By Alan Bloom, N1AL

A Modern GDO––
The “Gate” Dip Oscillator
Building an RF circuit? Use an up-
dated version of an indispensable
and legendary piece of equipment
to test it––the “dipper.”

still have my first grid-dip oscillator, purchased at a flea

I market about 30 years ago. It’s an Eico model 710 that

covers 0.4 to 250 MHz with 8 plug-in coils. It weighs over
2 pounds and must be plugged into the wall to obtain power
for its 6AF4 vacuum tube. It still works, although the coil
socket has become intermittent over the years, which makes it
a bit frustrating to use.
Amateurs have been using grid-dip oscillators (GDOs) at
least since 1947. 1 While these days they no longer have a
vacuum-tube “grid” to dip, the name seems to have stuck. A
GDO is a tunable oscillator with the coil mounted outside of
the chassis. The external coil allows you to measure the fre-
quency of a tuned circuit without any electrical connection to
it. Just bring the coil close to the tuned circuit’s coil and tune Figure 1––The “gate dipper” fits comfortably in the hand.
the GDO while watching for a dip in its meter reading.
This “no connection” measurement capability is handy for
other purposes as well. An example is measuring the parasitic When you are troubleshooting a receiver, a GDO can help
elements of a beam antenna. Since the parasitic elements have isolate the fault. Tune the GDO to each of the IF frequencies,
no feedpoint, you can’t measure the resonant frequency the starting with the last and hold the coil close to that part of the
normal way, with an SWR meter. A GDO is also perfect for circuitry. If you can hear a signal, then that IF and all the cir-
tuning the traps to resonance on a multi-band Yagi. cuitry after it are working. This is especially handy on densely
A GDO can “sniff out” spurious resonances in the tank cir- packed, surface-mount boards that are difficult to probe.
cuit of a high-power vacuum-tube RF amplifier. Be sure to If the GDO has an RF output connector, it can be used as a
turn off the high voltage first! “poor ham’s signal generator” in applications where good fre-
A GDO has other uses you might not think of. For example, quency stability and calibrated level accuracy are not needed.
it can measure inductance. Just temporarily connect a capaci- An example is functional “go/no-go” testing of devices such
tor with a known value in parallel with the coil to be tested, as amplifiers, mixers and wide-band filters.
measure the resonant frequency and use the formula L = With the oscillator turned off, the instrument functions as a
1/(2πf)2C or use the graph in The ARRL Handbook.2 (L is in- type of tuned RF detector known as an absorption frequency
ductance in µH, C is capacitance in µF and f is frequency in meter. The obvious use is to check if there is RF energy of the
MHz.) It is just as useful for measuring unknown capacitors. proper frequency in a tuned circuit. It also makes a handy fre-
Variable capacitors normally are not marked with their value. quency-selective field-strength meter using the capacity probe
To identify that “bread slicer” you bought for a song at the as an antenna. Another use is to check for feed line radiation.
local hamfest, just connect it in parallel with a known If you run the coil along the coax cable while transmitting, the
inductance, measure the frequency, and use the formula meter will show the peaks and valleys of the standing waves
C = 1/(2πf)2L. of the unwanted RF current flowing on the outside of the coax
shield. A GDO also makes an excellent “sniffer” to check for
1
Notes appear on page 58. RF leakage from a shielded transmitter chassis, both at the
54 May 2003

