‘Then load, compile and flash the Version 2 code onto the MCU. Version 2 does less flashing of the LED, just four flashes at the tend of each temperature reading. I took my readings once the current stabilized after the four flashes, and was delighted to ‘see a measurement of only 59 JA. This kind of current draw should allow an 800-mAh battery to power the project for nearly a year and a half. But there Is still more we can dot ©, Shut Down Unused Peripherals = The ATmega328 has a register that allows you to power off unused peripherals - the Power Reduction Regis- ter (PRR). The function reducePower () disables the Timer0, Timeri, SPI and the USART. The TWI and Timer2 are left enabled as we need them to communicate with the EEPROM ‘and wake the MCU from sleep respectively. The ADC Is already ‘managed as part of the conversion process, so no further ‘optimization needed there. Additionally this function enables the pull-up resistors on the unused pins to prevent them from floating—floating pins can result in additional current draw. With the DMM I was using I wasn’t able to see any additional power saving from using these settings—looking at the datasheet percentages it seems that the real saving 1s achieved when the CPU Is in active mode as the percent- age savings are quite low. nd A Final 10. Change ‘After some comparison to the datasheet I wasn't happy with the power consumption Twas achieving. Of course the MCU isn't the only element drawing ‘current—the EEPROM and the LM60 also have an impact. A few ‘quick tests, and references to thelr datasheets, and I noted that the EEPROM only draws 100 nA in standby mode while the LM60 has ‘a typical quiescent current draw of around 82 WA. By altering the circutt ‘so that the LM60 was powered off the T-Board's PD7 (well within the pin's ‘maximum output current of 20 mA), ‘and modifying the code to power the sensor only for the period needed to take a reading, I managed to get the current draw down to an incredible 1 yA (the lowest my DMM could measure). After seeing this reading T popped the champagne and considered the experimental process a success! I took a few additional measurements, and arrived at an estimated peak current draw of approximately ‘65 WA during temperature measurement. I rounded this up to 1 mA for good measure, and then by calculating a few averages I arrived at an estimated battery lifespan of six ‘years on an 800 mAh battery. Six years on 2 single battery Is ‘more than enough for our remote sensor. Experiment | Description cpu mituivottace | LFUSE | EFUSE on Tema Sv FEO SO lass EES a oe Se aoa a a 4 Use 4 MHz External Crystal 4 000 000 UL 3300 UL OxFD ox0S | 2.130 5 Divide Clock by 8 500 000 UL 3300 UL 0x7D oxos | 0.560 6 Internal 8 MHz oscillator + Divide by 8 _| 1 000 000 UL 3300 UL 0x62, 0x05 | 0.748 $—|aane oon tsoouno Uses ur —— [etfs an a ee creas ——|reoseso sara ——| nests FS eS a — [tonto [soos — [estos fos Experiment | Crystal Divide by 8? Effective clock speed F_cPU LFUSE Current ee ee oe a [ime — fre eo a se —fe = a aDar ar ttestr98 aww teestr ing send 3 m8 ay User Se sends tO ayaa even tee 102 wait WRITE = we nee vice Aaarest Seagrass 17 send me S00 cep OEE I result = 120 con 8) . xo cand f pingacess MO j1send Newory ceation ADHTeSS Web Links [2] Atmel Studio: www.atmel.com/atmelstudio [1] T-Boards 8/14/28, Elektor September 2014, wonw.elektor-magazine.comy 130581 [3] USB Tiny: www.crash-bang.com/using-usbtiny-with-atmelstudio/ [4] Fuse Calculator: www.engbedded.com/fusecalc/ [5] AVRDude: http://savannah.nongnu.org/projects/avrdude [6] Baud Calculator: www.wormfood.net/avrbaudcalc. php [7] Version 1 & 2 download: www.elektor-magazine.comy 140413 1 challenge readers to take it further and share their thoughts on www.elektor-labs.com and http://forum.elektor.com (Topic: Microcontrol- lers & Embedded). T-Board helps with optimization In order to reduce current consumption on a ‘microcontroller project, there are (at a basic level) four common areas to focus on. The first Is to reduce the input voltage, resulting (simplistically) In a physical illustration of Ohm’s Law. The sec- ‘ond Is to reduce the clock speed inet condition > ay EEPROM FIRST ADORESS resutt = 12¢-sen¢ reat $608 LU) scges LH vy vacation nsress, 2 (Ee ope I! 11 5end Meme end (uint® EEN result = 120 of the CPU, as less clock cycles per second result {n lower current draw. The third is to put the CPU Into a sleep mode when It isn’t needed. Finally, disabling unused peripherals and components will result in further reductions in power. ‘The T-Board 28 has been designed with the flex- ibility to help with the first and second items dis- ‘cussed above: the ability to select input voltage, and a means to switch out crystal oscillators to. change the CPU clock speed. The second and third Items are implemented in the pro- ye veegram code. Finally, the voltage selector of the ‘T-Board acts as useful point to measure the current usage in order to assess the effectiveness of the optimizations. In addition to the items outlined above, a number of other efficiencies can be gained—for example more efficient coding, which results in fewer lock cycles to achieve specific tasks, will reduce the current requirements. As discussed earlier, the code efficiency has not been a focus of this project—optimizing your coding in C/C++ isa whole series of articles on its own! Measuring the current consumption It Is assumed that many readers using the ‘T-Board (Including myself) may not have access to 2 fully kitted lab. Therefore I used a relatively crude, but nevertheless effective, means of measuring the current usage—a low-cost digital multimeter (DMM). The DMM was switched to measure amps, and the probes then connected (using female connectors) to the voltage-selection header pins on the T-Board. This allowed me to measure the current usage of the T-Board at point in time. The positions of the jumpers in the circult Isolate the draw of the project, and are not impacted by the power supply circuitry. In order to measure the current more accurately, I desoldered and removed ‘the power-indicator LED as I found Its current draw far outweighed that of the MCU Itself. This also reduced the consumption sufficiently at higher levels of optimization to enable me to switch my DMM to the A setting, allowing me to better measure the Impact of some of the smaller changes. Of course the limitation with this measurement method (apart from a reduced level of accuracy) is that the current Is only measured at a point in time. Ideally one would like to record the measurements over time in order to calculate an average current draw (this Is in itself a useful project to include in a future edition of Elektor). This is where the LED connected to PBO came In useful as an indicator of the activity of the program, and therefore a context to the current being measured. The experiments ‘The spread on the previous pages of this article focus on the experiments carried out to reduce current draw, detailing what changes were made and the measured impact of these. In order to implement each of the progressive optimizations, these steps were followed: 1, the ISP programmer was connected and any changes to fuses made; 2. any changes to the code were compiled and flashed to the T-Board; 3. the ISP programmer was disconnected; 4. any physical changes to the circuit were made; 5. a 9-V battery was connected to the 2.1-mm Jack and the current measured with the DMM. Any FTDI cables/boards should not be connected when the measurements are taken. The results of the various experiments are sum- marized in Table 1, along with the changes made. Before we get started, @ quick note on the programmer: Atmel Studio [2] supports a number of ISP programmers natively, which require little or no special configuration. There are also a number of other programmers (usually available at a lower cost) that are not natively supported but can be configured to work under Atmel Stu- dio (eg. USBTiny, USBasp, etc.). Apart from the flashing of the program onto the MCU, a key difference Is the way the fuses need to be set (Refer to the text box “Don’t Blow Your Fuse” for more information). There are various resources online explaining how to configure Atmel Studio to support these third-party programmers [3]. Next Steps I found this process enormously helpful, and it really highlighted to me the benefit of a flexible, lean & mean platform like the T-Board. Overall, consumption was reduced from an Arduino-class 12.3 mA to an average of 13 WA (averaged based on time spent in sleep vs. active mode); a reduc tion of 99.89%! I challenge other readers to take this to the next level—perhaps migrating to the T-Board 8, or Including an RF sensor to transmit the readings. Either way, with your t-shirt on. (140613)Programmable Oates nat- ae ha Brings (blue) light into the darkness eee Co) So osmeatheen lesions * proaching rapidly. So it has become Deon Ce aoe ek Vi Cec ba al unusual mood lighting for the YA eee ete eee eee’ Pee es oS Ceo rT sorts of beautiful pat- PeeteElectronics hobbyists are always keen to make a festive gadget for the celebrations around this time of the year. Not something read-built from a garden center or home improvement store, but a circuit which has something special and that you can’t just buy anywhere. Over the years Elektor have published numerous Christmas circutts, the majority of them projects that contain a number of LEDs which provide the ambiance. This year, in conjunction with the printed circult board manufacturer Eurocircults, we have designed a very nice circuit for you, which you can bulld yourself of course if you have the time and inclination, but we also offer this Christmas tree ready-bullt in the Elektor web store for an attractive price, You can always tell your visitors that you soldered the teeny-weeny SMD components in the little tree yourself with the ald of a microscope and a very small soldering iron during the annual Christmas function at your work... The Iittle Christmas tree that you see here is provided with no less than 62 bright blue LEDs on the front that, with their color alone, will already generate a pleasant Christmas atmosphere. The control electronics is on the back, mainly comprising a powerful microcontroller. True, we have had other flashing Christmas trees in our magazine before, but this particular specimen has an ‘extensive light show built in, which causes all Kinds of patterns or texts to ‘run’ across the LEDS. A number of patterns have already been pre-pro- ‘grammed, but a great feature is that you can very easily design your own pattems and effects using your PC and then simply store them in the memory of the Christmas tree via the USB connection. Powerful microcontroller ‘An ARM Cortex-MO microcontroller from STM has been used for controlling the LEDs (see the schematic In Figure 1). This relatively cheap 32-bit Figure 1 ‘The schematic for the Christmas tree mainly consists of a microcontroller ‘and 62 blue LEDs.Figure 2. ‘The circuit board is double- sided and screened on the front with a so-called PCB-Pixture solder mack [1] depicting snowflakes. (60% of re ste) Component List Resistors RI-RIG = 62-1000 (SMD 0805) RA7-R24 = 1009 (SMD 0603) R25-R2S = 10kQ (SMD 0603) Capacitors C1,C2 = 10uF 6.3V tantalum (SMD A) C3°C9 = 100nF 40 ceramic (SMD 0603) Semiconductors LED1-LED62 = LED blue (SMO 1206) 1-94 = FOV304P (SoT23) ICL = STMa2F042Cs (LOFPA8) IC2 = W2SQ80BVSNIG (508) ICG = LO1117833TR ($0223) Miscellaneous USBI = micro USB connector, PCB mount (47346-0001) ‘SW1 = pushbutton, PCB mount (TACTB-64K-F) PCB (files only) # 140371-11 microcontroller offers, besides much computing ower, an integrated USB2.0 interface. This allows for a very easy connection of the microcontroller to a PC. Thanks to the built-in bootloader its very straightforward to download new LED-patterns or new firmware. In order to provide the user which a generous amount of space for patterns or animations of their own design, a serial flash memory with a capacity of 1 MB (8x1 Mbit) Is ‘connected to the microcontroller. This may appear ‘small compared to the typical size of memory sticks and memory cards these days, but nevertheless provides sufficient space to store more than 90,000 complete LED patterns. The LEDs are driven in a 4x16 matrix (not entirely, 3x16 + 1x14). Therefore 16 (14) LEDS are driven alternately by port pins PO through P15 of the controller. The four columns are con- tinually multiplexed. Because of the persistence of vision of the human eye, this fast switching is, not noticed and It appears that all the LEDs are ‘on continuously. The LED columns are switched by four P-channel MOSFETS of the type FDV304P. ‘The cathodes of the LEDs are connected via current-limiting resistors (R1 through R16) directly to the port output pins of the microcontroller. ‘This saves a large number of components, but we do have to take into account that the maximum Ce et ea Ce et eetotal current that the microcontroller can handle {Is only 80 mA, which therefore means a max\- ‘mum current of 5 mA per LED. If we use LEDS with sufficient efficiency then they will still be very visible in a reasonably bright environment. The entire circutt is powered from the micro-USB connection. A 3.3-V regulator (IC3) provides a regulated power supply voltage for the microcontroller and the LED matrix. Figure 2 shows the printed circuit board design for the little tree. The LEDs are on the front, the remaining components including the micro-USB ‘connector are on the back. For this occasion the circuit board Is screened with a PCB-Pixture solder mask [1] depicting snowflakes. For those of you who are keen to etch the circuit board yourself and populate it with components we feel obliged to warn you that some of these ‘SMD components are difficult to solder by hand! The source- and hex code files are available from the Elektor website [2]. The ARM software is written in C using the Kell compiler. The largest part of the code is related to the USB communications and this is based on the ST firmware library. The bootloader mode will be activated when the pushbutton on the back is pressed while the tree Is connected to a PC. It Is then possible to download or update the firmware with the ald of the ST software DfuSe [3]. Software By default, the ready-built Christmas tree from the Elektor shop Is already provided with a wide range of animations. If you are happy with those then there Is nothing more you need to do. If, however, you would like to create your own patterns, animations or running texts and whatever else you can come up with then we direct you to a special interactive web page that we have bullt for you where you can indulge yourself [4]. This page contains a picture of the Christ- mas tree with all its LEDs. You can make your ‘own animations by using your mouse to click the LEDs on the screen on or off. This way you create a pattern, where you can also select the desired duration and brightness. All generated patterns appear at the bottom of the screen next to each other. There is also the possibility to enter a scrolling text (text symbol at top left). In this way you can build your own animations. You can store an animation for later use. Using the preview button you can view the animation on the screen. Program your own animations and win! Cee eee ea ea ee eat an Seen a eee ee aoe: Cee ena ee Cen Perea en Pee Seana 2015—winners will be notified personally. You can make multiple of these animations and ‘thread’ these together into one long light show. Once that is done you can download the entire sequence to the Christmas tree. ‘The tree is connected to a PC using a micro- USB cable. When you do that for the first time Windows will install a driver. Windows recog- nizes the little tree as a HID peripheral. After that you can ‘flash’ the new animation to the tree from the web page. For use in the living room you can connect the tree to a 5-V wall adapter which Is fitted with a micro USB cable. Enjoy and many happy returns. Figure 3. ‘The components mounted on the back ofthe Christmas Web Links [1] www.eurocircuits.comYblog/171-PCB-PIXture-launched [2] www.elektor-magazine.com/ 140371 [3] www.st.comy/ web/en/catalog/tools/FM147/CL1794/SC961/$S1533/ PF257916 [4] www.eurocircuits.com/x-masVariLab 402 (2) Control and display board By Ton Giesberts (Elektor Labs) The control portion of the VariLab power supply uses a microcontroller to set the desired output voltage and current and to measure all significant voltages and We designed a separate control board to configure and monitor the output of the VariLab power supply board described in the previous issue and display the output voltage and current ‘as well as other data. Its main components are an ‘Atmel ATxmegal28A4U-AU microcontroller and ‘an LCD module with four lines of twenty characters. The microcontroller has onboard ADCs ‘and DACs for setting and measuring the output voltage and current, so the component count ‘can be kept relatively small. Circuit description ‘As you can see from the schematic diagram (Fig- ure 1), the circuit is built around an ATxmega microcontroller (IC1). The balanced supply volt- ‘ages for the entire circuit (#5 V) are taken from the supply board. The microcontroller operates from a 3.3 V supply voltage, which Is provided by voltage regulator IC3. All measurement and control signals are present ‘on connector K3. The circuit also receives the necessary supply voltages through this connector. The sense signals for the output voltage and ‘current of the supply board are fed to the ADC3 currents in the circuit. All measurements and settings are shown on a four-line LCD module. Naturally, there is also a USB port for communication with a PC. ‘and ADC4 inputs of the ATxmega, with the current sense signal first buffered by IC2d. Resistor RS is dimensioned so that the range of the sense signal for the power supply output voltage is virtually the same as the measuring range of ADC3. ‘The voltages Vsmps’ and 48Vin are also routed to the control board and measured by ADCS and ‘ADCS. Resistor R6 Is dimensioned to apply 5% of Vsmps’ to the ADCS input so that the input signal level remains within the measuring range. Diodes D2-DS protect the ADC inputs against overvoltages and negative voltages. ‘Two signals are necessary to set the output volt- ‘age and maximum output current of the power supply: 0-4 V for the voltage and 0-2 V for the current. The reference voltage for the ADCS Is provided by an external 2.5 V reference source {(IC6, for which we chose the old standby LM336) because the internal reference voltage of the ‘microcontroller is not sufficiently accurate. The reference voltage can be adjusted precisely with 1, but this trimpot can usually be left at its maximum setting because the calibration can also bbe done in software. The output voltages of the DACs are amplified by a factor of 2 by a pair of ‘opamps (IC2A and IC28), which in theory yields‘a maximum control voltage of 5 V. ‘A disadvantage of using the internal DACs is that the output range does not extend all the way to OV (the data sheet states a minimum value of 0.15 V). Since we definitely want to be able to adjust the output voltage down to 0 V, an offset voltage of 0.3 V Is applied to the two opamps to ensure that the control voltage range on the ‘opamp outputs can start at 0 V. The 0.3 V offset voltage Is derived from the 2.5 V reference volt- ‘age by R11/R12 and buffered by 1C2c. For the opamps we chose a Microchip MCP6004-E/ SL quad device, which features rail-to-rail inputs and outputs. This IC Is powered from +5 V and -0.6V (the latter via 08) because its maximum rated operating voltage is 6 V. This arrangement allows the output voltage to be adjusted all the way down to zero. The four-line LCD module is driven in 4-bit mode by the ATxmega. The display Is powered from 5 V. Although the logic high level of the drive signals from the microcontroller is only 3.3 V, that Is adequate for reliable operation. The display contrast can be set with P2. The microcontroller can switch on the display backlight via T1 when a jumper is fitted on JP3. Since the voltage reg- tlator on the supply board that provides the+5 V supply voltage does not have an extra heat sink, It Is not advisable to leave the backlight on all the time. In the current version of the firmware the LCD Is only illuminated when current limit ing Is active. The backlight can be permanently disabled by removing the jumper at JP3. A pair of rotary encoders with built-in pushbuttons, (S1 and $2) and a separate pushbutton ($3) are provided for the power supply operator interface. ‘The separate pushbutton can be used for several functions, including quickly switching the power supply on or off. By default the output voltage is set with S1 and the output current is set with $2. Figure 1. ‘Schematic diagram of the contral and display board. All ofthe measurement {and computation tasks are handled by an ATkmegat28 ‘microcontroller. J J tr sarnerouscatrouci| ostnctnnccenttinsomnescy ne eneiaaas| im ‘Aamo 20ANUAUComponent List Resistors Default 0.125W, SMD 0805) RI-R6,RG,R14,RI6,R22-R25 = 1k 1% RS = 2.87kO 1%, O.1W L.BOKA 1% 10MR 1% 10k0 5% Rio = 8209 1% Figure 2. Rid = 8.870 1% The component layout of the | Ri2 = 1.21K0 1% double-sided control board. | R43 = 1.30k0 1% RLS = 150 1%, 0.25 Ri? = 100 5% 18,20 = 2.240 5% Ri = 470.5% R21 = not fitted RDG = 33K0 5% RI7 = 56K 5% PI = 5008 10%, 0.5W, 25-tum preset, top adjustment (Bourns 3296Y-1-SO1LF) P2 = 10kR 20%, 0.15¥, preset, top adjustment Capacitors C1,C2 = 229F SOV, 5%, SMD 0805 COG/NPO C3..C9,C13 = L00F 25, 10%, SMD 0805 X7R C10 = 10UF 63V, 20%, diam. 6.3mm, 0.1" pitch C11 = 100uF 25V, 209%, diam. 6.3mm, 0.1” pitch C12 = 4.7UF 6.3V, 10%, SMD 0805 tantalum (case R, AVX TAJRE7SKOOGRNJ) Inductors, L1,L2 = 3309 © 100MHH, 0.089 1.78, SMD 0603 Semiconductors Di = LED, red, 3mm, wired 1D2,03,D4,05 ~ BAS70-04 (SMD SOT-23) D6 = PRTRSVOU2X (SMO SOT-143B) 7,08 = TS4148 RY (SMD 0805) Ti,72 = 8C817-25 (SMD SOT-23) Ici = ATkmegai28A4U-AU (SMD TQFP-44), programmed, Elektor Store #120437-41) 1C2 = MCPGOO4 (SMD SOIC14) IC3 = KF33RDT-TR (DPAK-3) ICA = 4N25 (DIP-6) ICS = LM35 (10-32) C6 = LM3362-2V5 (70-92) Miscellaneous Kt = USB connector, type B, PCB mount, wired, vertical model (Lumberg 2411 01) K2 = 6-pin (2x3) pinheader, 0.1" pitch KS = 14-pin (2x7) pineader, 0.1" pitch Ka = 2-way screw terminal Bock, 0.2" pitch KS,K6,1P4 = 3-pin pinheader, 0.1" pitch K7 = 4-pin pinheade, 0.1" pitch 3P1,302,1P3 = 2-pin pinheader, 0.1* pitcn jumpers for 31,12,3P3,3P4 $1,S2 = 18PPR rotary encoder with integrated Pushbutton (e.g. Alps EC1iE183440C) 3 = pushbutton, PCB mount (€.9. Multimec RASFTLS) 3 = cap for pushbutton, h=19mm (e.g. Mutimec 1503-19.0) B21 = piezo buzzer, 1.5-16 Voc, max. 8 mA, diam. ‘4mm, 7.62mm pitch Xi = 16MHz quartz crystal, C, 18pF; dim. § x 3.2mm. (Abracon ABM3-16.000MHZ-B2-T) LCD1 = 4x20 characters, with bacight, Elektor Store #120061-73 pce # 120437-2Buzzer B21, which is driven by the microcontroller through T1, generates an acoustic signal in situations such as current limiting. Indicator lamp LED D1 Is lit when the power supply output Is enabled. Temperature sensor ICS measures the temperature In the enclosure, which can be displayed on the LCD. The sensor is located directly on the PCB, and when the board Is installed as Intended in an enclosure the sensor Is close to the heat sink on the supply board. For other applications the sensor can also be mounted somewhere else and connected to the board by a few wires. There is also a galvanically isolated output (IC4) avallable for switching purposes, although this capability is not used in the current version of the firmware. USB port for connection to a PC Is additionally present. This can be used to send measurement data to the computer, or the voltage and current can be set from the computer with the ald of a terminal emulator program. The USB port is normally connected directly to the microcontroller over 3P1 and 3P2, but if desired it can be fully Isolated by removing these two jumpers. Diode D6 protects the microcontroller against potential overvoltages on the USB port. Voltage divider R26/R27 Is included to prevent the microcontroller program from hanging if the USB connection Is suddenly broken. The microcontroller can be put into programming mode (ISP or PDI) by fitting a jumper on JP4. POI Is usually the better option, and it Is also supported by the Atmel AVRISP mkII programmer. With both settings the microcontroller is programmed over connector K2. The voltages Vset, Iset and ground are present on connector KS for convenient measurement during calibration. If the control board Is used in a project other than the VariLab power supply, the balanced +5 V supply voltages for the control board can be connected to K6. The maximum current consumption is +50/-5 mA with the backlight off or +105/-5 mA with the backlight on. Finally, K7 provides access to a number of free 1/0 pins of the microcontroller (PAO, PAL and PA2). Construction The PCB layout shown in Figure 2 Is compact and just big enough to hold the four-line display module, the rotary encoders and the pushbutton (see also Figure 3). Components are mounted on both sides of the board. Most of the components are on the component overlay side, Including all of the SMDs, connectors and leaded components. The only components mounted on the other side are the pushbutton, the two rotary encoders, the vertical USB connector and the Indicator LED, and of course the display module. Be sure to use the components shown in the components list when assembling the board. Substitutions are not allowed. The comments in Figure 3. Top and bottom sides of the fuly assembled control board. All of the controls ‘and indicators, including the LCD module, are located on the same side,Figure 4. The LCD module is secured to the PCB by a pair of M3 screws and nuts, with extra ruts forthe right spacing. the previous article (for the supply board) about ‘mounting the SMDs also apply here: you need a certain amount of experience in order to solder these devices properly by hand. If you do not have sufficient experience, we recommend that you do not start on this project or that you get help from a more experienced electronics enthu- slast or technician, ‘Mounting the components on the side with the encoders, pushbutton and display module is eas- ler because they are all conventional components. Use short wire bridges or a 16-way pin header to mount the LCD module a few millime- ters above the PCB, so that It cannot touch any protruding wires on the control board. Secure the LCD module to the PCB with M3 screws and nuts in the two holes at the bottom edge of the module. A pair of nuts between the module and the PCB provide exactly the right spacing (see Figure 4). Screws are not necessary at the top edge of the module, since the connecting wires