bloo.pmd 54 3/26/2003, 9:53 AM

 fundamental and harmonic frequencies.
Headphone output was indispensable in the “old days” to
check the modulation quality of AM transmitters. Hams don’t
use AM very much any more, but headphones are still useful
to listen for key clicks, hum and low-frequency parasitic os-
cillations. And, if you couple the coil to an antenna, you have
a wide-range “crystal” set!
My favorite, literally “off-the-wall,” application was men-
tioned in a comprehensive tutorial on grid-dip oscillators in
January 1974 QST.3 A GDO makes a sensitive detector of pipes
and other metallic objects inside a wall. If you listen to the
signal on a CW or SSB receiver, you can easily hear a change
in frequency whenever the coil passes near a metal object.
The Design
I got the idea for this design while reading an article in
Amateur Radio, the journal of The Wireless Institute of Aus-
tralia, by Lloyd Butler, VK5BR.4 The nice feature of his cir-
cuit is that it uses two-terminal plug-in coils. One terminal is
grounded so you can use a rugged, reliable coax connector for
the coil socket. Most GDO designs require three connections
to the coil or two connections that must both be insulated from
ground.5 Many of the latter also require a two-section tuning
capacitor, which can be hard to find.
The old Heathkit “Tunnel Dipper” had two-terminal coils
using a tunnel diode in a negative-resistance oscillator. Tun-
nel diodes are very hard to find nowadays, so VK5BR made a Figure 2––Photograph of the GDO with the cover off. The
negative resistance element by connecting N-channel and tuning capacitor was oriented for the shortest possible
connection to the coil connector.
P-channel JFETs back-to-back (gate to drain). I tried it and it
oscillates beautifully. In fact, that is just the problem; it oscil-
lates so strongly that it is difficult to get a good dip. VK5BR’s
solution was to connect a selected resistor in parallel with each
coil to reduce the oscillation amplitude.
I decided to go a different route. The resistor kills the coil
Q, which reduces the sensitivity and widens the bandwidth in
wavemeter mode. Also, P-channel JFETs are becoming hard
to find. Finally, I wanted to eliminate the internal calibration
potentiometer that sets the operating point of the oscillator.
Figure 1 shows the completed dipper, Figure 2 is a view in-
side the case and Figure 3 pictures the complete coil set. The
schematic and parts list are shown in Figure 4. A pair of source-
coupled N-channel JFETs (Q1 and Q2) form the oscillator
portion of the circuit. Unlike some GDO designs, no RF chokes
are required. Because of self-resonances, chokes don’t main-
tain a high impedance over the wide range of frequencies re-
quired for a GDO.6 This unit has no false dips anywhere in the
frequency range.
Figure 3––This is how the coils looked before the heat-shrink
Q4, a bipolar 2N3904 transistor, serves a dual purpose. Its tubing was applied. Small pieces of electrical tape hold the
base-emitter junction acts as the RF detector. Further, it ampli- turns in place during construction.
fies the rectified current flowing in the base and sends it to
emitter-follower Q5, a 2N2907 transistor. I used a 200 µA meter
because that was what I found in my junk box. If you use lator and wavemeter mode and also controls the volume to the
another value, change the value of R5 using the formula R5 = headphones. Battery current is approximately 3 mA with the
(1.5 / Im) – Rm, where Im is the full-scale meter current and Rm oscillator on and zero in wavemeter mode, when no signal is
is the meter resistance. Be careful when measuring meter inter- being received.
nal resistance. Measure the test current of your ohmmeter (on
the scale you will use) with another meter. A sensitive meter Construction
can be damaged by excessive measurement current, although This project was built almost entirely from junkbox parts.
most newer digital meters use test currents in the range of 50- The parts list in Figure 4 consists of parts selected to be simi-
200 µA for their ohmmeter scales. Transistor Q3 is a JFET lar to those used in the prototype. Feel free to substitute what
source-follower amplifier for the RF output connector. The RF you have on hand. The exceptions are transistors Q1, Q2 and
output may be used as a signal source or to drive a frequency Q4, which should be the types specified or their equivalents.
counter for more accurate frequency readout. The 2¼ × 2¼ × 5 inch chassis is a compromise; it’s large
In detector mode, the power to the oscillator and RF buffer enough to allow all parts to fit easily and small enough to fit
is turned off. Q4 detects and amplifies any signals picked up comfortably in the hand. The coil was mounted off-center, both
by the coil. The meter sensitivity control works in both oscil- to allow the shortest connection to the tuning capacitor and to
May 2003 55

bloo.pmd 55 3/26/2003, 9:56 AM

 afford easier coil coupling to an external circuit. chased at the local hardware store. The inside diameter is ac-
The prototype used a perforated test board for the RF por- tually slightly less than 3/8 inch, which makes a nice force-fit
tion of the circuitry and a solder tie strip for the meter cir- onto the 3/8 inch mounting threads of a male chassis-mount
cuitry. A printed circuit board pattern and parts layout are avail- BNC connector. The lowest-frequency coil is wound on a large
able at www.arrl.org/files/qst-binaries/gdo.zip, should you pill bottle with two 1-inch diameter aluminum washers at the
choose this construction route (the GDO can be hand-wired). connector mount for added strength.
Note that the battery is connected with the positive terminal First run the wire through the center of each form and sol-
to ground, backwards from the normal arrangement. der it to the BNC connector’s center conductor. Then press
The coil forms were made from “3/8 inch” water tubing pur- the form onto the connector threads and cut a small notch in