or the pin header provide the right spacing and firm attachment. Once all the components are mounted, you can fit jumpers at positions P1 and JP2 for the USB ort, 3P3 for the LCD backlight, and P4 for PDI programming mode. Now the board Is ready for use with the supply board. Use a short 14-lead flat cable to Join connector 3 to connector K3 on the supply board. Keep this connection as short as possible to avoid inter- ference. No other connections between the two boards are necessary. Actually you don’t have to Join the boards together just yet, since we plan to tell you how to mount the boards In a sultable enclosure in next month's Issue of Elektor. {As usual, there are two options for programming the microcontroller. You can order a prepro- grammed microcontroller from the Elektor Store (item number 120437-41; see [2]), or you can rogram the microcontroller yourself from Atmel Studio after you assemble the board. For the latter option you need a sultable programmer, such as the Atmel AVRISP mkII, connected to programming connector K2. As usual, the source ‘code and hex files (120437-11) can be downloaded free of charge from the web page for this. project [2]. Next month we will tell you more about the firmware, fitting the boards in an enclosure, and power supply calibration and adjustment. (140373) Web links [1] www.elektor.com/ 120437, [2] www elektorcom/ 140373 [3] www.elektor-labs.com/120437‘By Martin Ossmann_ (Germany) Figure 1. ‘The Red Pitaya module is very compact. Figure 2. Block diagram of the Red Pitaya module. Red Pitaya Not just a USB scope module Red Pitaya is the first open-source instrumentation platform that can be configured for various measurement tasks. Ready-made apps are available on the Web for downloading, but you can also program your own specific measurement processes. Several potential measuring tasks and demo programs are described in this article. To get an idea of what the module (Figure 1) can do, have @ look at the block diagram in Figure 2. The central component is a Xilinx 2C7Z010 system-on-chip (SoC) device. It contains a dual- ea “SH ‘core ARM9 processor clocked at 800 MHz, which hosts a Linux distribution on the Red Pitaya board. The SoC also contains an FPGA with 28k logic cells, a bunch of block RAM (BRAM), and several DSP slices. The FPGA can be used to implement functions, such as the functions used in digital signal processing, for extremely fast execution. The SoC Is surrounded by a variety of peripheral devices and 512 MB of RAM. There are two 14-bit A/D and D/A converters for measurement tasks, with a maximum sample rate of 125 MS/s. with those specs you can handle signal conversion for fairly demanding tasks, and the FPGA gives you the necessary processing speed. The board features an Ethernet port, a USB Host port and a USB Com port for easy ‘communication with a PC over a serial interface. There is also a slot for an SD card, which the Linux OS used for mass storage Instead of a hard disk drive. General purpose 1/0, IC, RS232 and ‘additional relatively slow A/D and D/A converters, ‘can be accessed through extension connectors. In short, the module has everything you need for demanding measurement tasks and high- performance signal processing with relatively high-frequency signals. That's exactly what Red Pitaya Is supposed to be: a general-purpose Instrumentation platform for signals up to 50 MHz. Getting started Getting started does not present any problems because a good user guide Is available on the website, Simply copy all the software and data that is supposed to be on the SD card to a MicroSD card and insert It in the card slot on the Red Pitaya board. Then connect the Red Pitaya to your LAN and switch on the power. The moduleboots up automatically, and you're ready to go. You might want to monitor the boot process on your PC via the USB Com port, since that way you can see which IP address Is assigned to the Red Pitaya. After the module has booted up, you can communicate with Linux directly over the USB Com port or you can use Ethernet for further communication. The latter is the preferred method. Instruments from the Bazaar One of the main objectives of the Red Pitaya project is to provide ready-made applications in the form of apps available online. You can find the apps in the Red Pitaya Bazaar. After you point your web browser to the Red Pitaya website, you can go to the Bazaar from the home page and select the application you want. The apps listed in Table 4 are currently available, Of course, many other interesting apps for various ‘measurement tasks are likely to show up In future. All of the projects are open source, so You can use them as starting points for developing your own applications. We chose the oscilloscope and spectrum analyzer apps to get familiar with the device. We used them to make measurements on the circuit shown In schematic form in Figure 3 and in physical form in Figure 4. ‘The Inputs of the Red Pitaya A/D converters are buffered by opamps with 1 M& input impedance, 's0 you can connect normal scope probes directly tothe inputs. You can select a measuring range of 2 Vpp OF 20 Vpp by setting the input gain jumper. If you run the oscilloscope app on the Red Pitaya and connect the probes to TP1 and TP2 of the cuit in Figure 3, you will see the screen shown in Figure 5. All of the usual oscilloscope settings are available. When you run the spectrum analyzer app, the screen shown in Figure 6 Is displayed. You can track and compare the two spectra in real time. In Figure 6 you can easily see the harmonics of the crystal oscillator, with stronger harmonics on ‘P1 at the gate output than on TP2. After getting an Initial impression of what can be done with the Red Pitaya platform, we looked at what you need to know in order to program applications for this device (Figure 7). Web-based instruments. The A/D and D/A converters usually interface directly with the FPGA. The FPGA Is also used © Two-channel signal generator, (0-50 MHz © Two-channe! oscilloscope, 2x 125 MS/s BER OP es eR a ee * Two-channel spectrum analyzer '* Frequency response monitor © PID controller Figure 3. A aystal oscillator circuit. Figure 5. ‘The Red Pitaya scope app in action. Figure 6. ‘The Red Pitaya spectrum ‘analyzer app in action.Figure 7. The components involved ina Red Pitaya web-based Instrumentation app. Figure 8. Block diagram of the signal ‘generator. Figure 9. Block diagram of an FM ‘generator consisting of two DDS stages. o NS for any required pre-processing, such as storing frames in Its fast internal memory. This means that you have to write suitable code to program the FPGA. The FPGA usually communicates with a program running on the ARMS processor. In most cases this program handles the control tasks, which do not require high processing speed but are ‘more complex. Naturally, this code also has to >be generated for any new application. In the case of a web-based app, this code is linked in on the Linux web server as a shared library. The web server handles data transfers between the web browser on the client system and the program ‘running on the ARM processor. The graphical user interface (GUI) on the browser Is implemented using a language such as JavaScript. Here again, this code must be developed for a new app. ‘As you can see, you need to know a good deal about programming if you want to develop ‘complex instrumentation apps. However, you can take advantage of the fact that all existing apps for the Red Pitaya are open source, so there Is plenty of starting material you can use to work your way toward a solution by making small changes. Now let's see how you right get started. Programing the ARM processor First let's look at what you can do with simple software for the ARM processor in combination with existing FPGA code. The standard FPGA code includes functions for a two-channel signal generator. The waveform for each channel Is ‘generated using a lookup table stored in memory with a total of 16,384 samples. Samples are taken from the lookup table incrementally in the same way as a direct digital synthesis (DDS) signal generator. These samples are fed to the 14-bit D/A converters. The sampling rate is 125 MHz. The block diagram for one channel is shown in Figure 8. The configuration parameters are stored in FPGA. registers, which can be read and written by the processor at specific addresses. This is done by aC program. FM RTTY signal generator ‘Suppose you want to shift the frequency of the signal generator back and forth under software control. To do this, you change the step size of the DDS signal generator. The result Is an FM RTTY signal generator. The wait time between ‘each change Is 20,000 1s, which corresponds to a data rate of 50 bits per second. The two transmit frequencies can be set to any desired value between 0 and 50 MHz. In the demo software the bits to be transmitted are first calculated ‘as Baudot code and then output in a loop. The program code for this loop is shown in Listing 1. You can directly modify the FPGA register by simply accessing the variable chb-count-step. This example shows that with a bit of simple of C programming, you can implement truly interesting applications with the Red Pitaya. Of course, you have to generate the C code using a complier that is compatible with the ARMS CPU. For this purpose the author used the ARM9 GCC compiler running under Lubuntu on a virtual ‘machine under VirtualBox. ‘Another option is to generate C code for the Red Pitaya under Windows, for which good descriptions are now available. In that case you can use SCP/WINSCP or another suitable utility to download the code from the PC to the Red Pitaya and then run the program,FPGA programming ‘The next step is programming the FPGA, for which ‘we also have demo software. The author used the \Vivado tool from Xilinx, in the form of the free |WebPACK edition, to generate the FPGA software. ‘The standard FPGA code for the Red Pitaya is also available as a Vivado project, so you can Use it as a starting point for your own designs by modifying the original code. In particular, that saves you the effort of reinventing the interface to the ARM processor and the other hardware. It Is also helpful that FPGA register contents and signal samples can be viewed as variables in ARM processor memory. This enormously simplifies ‘debugging because you can peek into the FPGA ‘while it is running. Now let’s look at an actual application, FM signal generator ‘We already have a two-channel signal generator in the FPGA, with each channel arranged as shown In Figure 8. Now we want to modify this so that ‘one channel modulates the frequency of the other channel. Since the waveform of the modulating signal Is also stored in memory, you can use the device as a sweep generator by making the modulation linear. Figure 9 shows the block diagram of our FM signal generator The upper portion Is the modulator DDS. The samples in the table have a ‘word size of 32 bits to enable smooth frequency modulation. The samples read from the table are sent to the second D/A converter for monitoring. Modulation is performed by sending the samples to the adder of the second DDS in the lower part of the figure, which forms a modulated DDS ‘oscillator. To allow an oscilloscope to be triggered ‘when the signal generator Is used in sweep mode, suitable trigger signal Is tapped off from the upper DDS. It is routed to GPIO pins and to a software-readable register. ‘The essential software for the modulator shown in Listing 2 consists of a few declarations and ‘a code block with six instructions. Figure 10 shows the spectrum of the output signal obtained by using this FM signal generator to produce a 10 MHz signal modulated by a 1 kHz sine wave with a specific modulation depth. The modulation depth is set to produce a frequency deviation of #2 kHz on the 10 MHz carrier frequency. The Verilog code of the frequency modulated oscillator is very similar to that of the modulator (see Listing 3). One of the main differences is {in the computation of the new sample position, since the modulation signal data fmInput—1 Is also added here. Listing 1. RTTY transmit loop. uint32_t delay1=20000 ; // 50 Bit/sec whiLe(1) for(int k=O ; kenBits 5 kr+){ ant theBit=biteuffer[k] ; ‘41f(theBIt){ g_awg_reg->chb_count_step else{ g_awg_reg->chb_count_step usleep(delay1) 5 ? ? step2 5 step ; 3 Listing 2. Verilog code of the DDS modulator. reg [ 32-1: 6] dac_buf [@:(1<<14)-1] ; // data buffer 16384x32 Bit reg [ 32-1: 8] dacrd — ; // DAC value reg [ 14-1: 0] dac_rp J] DAC read pointer reg [ 32-1: 0] dac_pnt ; // read pointer always @(posedge dac_clk_1) begin dac_rp <= dac_pnt{é-1+1 dac_rd <= dac_buf{dac_rp] ; // read data value signal_o <= dac_rd 5 dac_o Open Object File For Debugging. ‘The compiled files generated by Atmel Studio are located in the following folder under Windows 7: C:\Users\,000\AppData\Local\ \Witicro\Arduino\Butlds\SpeedTest\ ardutno_due_x_dbg\ There you will find the file speedrest.e1£. Now let's open it for debugging. If the sketch was compiled and saved with the‘Arduino IDE, a copy of this file Is also located in the debug folder for the sketch. It can be seen In the Solution Explorer window of the code gen- eration instance of Atmel Studio under “Output Files’ (Figure 8). Using the browse function, find and select the output file Speedrest..elf. You should see a window like the one shown in Figure 9. You can specify your own project name and storage location for the debug project folder. Your settings will be saved and used automatically the next time you open the debug project. Next you see a window for selecting the device family and target microcontroller (Figure 10). For the Arduino Due you should select “SAM3, 32 bit” for the device family and “ATSAM3X8E” for the microcontroller type. ‘Now the file Speedrest..cep appears in the Solu- ton Explorer window. This is simply the program file of the sketch, which originally had the extension . ino, Right-click the project name “Speed- ‘Test_Debug” to open the Properties window. When the Atmel ICE is connected to the PC over USB, the debugger can be selected under Too! -> ‘Selected debugger/programmer. Here you must specify “SWD" as the interface instead of “JTAG". Now the Atmel ICE Is ready. You can also use it as a programmer. To open the programming window, press Ctri-Shift-P or cick the chip symbol with the lightning flash. The desired functions are available after you click Apply. You can set the maximum SWD clock rate (2 MHz) on the Interface Settings tab. On the Memory tab, you can download the speedtext.e1f file directly to the flash memory of the microcontroller. The program starts running automatically after the programming process completes. However, it Is more efficient to use the Debug ~ Start Debugging and Break or the Start With- ‘out Debugging command (or the corresponding icon: the blue arrow with the vertical blue bar or the solid green arrow) to download a program for debugging or running, instead of the device programming function. Downloading the complete program to the microcontroller takes roughly 1 second with the Atmel ICE, which is quite fast. ‘The actual structure of the downloaded Arduino rogram can be seen in the Solution Explorer window, where the main. cop file is selected and the ‘cursor is positioned on the first line of this routine Figure 9, Dialog for selecting the ‘generated output fil. Figure 10. Dialog for selecting the device family and device type. Figure 11. ‘The true complexity of an [Arduino project can be seen {inthe Solution Explorer window. Figure 12. When ths error message appears ater you modiy 2 fle, simply cick “Reload” to dismiss the message.Figure 13. Oscilloscope screenshot of the 150-kH pulse waveform, Figure 14. Microcontroller PIO register. Figure 15. Oscilloscope screenshot of the 12-MHtz pulse waveform generated by version 2.0 of the program. 3 re Deron corte GF) (Bicrenraps race ape =) (Cina ered eat ot) ‘Great roataens cote Os) 31 rea rostoeps cern 2) = Qiiswntnnten coe BU * Ohecimeass on Saree mera eres Sree teers Saxe = Brose * Broren Cima. + nos preertimetr eda eat: ameas Gaal od nae one sees nee Sees Sem (Figure 11). As you can see, the speedrest..cop file is only a small part of the overall program, which is why a full compilation takes so tong. The debugger gives you access to all the debug functions of Atmel Studlo, such as single-stepping with F11, setting breakpoints with F9, defining breakpoint conditions in the Breakpoint window, and s0 on, If you edit and recompile the speedtest..1no file in the first instance of Atmel Studio, the message shown in Figure 12 appears in the second Instance (the debug environment). Press “Reload” to close the debug message and load the modified program into the microcontroller. Now the cursor Is again positioned on the first line of the program code and the program can be run again. If you have previously worked with 8-bit AVR microcontrollers in the Atmel Studio environment, you will not have any problems working with 32-bit ARM microcontrollers because the user Interface is nearly the same. Incidentally, clearing and rewriting the flash memory with the Atmel ICE does not affect the boot loader because It Is located in the ROM of the microcontroller and therefore cannot be deleted or overwritten. Con- sequently simply downloading a program over the USB interface still works the same way. SpeedTest Only the following two instructions are executed in the main loop of the example sketch (see the SpeedTest listing): digitalwrite(TP1,1);//0n digitalwrite(TP1,0); //ort With an oscilloscope, you can see a pulse waveform on this pin while the program is running, with a pulsewidth of 2.35 ps and a period of 6.6 Hs (Figure 13), which corresponds to a frequency of about 150 kHz. That is not very high, and you could hardly be blamed for thinking that there's no reason to use an ARM controller clocked at 84 MHz for this when a simple 8-bit AVR microcontroller can do it better. The reason for this poor performance is that the Arduino instruction digi talwes te is @ cover name for a bunch of C routines, as you can see by exe- cating the program in single-step mode with the debugger (F11 key). To actually achieve high processing speed, you have to avoid this over-head. Fortunately, the Arduino programming language Is not a separate language, but instead a form of C with a GNU compller in a special Arduino configuration. This means that a pin configured as an output can be addressed directly by setting or clearing a bit in the PIO register. The Instruction const int TPl = //Test pin selects Arduino pin 7 as the test pin. According to the Arduino pin mapping, this corresponds to pin PC23 of the microcontroller or bit 23 in the parallel input/output (PIO) register. With the program stopped you can go to the 1/0 View window of the debugger (Figure 14) and click bit 23 of the PIO_ODSR (set) register to set the pin high. To reset the pin, click the same bit in the PIO_CODR (clear) register. In the main loop you can simply use the instruction PIOC->PIO_SODR = 1<<23; to set the pin without any fuss or bother. To test this, and at the same time check whether conditional compilation works properly, the author generated a new version of the sketch (see the SpeedTest 2.0 listing). If the expression Direct Is defined in the code, the code segment after #1fdef Direct Is compiled, and otherwise the previous version with the slow digi talwrite instruction is compiled. Naturally, this change would also work in the normal Arduino IDE. Listing 1. Speed-Test. [RRR RRR EERE REE //Speed-Test with Arduino Due [oR aae aR Ra REA Ea const int TPL //Testpin void setup() { [+ add setup code here +/ // set the digital pin as output: PinMode(TP1, OUTPUT) ; } void Loop() { // put your main code here, to run repeatedly: digitalwrite(TP1,1); //on digitalrite(TP1,0); //Aus ?Listing 2. Speed-Test 2.0. [[eeeeteeeesneanannneanannanannnees //Speed-Test 2.8 with Ardutno Due /[[eeeetenenenennannnnaennnaneennees include yarduino.n« “ “ const int LED1 = 13; ‘Aint LED2 = 12; ‘ant LED3 = 11; ant TPL = 7; //Testpin define Direct void setup() t /* add setup code here +/ J/ set the digital pin as output: inode (LED1, OUTPUT) ; inode (LED2, OUTPUT) ; inode (LED3, OUTPUT) ; pintode(TP1, OUTPUT) ; // set output tow digitalwrite(LED1,0); digitalwrite(LeD2,0) Gigitalwrite(LED3,0) Gigitalwrite(TPL,0)5 void Lo0p() // put your main code here, to run repeatedly: #fder Direct //x= state of PIO_SODR with bit 23 ‘Ant x = PIOC->PIO_SODR | 1<<235 wntle (1)¢ //Omly for testpurposes, don‘t exit thts Loop PIOC->PIO_SODR = x; PIOC->PIO_CODR = x; ? else digitalwrtte(TP1,1)5 digitalwrite(TP2,0) 5 fenatt ? ‘The result of this code optimization, as shown in Figure 15, is stunning: the pulse waveform now has a pulse width of 23.2 ns and a period of 83.3 ns. This corresponds to 12 MHz ~ about 80 times faster than before. From that we can conclude that using an ARM microcontroller isn’t such a bad idea. Summary Working with Atmel Studio accelerates Arduino sketch development to a completely new level thanks to shorter turn-around times, since only modified code Is recompiled. Atmel Studio and the tools that can be used with it, such as the ‘Atmel ICE as a debugger and programmer, are worthwhile when you work with Atmel microcontrollers because they offer professional debugging features and because downloading programs to the microcontroller memory is significantly faster, Incidentally, you can also use Atmel's SAM ICE instead of the Atmel ICE for ARM projects. With the SAM ICE hardware debugger and program- ‘mer, microcontroller programming speed can be boosted by a factor of 4 because it can operate at 8 MHz Instead of 2 MHz. For this you need the adapter described in the inset. Ifyou are a professional software developer, the expenditure Quickly pays for itself even ifthe device can only be used with the AT91xx microcontroller family. (1303924) Web Links [1] Atmel Studio 6.2: wwew.atmel.com/microsite/atmel_studio6 [2] Arduino IDE for Visual Studio: wwew.visualmicro.com [3] Atmel ICE: www.atmel.com/tools/atatmel-ice.aspx Review: wwnw.elektor-magazine.com/ 140275 [4] SAM ICE: www.atmel.com/tools/atmelsam-ice.aspxby Thijs Beckers (Glektor Labs) Bulging Caps The other day I went to the local Used Chemicals Drop-off to dispose of some sad old inkjet printers when a fairly intact looking PC case caught ‘my eye. Without needing to dumpster dive I was able to spot an “Intel Core™ 2 Duo” sticker and another with the Windows Vista™ Home Premium ‘OEM product key, these gave me a clue about the ‘age of this PC. It wasn’t beyond hope considering that my current laptop was “Built for Windows Vista"—I'm reluctant to throw away stuff that still works, even though it might not be shiny and brand new anymore. Anyway, the PC case didn’t look battered or kicked around much, so I decided to take it home and to find out about the specs—CPU, RAM, HDD, that sort of thing Back home I opened it up. All components were present except for the hard drive. And tt looked like the previous owner had removed it decently ‘and hadn't just ripped It out. So I guessed it was time to try to fire it up, I connected a monitor and hooked up a key- board and mouse. I connected the power cord ‘and pushed the Power switch on the front and. It came alive! I entered the BIOS to see more detailed specs and was happily surprised. Here's ‘a breakdown of my find © CPU: Intel Core 2 Duo £4500 @ 2.2 GHz © RAM: 2 GB DDR2 @ 667 MHz © Mainboard: Fujitsu D2151-A21 © Graphics: NVidia Geforce 8400GS 256 MB ‘© Optics: DVD-ROM and DVD-R/RW DL drives © 350-W PSU «Internal Card Reader © Cooler Master case In 22 months we will have a good laugh about this configuration, but these days It’s good ‘enough for casual surfing, mailing and even some music making. Some more dumpster-diving in certain places Is sure to yield a hard drive and I can get this baby running. But first there was an issue I didn’t like very much. I noticed it when I inspected the mainboard. See those capacitors next to the ATX connector? Though the mainboard functioned perfectly, my electrical engineering hunch says them ‘caps look like they're about to explode. Which Is of course a bad thing. But if replacing a couple of bulging capacitors is the only thing needed to give this PC a second lease of life, I'm still a happy camper. I ordered a couple of decent low- ESR type replacement caps (Panasonic FM-series), removed the time bombs and soldered the new caps in. After Installing a fresh operating system, the PC ran fine. And after a month or so it still does. Did I mention I made a Hackintosh out of it? Maybe I shouldn’t have said that out loud ©.Elektor.Labs goes wearable by Clemens Valens (Glektor.Labs) Wearable is the new hype ‘The previous months have seen a lot of hobby- {st-related activity from Intel, the world’s largest manufacturer of PC microprocessors. Not only did Big Blue release their new Galileo (V2 is now available) and Edison boards with Arduino support, the company was also the main sponsor of the above mentioned Rome Maker Faire. At this show Intel made another announcement, besides the Edison board, that didn't really get the atten- ton that it deserves, in my humble opinion: Intel firmly intends to become a major player in the still pretty small wearable electronics market. To prove this they showed of their MICA (My Intelligent Communication Accessory) bracelet. 