Figure 4––Schematic diagram of the GDO. Part numbers listed are from Ocean State Electronics (OSE) (800-866-6626,
www.oselectronics.com), Alltronics (408-847-0033, www.alltronics.com), Digi-Key (800-344-4539, www.digikey.com) and
Mouser Electronics (800-346-6873, www.mouser.com). Resistors are available from all of the above sources.
B1, B2––1.5 V AA or AAA cell. Q1, Q2, Q3––MPF102 transistor, JFET, Digi-Key MPF102-ND,
C1––75 pF variable, OSE AVC75 or 365 pF variable BC14400. Mouser 512-MPF102.
C2, C8––5 pF disc capacitor, Digi-Key P11408CT-NF, Q4––2N3904 transistor, Digi-Key 2N3904-ND,
Mouser 75-10TCCV50, OSE CD5-5. Mouser 610-2N3904.
C3, C6––0.1 µF capacitor, Digi-Key BC1084CT-ND, Q5––2N2907A transistor, Digi-Key PN2N2907-ND,
Mouser 21RX310. Mouser 610-2N2907A.
C4––0.01 µF capacitor, Digi-Key BC1078CT-ND, R1––470 Ω, ¼ W resistor.
Mouser 21RX410. R2, R4––1 kΩ Ω, ¼ W resistor.
C5––0.001 µF capacitor, Digi-Key BC1072CT-ND, R3––1 MΩ Ω, ¼ W resistor.
Mouser 21RX510. R5––See text.
C7––10 µF, 10 V capacitor, Digi-Key P5134-ND, R6––100 kΩ Ω potentiometer, linear taper, Mouser 313-1000-100K.
Mouser 140-XRL10V10, OSE CER10-25. S1––SPST toggle switch, Digi-Key EG2350-ND, Mouser 1055-
D1––1N4148 diode, Digi-Key 1N4148DICT-ND, TA1120.
Mouser 512-1N4148, OSE 1N4148. Misc
J1, J2––BNC female chassis-mount connector, Alltronics #18, #22 and #26 gauge enameled wire, Mouser 501-MW18H-
CB122, Digi-Key ARFX1062-ND. 1LB, 501-MW22H-1LB, 501-MW26H-1LB, respectively.
J3––3.5 mm stereo phone jack, Digi-Key SC1094-ND, Mouser
10 BNC male chassis-mount connector (for coils), Alltronics
161-3402. CB111.
J4––Phone tip jack, Mouser 530-105-0802-1; phone tip plug, Dual AA or AAA battery holder, Digi-Key BC2AAAW-ND,
Mouser 530-105-0302-2.
BC2AAW-ND, Mouser 122-0421.
M1––0-1 mA, 15/8 inch square panel meter, OSE 60-158 or 2¼″″ × 2¼″″ × 5″″ chassis, Digi-Key HM611-ND, Mouser
0-200 µA, 11/2 × 7/8 inch edge-mount panel meter MS200 563-CU-3004A, OSE 1411KU.
(see text).
3 ft 3/8″ PVC water tubing for coils.
56 May 2003