4 Intel has invested in start-ups and teamed up with fashion designers to come up with reference designs (and products, why not?) to open the market and to show what's possible besides displaying the user’s heart rate and body temperature. Have a look at the Mimo Baby Monitor, built on Edison. Elektor cannot stay behind, of course, and so here's a call for wearable electronics projects to be published on Elektor.Labs. Have you designed something wearable? Do you have an idea for such a device? Post your wearable ideas and projects on Elektor.Labs and we will see that the best ones will be transformed into real products. .elektor-labs.com/wearable 3D printing—the story continues ‘Some months ago In this column I ranted about the useless things produced by so many people with 30 printers. I ended the article by suggesting that those people should print their own 3D printer filament or “ink” to save the planet from drowning in polymer 30 printed gar- bage. At the Maker Faire that was held at the beginning of October 2014 in Rome and that I visited I encountered once more many useless 3D printed objects. To be honest there were some very useful 3D printer applications on display too, sadly they were a small minority. During the keynote presentation the winners of a product design contest orga- nized by the Italian magazine Focus got ‘announced. To my big surprise the win- ning project was a... 3D Printer Filament Extruder! A kind of 30 printer that prints 3D printer ink. Being an Italian project, great attention had been given to Its looks, and the result is an appliance that sure looks great in any kitchen (unlike the ugly 30 printer). At the top you feed in polymer ingredients and out comes a filament that gets collected on a detachable spool. Very likely it’s sultable for making spaghetti too. Walking the show grounds I came across a few other similar machines, but not so nicely shaped ‘as the EWE Filament Extruder. Photos are available on this webpage: www.elektor-labs.com/3¢-printing (140365)by Neil Gruending (Conade) DesignSpark Tips & Tricks Day #16: Working with Components We've talked a bit about components in the previous installments so today let’s look at them in more detail. Tw oe oe tee De eo 46 ABE oe 6 a a ia ° a 4) A BBX Today I realized that we've talked about components in DesignSpark before without dis- cussing them in detail or how to use them efficiently, so that will be our topic for today. Back in DesignSpark Tips & Tricks Day #3 (Elek- tor September 2013) I talked about libraries and how components in DesignSpark linked all of the design information for an electrical component in one place. That’s how the schematic and PCB tools can find the component information they need without interfering with each other. Creating and editing components ‘Creating and editing components can be a lot of ‘work if you draw them from a blank page every time. If you can find a similar component that’s, already in the library then it’s easiest to copy It using the “Copy To” button in the Library Man- ‘ager window and then modify it. But when you ‘can't, then the DesignSpark component wizards ‘are a great way to make new components quickly. Let's bulld an audio playback IC component to see how the wizards work. always like to start with the schematic symbol first so we will open the DesignSpark schematic wizard by clicking on the Wizard button in the Library manager. Answer a few questions and It will automatically create simple schematic symbol that we can modify. After alittle editing I came Up with the schematic symbol in Figure 1. The pins are where I want them but if you look closely ‘you can see that they have the default names N1 to N16 which Is okay for now because they will be changed later at the component level. One Useful trick that I used is to place temporary text strings over the pin names to make sure that the proper pin names will fit inside the symbol. DesignSpark also has a PCB footprint wizard that can create a wide variety of PCB footprints for you. ‘The best part about itis that it can create all of the pads for you along with the correct spacings. The last step Is to use the component wizard to asso- diate the schematic symbol to the PCB footprint. ‘Once you choose the schematic symbol and PCB footprint you've just created you will get to the ‘Assign Pins screen lke in Figure 2 which is where you can label each pin name and associate the ‘schematic symbol pins to the PCB footprint. In my case I designed the schematic symbol so that its pin numbers match the PCB footprint pin numbers so I would click the “Assign 1-to-1" button, But DesignSpark would let you have any pin mapping you want. This Is also where you ‘can enter all of the pin names in the Terminal Name column. Clicking on the Next button will create the component for you and let you save It to your library. Figure 3 shows what our final ‘component would look like in a schematic. But what if we needed to make changes and update ‘our design? We will cover that next. Updating components When you make changes to a component In a library DesignSpark will only update the library files and not the schematic and PCB files that use ft. ‘That's because every design file keeps its own copy ‘of the component so that changes in a component. library can’t accidentally affect an existing design. ‘So how do we update those local component cop- les with the new component library information? ‘One way Is to delete a component and then add it back into the design but this is a dangerous‘operation that could have unintended side effects like breaking net connectivity in your design. This would be especially risky In a PCB file because It could break the link to the schematic for that file Instead you always want to go into the Tools menu and select “Update Components” and then ‘choose which components you want to update. This works in schematic and PCB files. ‘You will almost always want to either browse for the components or choose to update the selected ‘components. You don’t usually want to update all components automatically unless you review all of the pending changes to make sure that there won’t be any unexpected updates. Either way you will get the Update Components window in Figure 4 if you update components in a ‘schematic. The PCB version of the Update Com- ponents windows is very similar. ‘The “Only update item if version Is different in library” option tells DesignSpark to only update ‘a component if It’s different from what’s in the brary. The “Keep value positions” option tells DesignSpark to try and leave all of visible component fields in thelr current locations. This usually works but you will still want to double check that nothing has moved on you after the update because even though DesignSpark will keep the ‘same anchor point sometimes the text justifica- tion will revert to what was used in the library. Enabling the “Keep existing component values” ‘option will keep all of the component value fields In your design, otherwise the values will be overwritten with the values in the library. This option Is important when you modify a component's value fields in the library and you want to trans- fer those changes to your design. The Update ‘Components window adds another option called “Remove pad style exceptions” when you update ‘@ PCB component. You would normally leave it selected so that DesignSpark will clear all of the errors it detected with the previous footprint. Replacing and changing components ‘Sometimes you need to replace or change a component as you work on a design. For example, 1 like to have a different resistor component for each value In a design so that I can put all of the manufacturer’s part numbers in the component value fields and make a proper BOM later. That means that I just change the component when, I want to change a resistor’s value. If you haven't already placed the component on the PCB then it doesn’t matter if you Just delete ‘a component and then add a new one to replace It. But if you do this and the part Is on the PCB, DesignSpark will delete the component and then add the new one to the component bin. A better solution Is to replace a component without deleting It first. You do this by selecting the component and open its properties. Then you click ‘on the Change button which opens the Change ‘Component window where you can select the new ‘component. With this method DesignSpark will leave the component on the PCB and you also don’t have to worry about matching the reference designator. Conclusion Today we've taken a pretty close look at how to create components with the DesignSpark wizards ‘and how to update designs with the library changes. Next time we'll learn how to use multiple component gates and hidden power pin connections. ao3e7) Figure 4. Update Components window.8y Neil Gruending Web Linl Tantalum Bead Capacitors Weird Component # 10 What's the first thing you check when a plece of electronics stops working? I bet it’s the power supply voltages and if you are repalring a board that over 20 years old then you've probably your share of tantalum bead capacitor fallures. If you're lucky they will ust short out and current limit the power supply but I've also seen them explode Into a small ball of flame! But whence the fall. the noise. the vile smell? Tantalum capacitors like the dipped ones in Fig- ture 1 are a form of electrolytic capacitor that uses oxidized tantalum powder pressed into a pellet as its dielectric. The tantalum powder allows capacitors to be bullt with a very high capacitance for their size and since the powder Is solid they also have a very wide temperature range. But perhaps one of their best properties Is that they have a very low equivalent series resistance (ESR) compared to other electrolytic capacitors, types like aluminum electrolytic. The tantalum powder Is a dielectric because It easily forms an insulating oxide layer as part of the manufacturing process, The oxide layer thick- ness Is controlled by applying a polarized voltage to the capacitor called the formation voltage which {s usually a multiple of the working voltage. The capacitor will also get the same polarity as the formation voltage. It's this oxide layer that is crit- {cal to the tantalum capacttor's safe operation. It's thelr high capacitance and low ESR that ‘makes tantalum capacitors a good choice in power supplies and for a long time they have been used as a drop In replacement for small aluminum capacitors. But if you do that in a power supply rcult you could have a pretty high failure rate because it tums out that tantalum capacitors are sensitive to voltage transients and ripple current. ‘Tantalum capacitors can absorb small voltage transients within thelr operating range but if the transients are too large then they can punch through the thin oxide layer on the tantalum powder and damage the capacitor. The capacitor’s low ESR can unintentionally amplify the voltage spikes (ringing) and can also cause damaging current spikes. Once a capacitor is damaged it will degrade until it eventually shorts out. If the ower source Is large enough then the capacitor ‘can actually explode then burn the circuit board. But tantalum capacitors can be used safely as long {as you watch their voltage and current ratings. Normally you can get away with using an alumi- ‘num capacitor near its maximum rating but for a tantalum capacitor you need to derate its working voltage rating by at least 50%. You also have to ‘make sure that the ripple current won't exceed the capacitor ratings by adding enough capacitors in parallel. The downside to paralleling capacitors, Is that it also reduces the capacitor ESR, can create a small tank circult with the inductance in the PCB traces and create unexpected voltage spikes in power supply applications. Electronics roughly from the late 1980s through ‘mid-1990s used epoxy dipped and surface mount tantalum capacitors quite a bit in their power ‘supplies and for decoupling. Unfortunately these capacitors are now known to have a pretty high fellure rate (Figure 2). In some cases It’s now ‘common advice to test or replace all the tantalum. ‘capacttors when repairing 1980/90’s electronics. It’s hard to know exactly why the failures hap- pened but according to Kemet [1] early tantalum. capacitors weren't actually designed for power supply applications and they required a signif- cant amount of external resistance to operate reliably (usually several ohms per volt applies to the capacitor). It's also possible that many designs from that era were using the capacttors near their voltage ratings and that they simply just falled because of many small transient events. For example, it was common to see 6.3 V capacitors used for 5 V power rails. ‘Tantalum capacitors are being used less often in znew designs thanks to MLCC ceramic and better ‘aluminum electrolytic capacitors but if you need a lot of capacttance in a small package tantalums are hard to beat. Modern tantalums have dealt with the problems mentioned here with surge- tested and fused parts. You just have to make sure that that they are used within their limits. (140366) [1] http://canada.newark.com/pdfs/techarticles/kemet/Tantalum-in-Power-Supply-Applications.pdftas gt 8x8x8 cube * Every LED individually addressable '* Optional use of infinity mirror ‘= Animations can be simulated + Developed and programmed using Flowcode 6 and ECIO40P16 © Arduino patchable Infinite RGB LED Cube with Flowcode by Ben Rowland, ‘The 8x8x8 LED cube project published on the Instructables website [1] was very detalled and yet looked simple enough for me not just to complete but add substantial enhancements as well. After planning on creating a larger single color ver~ sion of the cube I quickly realized that to double the display dimensions to 16x16x16 meant octu- pling the number of LEDs and interconnections. The thought of processing 2,048 LEDs was too much for me so I scaled down my thinking and instead decided to repeat the &X8x8 (512-LED) design but Using full color RGB LEDs instead of single color LEDs. Even at 8x8x8 RGB LEDs it stil tallles up to ‘around 2,200 solder joints so be warned! From my research I also came across some great infinity mirror projects which are basically LEDs sandwiched between a mirror and a pane of glass with a one-way mirrored surface to create an effect that the LEDs go on forever (Figure 1). 1 decided to apply some of the simple techniques to my cube to try and extend the display beyond the bounds of the actual LEDs. With 512 RGB LEDS in the design each with their ‘own three internal LEDs a lot of control signals are required to switch the LEDs on and off. There- fore a method is needed to share control signals An Instructables project colorized & enhanced Matrix Technology Solutions Ltd. (UK) Here's how to take a plain single-color 8x8x8 LED cube project to the Next Level. The newly released Flowcode 6 software development suite was instrumental in a lively journey into understanding and achieving the simulation of animations on ‘full color’ LED cube that’s capable of mesmerizing visualizations. between LEDs (multiplexing). In the design the LEDs are cathode-grounded with a current lim- Iter resistor on the anode. Multiplexing tool 1: buffer ICs ‘The LED cube employs two types of multiplexing to allow all 512 RGB LEDs to be driven inde- pendently. The first method uses octal flip flop buffer ICs to allow us to create enough control signals to drive a full 8x8 segment, I.e. 1/8" of the cube. In an 8x8 section there are 64 LEDs which means we need 192 control signals to allow us to drive all the individual color channels. To drive all the signals we need to chain 24 of the buffer ICs together so that the output of the first buffer drives the input of the second buffer, the second drives the third and so on. igure 2 shows a simple schematic with two buffers chained together and connected to drive LEDs. The dock signal Is driven by the controller ‘and used to shift the output signals from one buffer IC to be passed to the next buffer IC. I made the buffer driver fairly modular to allow ‘me to scale up or down the size of the cube depending on how many LEDs I wanted. To do this I created a PCB with three buffer ICs on allowing me to control a single row of eight RGB‘One method of achieving LED mattiplecing isto “insert” the devices between ‘two buffer ICs to which intelligent control is applied Figure 2. implementation ofthe 8x8x8 RGB Color LED Cube. ish) buffer board. Eight of these are required for a full Figure 3. ‘Schematic of the modular(- 3993943 Sseossae 18S oa PEEEEEEe Ssasssss 2@28 5 a a) a}3} 3) 3) 3) 3)3) roe BevesaeeFigure 4. ‘One of eight NPN transistors used to switch 8x8 sections ‘on and off under software ‘control, completing the ‘overall multiplexing scheme for 512 RGB LEDs. Figure 5. ‘The stack of driver boards. LEDs. With eight of these driver PCBs chained together it allows me to control the entire 8x8 ‘segment of the display. The smaller 16-way ribbon cable is used to pass the 10 control signals plus power and ground between the PCBs. The wider ribbon cable is used to pass the signals ‘onto the LEDs via the on-board resistors. Each board Is fitted with Its own capacitors to allow a bit of current to be sourced locally as and when needed by the LEDs. Multiplexing tool 2: NPN transistors In Figure 3 each of the 8x8 LEDs in a single segment is driven via our chain of buffer ICs. To allow us to control the other seven 8x8 seg- ments of the display we need a second method ‘of multiplexing. To do this we take the cathode pins of all the LEDs within the 8x8 segment and ‘connect them all together. We then connect these to ground via an NPN transistor which allows us to enable and disable all the LEDs in the 8x8 ‘segment at once using a single control signal to the transistor’s base pin. Figure 4 shows the approach schematically; Figure § the resulting stack of eight driver boards. By connecting all the LED cathodes together for each horizontal 8x8 section we can then use ‘eight NPN transistors to enable or disable each of the &xs sections and therefore control each of the LEDs in each section individually. This is, great but obviously this means that we can only light up 1/8 of the display at any one time. By ‘constantly switching which 8x section is active at fairly high speed we can build up the entire display so it appears to the human eye that the entire cube Is lit and constant. The Build ‘So that’s pretty much it for the theory, It’s time to pull up our socks, make a brew and dive into the world of bending wire and soldering. There are two fairly laborious tasks that are required when building a LED cube, the first being bending LEDs and the second being soldering the bent LEDs together. Each step required a Jig to make the process more manageable and to help maintain quality while I slowly lost the will to live. Luckily I managed to do most of this stage during the Christmas vacation so there was enough going on to keep my mind interested while my hands. were occupied on the task. Bending LEDs ‘The first step for assembly was to take all 512 LEDs and bend them correctly. To do this 1 drilled ‘a Smm hole in a spare piece of wood and checked that a LED fit snugly and flush to the wood but could also be inserted and removed easily. Once Thad my hole correct I drew some guide lines to help with the bending. As pictured in Figure 6, the lines I drew were one vertical running through the center of the hole (A), one horizontal line run- ring through the center of the hole (B) and one horizontal line running through the very edge of the hole (C). A metal ruler came in very handy during the bend stage. ‘The photos in Figure 7 show the steps to follow when bending the legs. 1. Insert the LED into the jig ensuring you have ‘got the LED in the correct orlentation. The pins should all be sat in line with the horizontal line (B) and the cathode pin should be the second pin from the right. 2. First Bend: bend the cathode pin downwards ‘and the anode pins upwards parallel with the line (A). Ensure they are fairly flat to the board ‘and the bottom of the LED. 3. Second Bend: using the metal ruler bend the cathode pin to the left so it Is parallel with line (8). 4. Third Bend: using the metal ruler on the sec- ‘ond horizontal tine (C), bend the anode pins ‘upwards again. One LED down, 511 left to bend! Assembling 8x8 sections ‘Once you have at least 64 bent LEDs you are ready for the next stage which is to assemble one 8x8 section of LEDs. To hold the LEDs in place I created a jig using my trusty 3D printer and a design T uploaded to Thingiverse (2). If you don’t have ‘access to a printer then the Instructables guide [1] describes a different method to hold the LEDs inplace. The LEDs were soldered together by inserting ‘each LED one at a time into the jig before soldering the anodes together on the current column and the ‘cathodes together on the current row as shown in Figure 8. Take your time and ensure that every solder joint you make Is nice and strong, spending ‘a bit more time here will potentially save a huge ‘amount of time later on when it comes to getting the thing up and running. ‘As each complete panel was finished I made sure to give it a full test while it was stil in the Jig to ensure that every LED lit correctly and that there were no breaks or shorts in the panel. To do this I connected long lengths of wire to my 5-V power supply. At the end of one of the pieces of wire I soldered the larger of my two resistor values used on the driver board from Port 1 and this allowed me to drive all of the LEDs without ‘any risk of damage. Using my trusty metal ruler pressed against the unconnected common leads Twas able to temporarily short all the common pins together to ground meaning I only have to test each anode column once rather than having to test each and every LED individually. To get the panel of LEDs out of the jig I just gently pushed ‘on the top of each top LED in turn and repeated Until the entire panel popped loose. Setting the completed panel aside I went back to bending LEDS for the next &x8 panel. Combining the 8x8 sections Afterall the LED legs were bent and all the LEDs ‘were assembled into panels it was time to start Joining the panels together. I laid the first panel flat on a desk and soldered two lengths of straight ‘single core wire to each the cathode wires one at elther side of the panel to provide a good level of support and a bit of redundancy Just in case ‘one of the solder connections on one of the wires were to fall. To obtain the straight wire I used the technique of holding both ends with a pair of pliers and pulling hard. I had to experiment with a few different types of wire before I found one that would be happy with the current from 64 x. 3 LEDs and would also straighten nicely without too much effort on my part. I used a few small match boxes to space the next panel away from the first. Taking my time to ensure everything was as lined up as possible I repeated this process until all of the panels were together and then gave the LEDs another full test to ensure I hadn't missed anything. A virtual 3D view of the ‘cube in the making like in Figure 9 shows the Position of the 8x8 section. Preparing the base The cube once fully assembled Is fairly strong but all it would take would be a bit of weight applied in a wrong place and all that work would be ruined. I used the same piece of MDF board that I used to create my jigs and drilled 5-mm_ holes into this at the correct pitch for the legs of Figure 6. Guide ines (iter) to help with the LED pin bending. Vas Figure 7. Method of bending the LED pins for easy assembly into a cube shape. Figure 8. ‘jig is used to hald the 88 LED arrays together for the purpose of soldering the ‘anodes and cathodes thus creating rows and columns.Figure 9, AD rendering showing the whereabouts of an 8:8 section within the cube. Figure 10. ‘The program for the RGB LED cube was developed using Flowcade 6 shown In action here. Note the ‘simulation in the right hand ppane. The full program ‘and more is contained in archive fle 140149-W.zip downloadable at [3]. the bottom row of LEDs to go through the MDF. ‘To work out the pitch I took measurements from ‘outer left LED to outer right LED on both sides of the cube and then averaged the spacing, then repeated for front to back. I then drilled a further ‘eight 3mm holes along one edge for the cath- ‘ode wires to come through to the cube. Finally 1 routed 2 6-mm groove for my glass cover to fit Into and spray painted the MDF. Details on inserting the cube onto the base, adding the common-cathode rails and the project. wiring may be found in a supplementary document available for free downloading from the Elektor magazine website [3]. The Brain T used some pieces of veroboard to allow me to connect an ECIO40P16 ‘brick’ directly to my 5-V power supply and to create the NPN com ‘mon driver circutry. A S-V power supply must be used to allow a fairly large current to flow without ‘generating any additional heat. The Flowcode LED Cube component is flexible so will work with pretty ‘much any microcontroller family apart from 8-bit PICs as these cannot have arrays greater than 256 bytes in size. The ECIO40P16 packs a lot of, punch for its size but with sultable code patching a standard AVR based Arduino should also be able to drive the cube if that’s what you have available. Whatever method you choose of packaging or securing the control electronics to the cube assembly, remember to add a hole to allow your power supply to be connected. Some of the pop- ular SMPSUs come in small metal cases so this could also be Included inside the box to neaten things up a bit. You may also want to leave the USB cable connected to the ECIO to allow you to a) reprogram without having to re-open the casing and b) for communications via USB should you want to go down this route using the Flow- code USB components. Enter Flowcode 6 ‘To create the simulation and the code to drive the display I used the Flowcode v6 software and this had everything I needed to allow me to create ‘a masterpiece. I began by creating a sphere on the panel, making it invisible and then cloning the sphere enough times to match the x, Y and Z dimension properties. I then went about creating routines allowing me to get and set the color of, each individual sphere. This then allowed me to generate the line and cuboid drawing routines as well as the slightly more complex shift and rota~ tion macros. By using a previous graphical LCD component I was able to rip out the text drawing ‘macro and by finally adding a double buffering system the simulation side of the component was about feature complete. To add the embedded functionality I created an array of 16-bit variables to store each of the LED color values in the cube. I then created a routine which could be called repeatedly to handle things like clocking out the data to the buffers ‘and switching between the eight common channels. Finally all that was left to do was to add code into the get and set color macros to allow the array to be read and written when not running via the simulation. The final component is ‘now available [3] which can simply be draggedonto the Flowcode panel. The Flowcode sketch Ing ‘sheet’ Is shown in Figure 10. Visualizations The Flowcode tool has been used to develop a set of test programs to generate a rain effect with lightning, a fireball, two interacting plasma balls, a textual display and a vector based animation. A video Is available on YouTube [4] showing the visualizations running on the hardware. The next Item Description RGB LED, Smm, diffused lens, common cathode 5V SA switch mode DC power supply Microcontroller Brain (ECIO40P16 or Arduino) SN74ACS74N octal flip flop 16-Way IDC ribbon Cable 24-Way IDC ribbon cable 1L0uF aluminum electrolytic capacitor 100nF ceramic disk capacitor 1009 0.25W resistor (green/biue LEDs) 1800 0.25W Resistor (red LEDs) 16-way IDC socket 16-way DIL header 24-way IDC socket 24-way DIL header T1P122 NPN SA transistor ‘Small piece of Veroboard Tempered glass shelving émm / 0.25" thick Sheet of fiberboard / MDF 20mm/ 0.8" thick Spray paint to color MDF Reflective car window tint step Is to add a microphone to the internal cir- cultry and use this to control the movement of the visualizations using techniques such as FFT