bloo.pmd 56 3/26/2003, 9:56 AM

 largest were close-wound. On the 0.9-1.5 and 1.5-2.8 MHz
coils, there were too many turns to fit in a single layer so I
“scramble” wound them by overlapping turns, a few at a time,
as I wound the coil. It would be neater to lay the wire in two
overlapping close-wound layers, but that increases the inter-
winding capacitance which can cause spurious resonances and
reduced tuning range.
I extended the range of the lowest-frequency coil by con-
necting it to the GDO through a BNC “T” connector with a pair
of 53 pF capacitors attached with clip leads. (Of course it also
would have been possible to wind another coil, but I only had
one large pill bottle.) Even for the other coils, adding extra ca-
pacitance is a useful trick because it slows down the tuning
rate, handy for measuring narrow-band devices like crystals.7
With the smallest coil, the GDO oscillates over only a small
portion of the capacitor tuning range. Fortunately, it happens
to cover the two-meter ham band. With the second-smallest
coil, the oscillation amplitude drops off at the low end, but it
is usable down to about 62 MHz. The circuit would probably
work better at VHF if it were constructed with a lower-induc-
tance tuning capacitor.
The small 75 pF tuning capacitor gives only about a 2:1
tuning range. The advantage is that tuning is not so critical
and the frequency dial is easier to read. The disadvantage is
that it takes more coils to cover the desired frequency range.
If you substitute a 365 pF AM broadcast radio tuning capaci-
Figure 5––The dial scale for the prototype GDO. If you make tor, you should get better than 3:1 tuning range from each coil.
your own GDO you will almost certainly need to lay out your The tuning dial was made from a circular piece of ¼ inch
own scale.
clear plastic glued to a knob. Making the diameter slightly
larger than the chassis width allows one-handed operation us-
ing your thumb to adjust the tuning. The scale, cemented to
the other end for the wire to go through. After winding the the chassis under the clear plastic dial, was drawn using a com-
coil, cut and tin the wire end and solder it to the ground lug puter graphics program. Measuring the calibration data and
mounted on the connector. I covered each coil with a layer of designing the scale were very time-consuming. A better solu-
heat-shrink tubing or electrical tape to hold the turns in place tion would be to include a built-in frequency counter. A 4 or
and to protect the wire. Figure 3 shows the coils before the 5-digit readout would be accurate enough since a GDO’s fre-
heat-shrink tubing was added. quency stability is not good enough to make a high-accuracy
Winding data is listed in Table 1. Unless you happen to readout useful. Figure 5 illustrates the homebrew dial.
duplicate my unit exactly, the frequency range of each coil is
likely to be different. However, this data can be used as a start- Operation
ing point for figuring out your own coil designs. To measure the resonant frequency of a tuned circuit, switch
The wire for the two highest-frequency coils is bare cop- to oscillator mode and adjust the meter sensitivity to about ¾
per, salvaged from scraps of Romex house wiring. The other scale. Then orient the GDO coil close to, and approximately
coils are wound with enamel-insulated magnet wire. The ex- parallel to, the coil under test and tune the dial until you get a
act wire gauge is not critical (although it may affect the num- strong dip on the meter. Tune slowly or you may miss the dip.
ber of turns required). A good source for fine-gauge wire for Overly close coupling to the circuit under test causes such a
the low-frequency coils is a deflection yoke removed from the strong dip that the oscillator is pulled far off frequency. For
cathode-ray tube of a defunct television set. the most accurate measurement, use close coupling to find the
The three smallest coils (29.5-150 MHz) were space-wound approximate frequency, but then move the GDO coil farther
to the lengths listed. The middle 4 coils (2.8-35 MHz) and the away until the dip is barely visible and retune. Figure 6 shows

Table 1
Coil Winding Data for the GDO
Wire Gauge Form Diameter Form Length Coil Length Number of Turns Frequency Range
#12 0.375" –– 0.125" 2 130-150 MHz
#14 0.5" 1.5" 0.5" 3 62-108 MHz
#18 0.5" 2.0" 0.4" 5 29.5-62 MHz
#18 0.5" 2.5" 0.5" 10½ 16.5-35 MHz
#22 0.5" 2.5" 0.6" 21 9.2-19 MHz
#26 0.5" 2.5" 1.0" 46 5.1-10.5 MHz
#30 0.5" 2.5" 1.5" 100 2.8-5.6 MHz
#30 0.5" 2.5" 1.6" 180 1.5-2.8 MHz
#30 0.5" 3.9" 3.3" 390 0.9-1.5 MHz
#30 1.25" 3.9" 3.4" 230 620-980 kHz
(+53 pF) 505-640 kHz
(+106 pF) 440-525 kHz