to chop up the audio signal into frequency bins. Plenty of challenges there—let us know how you get on. (140149) Quantity _| Source Part Number 512 | eBay 1 eBay 1 Elektor Store 24 Farnell 165-2006, 2m/7t_| Farell 120-7442 ‘4m/13ft_| Farnell 120-7439 8 Farnell 211-2751 8 Famell 187-1077 128 | Farell 933-9043 64 Farnell 933-9230 16 Farnell 221-5235, 16 Farnell 102-2240 8 Farnell 221-5253, 8 Farnell 159-3448 8 Farnell 980-4021 1 Farnell 120-1473, si B&Q/ Lowe / Home Depot / DIY Store A B8Q/ Lowe / Home Depot / DIY Store a B8&Q/ Lowe / Home Depot / DIY Store 1 Halfords / Auto Partsfaeries Distance and Level Gauge With Alarm Function Do you need to know how full your oll tank or water tank is, without requiring vast compu- tational effort or expert knowledge? You fill a large barrel with water from the garden hose ‘+ Measures the level status of liquids '* Watchdog function monitors fill level, with relay output and alarm LED ‘+ Programming of min./max. alarm levels infinitely variable ‘+ Stores min./max. calibration values for up to ten tanks or containers in memory ‘= Also measures distance + Intuitive menu guidance using LCD display By Jérg Trautmann (Germany) ‘Small, widely available and dirt-cheap ultrasonic sensors enable us to design both simple and lavish measuring devices for all kinds of ranging. Take one of these modules, add a display, a couple of pressbuttons and a microcontroller loaded with software, and you now have all the ingredients for a circuit of this kind. and would like to cut off the water flow using a ‘magnetic valve when it has reached the desired level? Or maybe you would just like to measure the distance between two objects? The possibil- ities using the level and range meter described here are virtually limitless! The extent to which cylindrical or cubold contain ers are full can be determined either mechanically with a float and potentiometer, or else capaci- tively with ultrasonics or a laser device. Both approaches, mechanical and capacitive, require ‘auxiliary components such as floats and sensors. If accuracy of measurement Is your prime con- sideration a laser Is definitely the best solution, yet even this method has its pitfalls. For exam-ple, fog and steam can cause reflections for the laser, falsifying the result. Since influences of this kind affect ultrasonics hardly at all, I settled on ultrasonic sensors for this cost-effective device. A further advantage of this solution is that it requires no physical contact to be made with the liquid medium. Into the bar- gain we also obtain a tool for measuring distance. ‘The frequency of 40 kHz normally used by ultrasonic transducers corresponds, at an ambient temperature of 20 °C, to a wavelength of 8.5 mm. ‘Sound waves in this frequency spectrum do not disperse to any great extent and diverge (in a dlub- shaped form) with an aperture angle of around 15°, Perfect qualifications for our purpose then, since the side walls of the container will have effectively no influence on the reflected signal. The ultrasonic sensor used here was already described exhaustively in Elektor a few months ago in the 2014 Project Generator issue [1]. ‘Accordingly we will Keep the functional description short and to the point. A trigger pulse causes the ultrasonic module to transmit a burst signal. This signal is reflected by the object to be measured, say the water surface in a rainwater barrel, and received back by the ultrasonic module as an echo signal. Since the sound traverses the distance to the reflecting object twice during the echo period, we can state that: distance = 0.5 x velocity of sound [m/s] x echo period [5] Circuit and components ‘The schematic in Figure 1 Is based on three components, the well-loved ATmega8 microcontroller, a LCD connected to this plus the ultrasonic module. As well as the HC-SR04 module (Figure 2) you can also use the almost identical = Ue 5 Figure 1. Simplified circuit of the Distance and Level Gauge.Figure 2. ‘The compact ultrasonic ‘module HC-SRO4. Figure 3. Procedural diagram forthe US-020. Both modules are in the $5 / £3 price lass, making them unbeatably attractive for use in a do-tt-yourself measurement device. Ascertaining the manufacturers of these modules Is not an easy task but they are widely available in electronics stores, particularly those that spe- lize in robotics. When you check out the data sheets, however, you will discover wide varia- tlons. One time the detection distance for the ‘two models will be shown as 2 cm-3 m (.8 inch = 10 feet), another time as 3 cm - 7 m (1.2 inch = 23 feet). Further investigation will uncover other divergent detalls, even for one and the self-same model. Hence the current consumption will vary between 3 mA and 15 mA depending ‘on which data sheet you read. Reliable data can bbe determined only by doing your own research. Generally suppliers of these modules agree on the following data: ‘* Operating voltage: 5 V DC ‘* Temperature range: 0-70 °C ‘© Diffusion angle: 15° ‘© Operating frequency: 40 kHz ‘Trigger input signal: 10 ps ‘The minimum distance between transmitter and receiver capsules for the HC-SR04 module is so ‘small that measurements can be made in can- Isters (the ultrasonic module needs to be dis- charged when dealing with highly inflammable fluids, owing to risk of explosion!). If you are taking measurements in containers with particularly small aperture diameters you can resort to the ‘SRFO2 ([4] ultrasonic module, which although not. pin-compatible, is nevertheless equipped with a serial/I°C interface and uses a single capsule for both transmitting and recelving. As the transmity receive angle Is very acute, with this model you need to bear in mind that the minimum distance from the medium being measured amounts to 16cm. To simplify the menu guidance a four-line, 16-character LCD display is employed. Potenti- ‘ometer P1 adjusts the contrast. R3 is the series resistor for backillumination and should amount to around R3 = (5- Vi) / Trem [2].For this you need to consult the data sheet of the display. With many LCDs the series resistor Is already provided internally and in that case you can replace R3 with a wire bridge or strap. To save energy you might also consider a setup involving a switch. ‘The changeover relay RE1, whose contacts are connected to K3, are controlled by port pin PD7 and driver transistor T1, and the alarm LED (low-current) by port pin PDS. A low-dropout voltage regulator LM2940CT-S ensures a stable 5 V supply voltage from four 1.5 V batteries, This, somewhat exotic voltage regulator delivers up to 1A and can therefore be loaded much more heavily than the usual 78LO5 with its 100 mA. ‘Naturally you can also use a plug-in (‘wall wart’) AC power supply (<26 VDC) for this project, connected to K1. With the relay energized and the LCD illuminated the circuit draws around 150 mA. To achieve maximum accuracy of measurement, the microcontroller needs an external 16 MHz crystal. A couple of passive components complete the circuitry. The software The program in the microcontroller must cover the following functions: ‘ Indication of the level of liquids ‘© Watchdog monitoring of this level with relay output and LED alert ‘Infinitely variable programming of the mir/ max alarm level «Storage of min/max calibration values for up to ten containers or tanks « Distance measurement «Intuitive menu guidance using LCD readout * Offset correction for distance measurement ‘* Manual divisor correction to accommodate low and high temperature extremes The software follows the procedure shown in Fig- ture 3. In view of the limited amount of memory space In the ATmega8, readout messages are given in English only. Operation makes use of the four pressbuttons Return, Next, Plus and Minus. Because of the complexity of the program coding, which can be downloaded along with the hex file and the PCB layout at [2], it has been subdi- vided into meaningful procedures and provided with corresponding commentaries, so that even beginners should find their way through it rapidly. Ponca maul cuts eRe memer tl ‘The velocity of sound cy in air amounts to 331.5 m/s measured at a ‘temperature of T = 0°C. The velocity of sound is not dependent on ‘atmospheric pressure but purely on the air temperature 0 in degrees Celsius: =o: Via) where « is the coefficient of expansion of 1/273.15 = 3.661x10-? 1/9C. From this arises the approximation: 9 = 20.063-V0+273.15 m/s We can see that for precision measurement the influence of ‘temperature is by no means insignificant. Here are some examples of sound propagation times at 0 °C and 20 °C. nce incm Time elapsed in ms at 20°C Time elapsed in ms at 0°C 2 0.117 0.121 10 0.583 0.603 50. 2.915 3.017 100) 5.831 6.033 200 11.662 12.066 300 17.492 18.100 ‘The actual measurement process follows. To bring the ultrasonic module connected to K2 into trans mittting a signal, a pulse 10 pis long with a fall- Ing edge Is sent via port pin PD2 to the trigger Input. As a result, about 250 uss afterwards, the ultrasonic module transmits a burst signal with a frequency of 40 kHz and duration of 200 ys. The echo output, which is connected to the microcontroller via port pin PD3, switches now to High level and the ultrasonic module stands by to receive the reflected signal. If this occurs, the echo output fips back to Low. Timert Is ticking away throughout this process, so that after the timer Is halted, the distance can be calculated. After 50 ms if no echo signal arrives, measure ‘ment for the current cycle Is cancelled. To achieve the best possible results the mensuration cycle Involves 16 separate measurements. The program code for this can be examined in the procedure Gauge_distance(). Now a word of advice for people doing their own “burning’ and taking advantage of the programming interface for the controller at K4, BecauseFigure 4 Upper and lower sides of the PCB. Component List iow current, re, cs47C uM29407-4.0 Miscellaneous K1 = DC-power socket, 1.95mm pin (Farnell 1217037) der the program is extremely wide-ranging, the 8 K-capacity flash memory Is 100 % full. For this reason the demo version of BASCOM-AVR cannot be used for compiling, since it allows only 4K (Le. 50%) of memory usage. While burning you can rapidly blunder into a trap: in this circult the ATmega8 operates with an external 16 MHz crystal, making It necessary to set two fuses. CKOPT should be activated to assure reliable operation of the oscillator. This method increases current consumption somewhat but you do obtain stable oscillation in return. It’s essential to set SUT_CKSEL to Ext. Crystal/Resonator High Freq.: Start-up time: 1K CK + 64 ms! The list of options provides a wide selection of settings. If you make a mistake here, under some cdrcumstances you can ‘brick’ your ATmega8 once programmed, as it may well be unable to find its feet again afterwards. Better then to double check precisely which settings you have selected, A screendump of AVR Studio showing the correct fuse settings is included in the download package. Construction and commissioning Construction Is extremely simple thanks to the very manageable number of components. You can buy the double-sided printed circutt board (Fig- ure 4) from Elektor or etch it yourself at home. The relevant files are In the download package [2]. Some handiwork and skill are necessary to create cutouts in the case for the LCD readout, the ultrasonic capsules, the pressbuttons and the switch, How you arrange the ultrasonic capsules is a matter of personal taste. Installing the ultra- om seteae or) Perrot resonic modules separately on a moderately long screened cable Is also concelvable, in case reading needs to take place in another room. Once the programmed ATmega8 has been inserted and the ultrasonic modules connected, you can switch on the device for the first time. At very least the background lighting for the LCD should illuminate. If this is not the case, disconnect the power supply and double-check the component placing! During the start phase Dis- tance & Level Gauge 2.4 (Figure 5) should be visible on the display for two seconds. ‘Now you can perform the basic settings. To do this you switch off the device, then turn it on again while holding down the Return button. Having released the button, you will be asked the width of the container being measured. At this point you can press the Plus and Minus buttons In order to enter an offset in 5 mm increments for taking into account the distance from the top of the ultrasonic module to the leading edge of the container. This gives you a simple way, for example, of measuring the distance from one wall to the other later on. The best method for this is to take a2 m reference measurement with a two-meter wooden rule (or measure three feet using a carpenter's yardstick). The value Is confirmed using Return. Following this you are asked for the divisor value used for correcting the value measured. This Is necessary since the velocity of sound Is temperature-dependent. In the temperature range from -20°C to +40°C the velocity of sound varies from 312.85 m/s up to 349.32 m/s. The deviation is virtually linear, so that for a temperature variation of 10°C you can expect an inaccuracy of about 2%. Panel 1 contains some formulae and a table of values for the elapsed times to The latest on electronics and information technology ¢ Videos, hints, tips, offers and more Exclusive bi-weekly project for GREEN and GOLD members only Elektor behind the scenes In your