May 2003 57

bloo.pmd 57 3/26/2003, 9:57 AM

 the dip meter in use, measuring the frequency of an amplifier
tank circuit.
Coupling to toroids or shielded coils can be difficult. One
solution is to connect a wire from the capacitance probe to the
“hot” end of the circuit to be tested. For looser coupling, just
place the wire close to the circuit under test. When measuring
the inductance of a toroid with a test capacitor, the capacitor
leads often form enough of a loop to allow coupling to the
GDO coil. Some authors recommend coupling the GDO to a
one-turn loop through the toroid, but that makes the toroid act
like a transformer with a shorted secondary and that changes
its inductance.
Antenna measurements are best made at the high-current
point of the antenna conductor. For a half-wave dipole, this is
near the center. Orient the coil perpendicular to the conductor
for maximum coupling. Be sure to short the feed point of the
antenna before making the measurement. If you can’t find the
dip, make the shorting wire into a 1-turn loop for better cou-
pling to the GDO coil. You should also see resonance at the
odd harmonic frequencies as well as the fundamental.
To measure the electrical length of a transmission line,
couple the GDO to a small wire loop; connect it to one end of Figure 6––N3SU grid-dipping her linear amplifier’s tank coil.
the line and leave the other end disconnected. (For best accu- Note that the high-voltage power supply is not only off but is
racy use the smallest loop that gives sufficient coupling.) The disconnected from the amplifier.
line is ¼ wavelength long at the lowest resonant frequency, so
the electrical length in meters is 75/f, where f is the resonant
Circuit,” Amateur Radio, Jan 1997, p 15. Amateur Radio is the journal
frequency in MHz. Again, you will also see resonance at the of The Wireless Institute of Australia. For North American readers
odd harmonics. who may not have ready access to this magazine, the article is posted
The voltage level at the RF output connector varies from at users.tpg.com.au/users/ldbutler/NegResDipMeter.htm.
5
H. Olson, W6GXN, “A New Look at Dip Meters,” Ham Radio, Aug
coil to coil but typically runs about 250 mV RMS into an open 1981, pp 25-28. This is a good survey of different dip meter circuit
circuit and 50 mV RMS into a 50 Ω load. That is sufficient to designs.
drive a typical frequency counter or serve as a test signal when 6
The difficulty of designing a wide-band choke is discussed in F. Lewis,
troubleshooting. W1LKV, “Anatomy of a Solid-State Dipper,” QST, Dec 1972, p 23.
7
This trick is one of several described in R. Johns, W3JIP, “Add-Ons
A GDO is like a TV remote control––you don’t know you For Greater Dipper Versatility,” QST, Feb 1981, p 37.
need it until you have it, and then you wonder how you ever 8
M. Bradley, K6TAF, “What Can You Do with a Dip Meter?” QST, May
got along without it. One simple instrument can do many of 2002, p 65. A good overview on how to use a dip meter.
the measurements of a bench full of expensive RF test equip- Al Bloom, N1AL, was first licensed while in high school, in 1968.
ment. I’ve only touched on a few of the many and varied ap- He received a Bachelor’s degree in physics, then a Master’s degree
plications for this versatile instrument.3,7,8 in electrical engineering, the latter while working as a W1AW op-
erator. While at the ARRL he co-authored both the License Manual
Notes and the Operating Manual. He designed amateur and commercial
1
C. F. (Bud) Bane, W6WB, “...About Grid-Dip Oscillators,” CQ, Mar equipment for the R. L. Drake Company and has spent the last 23
1947, p 13. This seems to be the first grid dip oscillator design in the years at Hewlett-Packard, designing RF and microwave test equip-
amateur literature. It was a portable battery-operated unit using a
type 3A5 vacuum tube. ment. Al has written several articles for QST, including two cover
2
The 2003 ARRL Handbook, Fig 6.46––Inductive and capacitive reac- plaque award winners. Mainly a CW operator, Al is on nearly all
tance vs frequency, p 6.27. of the amateur frequencies from 1.8 to 2450 MHz, including satel-
3
B. Clark, WB4OBZ, “The Art of Dipping,” QST, Jan 1974, p 16. lite operation. He is a life member of the ARRL and an ARRL Tech-
4
L. Butler, VK5BR, “A Dip Meter using the Lambda Negative Resistance nical Advisor. You can contact the author at [email protected].

Price: $19.95 + shipping. For more information, contact

NEW PRODUCTS Radio Warehouse, PO Box 77001, Charlotte, NC 28271-7000;
tel 704-321-2300; www.radio-warehouse.com.

QSL INDEX DIVIDERS FROM RADIO WAREHOUSE

š Attention award chasers: Radio Warehouse has introduced
a set of QSL index dividers to sort your cards in a file. Radio
Warehouse has also introduced a WAS set of dividers, pro-
duced in the same familiar format as the original DXCC edi-
tion (see photo). Both sets include Mylar reinforced tabs with
a record keeping grid on each card. The dividers are sized to
fit in any standard 5 × 8 index box, available at office supply
stores.
58 May 2003

bloo.pmd 58 3/26/2003, 9:58 AM