email inbox each Friday @ektor Sean bee ems te Wo net Register today at www.elektor.com/newsletter, (DeteeTV goes naebe expected. Incidentally, atmospheric pressure has no influence on the veloc~ ity of sound. ‘Anyone prepared to accept up to 5% inaccuracy in this connection need do nothing at all. On the other hand, if you wish to take measurements in temperatures below 15 °C or above 25 °C and with a higher level of accuracy, a division factor must be introduced. The higher the temperature, the greater the divisor must be. The values for the width of the container and the divisor must also be retained after a battery change and can be entered afresh at any time, The Select mode: menu should now be visible. Pressing the Return button will take you to the tank selection menu, The opportunity is provided to carry out calibration functions for up to ten different tanks or other containers. The only condition made is that the container shape must be as close as possible to a cuboid or cylindrical form, otherwise linear measurement Is impossible. Hav- ing selected your tank number you land back at the main menu. After pressing the Next button the menu option CaF Ibrate tank appears. Hit- ting Return now goes to the calibration menu. Having selected empty-level calibration, you are asked if you are ready to perform the calibration. You don’t necessarily need to empty the vessel in question for this purpose; instead you can simply measure outside the tank the height from the filler neck down to the ground. Once the empty level has been calibrated, you reach the menu for full-level calibration (Fig- ure 6). The default value is set as a minimum of 3 cm, which you can accept by pressing Next. If, however, you wish to perform a calibration for the full-level, put the device in position above the filled container and press Return. After returning to the main menu you can perform the first level measurement. To do this, select menu option Gauge level and confirm with Return. Measurement begins now and the result is displayed. Next carries out a fresh measurement, Return leads back to the main menu. To test out level monitoring you should select the Level Watchdog menu. In the menu that follows (Figure 7) the alarm type can be selected: Full will activate the alarm when the permitted fil quantity is exceeded and Empty when it falls short. Having chosen your alarm type, you can alter the threshold value in the Level Alarm menu (Figure 8). For the default values we have defined the maximum level as 95 % and the minimum level as 5%, but these can be varied using the Plus and Minus buttons. All the calibration values made are stored in the EEPROM and are retained there even after a change of battery. The first measurement follows next and the current level Is displayed (Figure 9). Measurements are now taken every five seconds. If these fall below the minimum value or exceed the max- Imum value set, the display shows > Level alarm <, the red LED illuminates and the relay operates, for example to switch on a pump. If the level status returns into the permitted range after 5 s, the alarm function is cancelled. The 5s measurement cycle avoids the need for a hyster- esis function to eliminate relay contact bounce. To exit from this mode back to the main menu, ress the Return button (for up to 5 s). You can now select the Gauge dlstanae menu. Each press on the Next button initiates a distance measurement (Figure 10). As before, you can exit from this mode and go back to the main menu by pressing Return (140208)By J.T. van Es (Netherlands) Figure 1. Basic circuit. Figure 2. Output signal (10 Hz) without positive feedback. Figure 3. Positive feedback via R4RS. Figure 4. ‘Output signal (10 Hz) with positive feedback. Improved Current Transformer This report is a response to the two publications by Martin Ossmann in the April and May 2014 editions about the dimensioning and use of current transformers. Martin aims to reduce the lower cut-off frequency, determined by L/R, by looking at the core material, among other things. While the method is not incorrect, similar results can be achieved much easier as shown in the report below. Tr i f ay Son 8g Figure 1 shows the signal path of a current meter built in 2005 and which was appropriate for its purpose. It ended up in a closet after that, but ‘was retrieved again for this occasion. The current transformer Is a pick-up coil from a ‘Tubantor’ fuel-infection system from 1983. This contains a ferrite core with 80 turns. The -3 dB bandwidth of this assembly is 7 Hz to 5.5 MHz. Using it with 2 10-Hz square wave Is therefore not such a success, as Figure 2 shows. R3 has been added to correct the offset of the amplifier. By giving the amplifier a little bit of positive feedback, using R4 and RS (see Figure 3), the input resistance of the amplifier becomes negative (adjustable with RS) and this effectively compensates for Roy. When the circuit resistancereduces, L/R increases, Figure 4 shows this effect. It even works well down to 1 Hz, if the positive feedback Is increased a litle more, see Figure 5. However, the DC stability leaves much to be desired. Figure 6 shows a solution for this, using ‘a control circuit comprising an additional opamp, R7 and C1. The opamp needs to be a type with FET inputs so that It becomes possible to work with large values for R7 and C1. C1 needs to be a film capacitor in order to keep the leakage ‘current low. Because the secondary winding is almost perfectly short-circuited using this technique, there can be no voltage drop across it any more—and not ‘across the primary either, apart from its own ‘copper resistance. The flux in the core, which is already low in a current transformer, also ‘approaches the absolute minimum. The value for L can be much lower in this case, which Is beneficial for its HF behavior. The measurement of DC signals is, however, not possible, this requires ‘Hall effect sensor. The annoying feature of such a device Is that It requires a gap in the core, ‘which results in a much reduced mu and therefore value of L. Figure 5. Output signal (1 Hz) with ‘more positive feedback. In the past (1985, I think) we designed a beam ‘monitor using a Vitrovac ring core with an inside diameter big enough to pass a beer bottle through. The bandwidth we obtained then was 23 mHz to 20 MHz! Figure 6. Improved DCstability with ‘an additional opamp and RICA. (140266) ‘The author's amplifier setup,Safe High Voltage Power Transistor Measuring with Oscilloscopes ‘Sacig's CLIPPER CLP1S00V15AL is a high technology oscilloscope adapter that allows small voltages to be measured in the presence of very high voltages, such as those present on switching MOSFETs or IGBTs (Insulated Gate Bipolar Transistors). Making accurate measurements in the presence of very high voltages has previously been extremely challenging, but using the CLIP- PER CLPISO0V1SAL, the Vizou or Rycu Of @ Switching transistor can now be easily seen in high resolution e.g. 100 mV/div) on an ordinary digital esciloscope. Determining the switching speeds, Safe Operating Area (SOA), and temperature effects of MOSFETs and IGBTs now greatly simplified with the CLIPPER CLPISOOV1SAL. ‘The CLIPPER CLP1SOQVISA1 represents 2 leap forward in transistor switch testing, ‘essential both in the design verification phase and production process monitoring. In typical use, the CLIPPER is connected between Drain and Source terminals of a Power MOSFET. ‘The output signal provides a 1-to-1 picture of what is occurring across the MOSFET, without the

FM Transmitter
No ratings yet
FM Transmitter
1 page
Sony Icr-4800 Service Manual
100% (2)
Sony Icr-4800 Service Manual
13 pages
Everyday Practical Electronics 2000 03
100% (2)
Everyday Practical Electronics 2000 03
84 pages
Ultra-Simple Transmitter
100% (2)
Ultra-Simple Transmitter
2 pages
Low Pass Filter For FM 88-108 MHZ
No ratings yet
Low Pass Filter For FM 88-108 MHZ
3 pages
Polar T1200 Users Manual
100% (1)
Polar T1200 Users Manual
8 pages
2P3 Superheterodyne MW Radio Kit Manual English
No ratings yet
2P3 Superheterodyne MW Radio Kit Manual English
15 pages
Building The SAP-3 Rev 2.0 PDF
100% (1)
Building The SAP-3 Rev 2.0 PDF
58 pages
Electronic Mechanic 3 RD Semester NSQFQBCTS
No ratings yet
Electronic Mechanic 3 RD Semester NSQFQBCTS
45 pages
Dimmer Technology Overview
No ratings yet
Dimmer Technology Overview
75 pages
Persistence of Vision LED Clock Project
No ratings yet
Persistence of Vision LED Clock Project
10 pages
FM Phone Transmitter Circuit Guide
No ratings yet
FM Phone Transmitter Circuit Guide
2 pages
PC Scope Probe Circuit Design Guide
No ratings yet
PC Scope Probe Circuit Design Guide
4 pages
FG
No ratings yet
FG
2 pages
Manual Denon AVR-3802
No ratings yet
Manual Denon AVR-3802
82 pages
Elektor Electronics (UK) 1998 - Cumulative Index
No ratings yet
Elektor Electronics (UK) 1998 - Cumulative Index
5 pages
GE4139 Manualmain
100% (1)
GE4139 Manualmain
5 pages
Frequency Counter (Mine Worked Up To 70MHz) Using PIC 16F628A
No ratings yet
Frequency Counter (Mine Worked Up To 70MHz) Using PIC 16F628A
6 pages
Elektor USA - September 2014
No ratings yet
Elektor USA - September 2014
71 pages
Electronics Enthusiasts' Guide
No ratings yet
Electronics Enthusiasts' Guide
72 pages
Pico C (English) - Part1
100% (2)
Pico C (English) - Part1
6 pages
32-Bit Microprocessor (Stand-Alone) : Labvolt Series
No ratings yet
32-Bit Microprocessor (Stand-Alone) : Labvolt Series
5 pages
LCD KeyPad Shield For Arduino
No ratings yet
LCD KeyPad Shield For Arduino
9 pages
Understanding the 555 Timer IC
No ratings yet
Understanding the 555 Timer IC
7 pages
Marconi 2955A Service ID7810
No ratings yet
Marconi 2955A Service ID7810
433 pages
SC Analyzer 2005 Elektor Semiconductor Tester
No ratings yet
SC Analyzer 2005 Elektor Semiconductor Tester
9 pages
Elektor Electronics 1998-03
No ratings yet
Elektor Electronics 1998-03
58 pages
Philips Tube Codes
No ratings yet
Philips Tube Codes
14 pages
Re - 1982-11
No ratings yet
Re - 1982-11
140 pages
All-Waveband Ferrite Rod Antenna
100% (1)
All-Waveband Ferrite Rod Antenna
4 pages
UV916 TV Tuner
No ratings yet
UV916 TV Tuner
82 pages
DN-V210/DN-V310 DVD Player User Guide
No ratings yet
DN-V210/DN-V310 DVD Player User Guide
83 pages
1080P Security: Electronics Components
No ratings yet
1080P Security: Electronics Components
8 pages
Ee 1980 03
No ratings yet
Ee 1980 03
47 pages
DigitalElectronics WhenHardwareGreetsSoftwareV10k Nodn
No ratings yet
DigitalElectronics WhenHardwareGreetsSoftwareV10k Nodn
167 pages
SSAVI Cable Scrambling System
No ratings yet
SSAVI Cable Scrambling System
3 pages
ETI Project Index: SDVR-1503
No ratings yet
ETI Project Index: SDVR-1503
12 pages
Electronic - Schematic Circuits For The Hobbyist
No ratings yet
Electronic - Schematic Circuits For The Hobbyist
159 pages
Elektor - December 1974
No ratings yet
Elektor - December 1974
54 pages
Re - 1983-09
No ratings yet
Re - 1983-09
132 pages
4 Stage FM Transmitter
No ratings yet
4 Stage FM Transmitter
1 page
Robotics For Class VII (Revised)
No ratings yet
Robotics For Class VII (Revised)
17 pages
POLAR T1200 Service MANUAL
100% (1)
POLAR T1200 Service MANUAL
17 pages
Nuts and Volts 2008-07
No ratings yet
Nuts and Volts 2008-07
101 pages
4 X 50 W MOSFET Quad Bridge Power Amplifier: Features
No ratings yet
4 X 50 W MOSFET Quad Bridge Power Amplifier: Features
18 pages
Pic Graphics L.C.D. Scope: Ccoonnssttrruuccttiioonnaall Pprroojjeecctt
No ratings yet
Pic Graphics L.C.D. Scope: Ccoonnssttrruuccttiioonnaall Pprroojjeecctt
5 pages
One Transistor Pocket Radio
No ratings yet
One Transistor Pocket Radio
4 pages
Bit Scope Programming
100% (1)
Bit Scope Programming
16 pages
1W and 5W Power Amplifier
No ratings yet
1W and 5W Power Amplifier
4 pages
Parallel Port: Labeled Back View of Computer Case
100% (1)
Parallel Port: Labeled Back View of Computer Case
33 pages
Embedded Electronics Insights
No ratings yet
Embedded Electronics Insights
84 pages
Compact Stereo FM Transmitter Guide
No ratings yet
Compact Stereo FM Transmitter Guide
10 pages
DTMF Decoder Using MT8870
No ratings yet
DTMF Decoder Using MT8870
8 pages
Semiconductor Rectifier Specs
No ratings yet
Semiconductor Rectifier Specs
8 pages
Elektor USA - December 2014
No ratings yet
Elektor USA - December 2014
76 pages
Android Tablet: Electronic LED Candle
No ratings yet
Android Tablet: Electronic LED Candle
88 pages
Elektor Electronics UK 2014-01-02
No ratings yet
Elektor Electronics UK 2014-01-02
132 pages
Elektor Electronics 2014-01,02
100% (2)
Elektor Electronics 2014-01,02
132 pages
Elektor Electronics 2014-06
100% (4)
Elektor Electronics 2014-06
92 pages
Ee 1992 12
100% (2)
Ee 1992 12
89 pages
Ee 1992 04
100% (2)
Ee 1992 04
56 pages
Ee 1992 01
No ratings yet
Ee 1992 01
57 pages
Ee 1992 02
100% (1)
Ee 1992 02
59 pages
Ee 1991 07 08
No ratings yet
Ee 1991 07 08
92 pages
Ee 1991 06
100% (1)
Ee 1991 06
59 pages
Ee 1990 03
No ratings yet
Ee 1990 03
55 pages
Ee 1990 02
No ratings yet
Ee 1990 02
54 pages
Elektor Electronics 1998-11
No ratings yet
Elektor Electronics 1998-11
54 pages
Ee 1990 01
No ratings yet
Ee 1990 01
56 pages
Ee 1989 10
No ratings yet
Ee 1989 10
43 pages
Ee 1989 12extra
100% (2)
Ee 1989 12extra
80 pages
Ee 1988 03
No ratings yet
Ee 1988 03
47 pages
Elektor Electronics 1998-02
No ratings yet
Elektor Electronics 1998-02
51 pages
Ee 1987 11b
No ratings yet
Ee 1987 11b
76 pages

Elektor Electronics 2014-12

Uploaded by

Elektor Electronics 2014-12

Uploaded by

You might also like