MOOG INC.

CONTROLS DIVISION, EAST AURORA, NY 14052

TRANSFER FUNCTIONS FOR MOOG SERVOVALVES

W. J. THAYER, DECEMBER 1958
Rev. JANUARY 1965

INTRODUCTION
It is often convenient in servoanalysis iables produce significant differences in with operating conditions such as supply
or in system synthesis work to represent the actual dynamic response. Consider pressure, input signal level, hydraulic
an electrohydraulic servovalve by a sim- the variables of the valve design. It is fluid temperature, ambient temperature,
plified, equivalent transfer function. Such well known that internal valve paramaters valve loading, and so forth. These effects
a representation is, at best, only an ap- (e.g., nozzle and orifice sizes, spring are insignificant for small variations
proximation of actual servovalve perform- rates, spool diameter, spool displace- about design values, but should be con-
ance. However, the usefulness of linear ment, etc.) may be adjusted to produce sidered where wide excursions are antici-
transfer functions for approximating serv- wide variations in dynamic response. An pated. It is important to appreciate and
ovalve response in analytical work is well analytic approach for relating servovalve control these and other operational vari-
established. dynamic response to internal valve para-
ables when performing measurements of
meters is given in Appendix I of this tech-
The difficulty in assuming an explicit servovalve dynamics. If such precautions
nical bulletin.
transfer function for electrohydraulic ser- are not taken, misleading and inaccurate
vovalves is that many design factors and Once a servovalve is built, the actual results may be obtained. Appendix II to
many operational and environmental var- dynamic response will vary somewhat this Bulletin describes the production
equipment presently used by Moog to
measure servovalve dynamic response.

Another difficulty in assigning simplified,

linear transfer functions to represent
servovalve response is that these valves are
highly complex devices that exhibit
high-order, nonlinear responses. If a first,
second, or even third-order transfer function
is selected to represent servovalve
dynamics, still only an approximation to
actual response is possible. Fortunately, for
most physical systems, the servovalve is not
the primary dynamic element, so it is only
necessary to represent valve response
throughout a relatively low frequency
spectrum. For instance, if a
servovalve-actuator is coupled to a load
which exhibits a 50 cps resonant frequency,
it is meaningful only to represent valve
dynamic response in the frequency range to
50 cps. Similarly, for lower response
physical systems, the contribution of valve
dynamics throughout a correspondingly
smaller frequency range need be
considered. This simplification of actual
servo response should be applied whenever
practicable, for the reduced analytical task
associated with the system analysis is
obvious.

These approximations to servovalve

response have resulted in such expressions
as "the equivalent time constant of the
servovalve is - seconds" or "the apparent
natural frequency of the servovalve is -
radians /second." If a representation of
servovalve response throughout the
frequency range to about 50 cps is sufficient,
then a first-order expression is usually
adequate. Figure I shows a typical valve
dynamic response, together with the
response of a first-order transfer function.
The first-order approximation is seen to be
quite good throughout the lower frequency
region. The time constant for the first-order
transfer function (i.e., the equivalent
servovalve time constant) is best established
by curve fitting techniques. If a quick
approximation is desired, the equivalent time
constant should correspond to the 45°
phase point rather than the 0.7
amplitude point (-3 db). In general, these FLOW CONTROL
points will not coincide as the higher- SERVOVALVES TORQUE MOTOR
order dynamic effects contribute low fre- This basic servovalve is one in which
quency phase lag in the servovalve re- the control flow at constant load is pro-
sponse, while not detracting appreciably portional to the electrical input current.
from the amplitude ratio. Flow from these servovalves will be in-
fluenced in varying degrees by changing VALVE
If servovalve response to frequencies load pressures, as indicated in Figure 4. SPOOL
near the 90 ° phase lag point is of in- For null stability considerations, only the
terest, then a second-order response region of this plot about the origin need
should be used. In a positional servo- be considered. Here, the influence of the
mechanism, a second-order representa- load on flow gain of the servovalve can
tion of the servovalve response is usually be considered negligible. In general, the
sufficient, as the actuator contributes an assumption of zero load influence is con-
TO ACTUATOR
additional 90 ° phase lag from the in- servative with respect to system stability
herent integration. Figure 2 shows a sec- analyses. FIGURE 3
ond-order approximation to the servo-
valve dynamics of Figure 1. Here, the
natural frequency is best associated with SYMBOLS
the 90 ° phase point, and the damping FREQUENTLY USED
ratio with the amplitude characteristic.
Other factors will often weigh more heav- i differential current m a
input to servovalve
ily in the choice of an approximate nat-
ural frequency and damping ratio. For Q servovalve flow in’/sec (cis)
to the load
example, it may be desirable to approxi-
mate the low frequency phase character- P servovalve differential Ibs/in’ (psi)
pressure output
istic accurately and, to do so, a second-
order transfer function which does not K servovalve sensitivity,
as defined
correlate with the 90 ° phase point may
be used. A good deal of judgment must, T time constants sec.

W” natural frequencies
therefore, be exercised to select the most rad/sec.
appropriate transfer function approxima-
r damping ratios nondimensional
tion.
S Laplace operator
FIGURE 4

SERVOVALVE
Another linearity assumption which is but within this range, flow gain may be
TRANSFER FUNCTIONS
often made is that servovalve flow gain from 50% to 200% of nominal.
Appropriate transfer functions for is constant through null. This is theoret-
standard Moog servovalves are given be- ically true for an ideal “zero lap” null The change in servovalve flow gain at
low. These expressions are linear, em- cut of the valve spool; however, the ac- null may sometimes cause system insta-
pirical relationships which approximate tual lap condition will vary with produc- bility; or, in other cases, poor positioning
the response of actual servovalves when tion tolerances. If the spool becomes accuracy, or poor dynamic response of
operating without saturation. The time overlapped, the servovalve flow gain is the actuator at low-amplitude input sig-
constants, n a t u r a l f r e q u e n c i e s , a n d reduced at null. Likewise, an underlap nals. This situation can be varied one
damping ratios cited are representative; produces higher-than-normal servovalve way or the other by holding a nominal
however, the response of individual serv- gain. Normal production tolerances main- overlap or underlap, as appropriate.
ovalve designs may vary quite widely tained at Moog hold the spool lap within
from those listed. Nevertheless, these ±O.OOOl inch for all four null edges. This The dynamic response of Moog flow
representations are very useful for ana- close control gives a very small range of control servovalves can be approximated
lytical studies and can reasonably form possible nonlinear flow control through in the frequency range to about 50 cps
the basis for detailed system design. null (about ±3% for an “axis” null cut); by the following first-order expression:

3
 The first and second-order transfer sensing the relationship of load pressure
+W=K (l ;,,> function approximations for servovalve to input current. A second-order transfer
dynamic response listed in the above function closely approximates the meas-
table give reasonably good correlation ured response in the frequency range to
where
with actual measured response. It is about 200 cps.
K -= servovalve static flow gain cis possible to relate servovalve response to
at zero load pressure drop m a internal valve parameters, as discussed
in Appendix I. However, the analytical + (s) = K,
7 = apparent servovalve time [1+(%ls+(%)j
approach to servovalve dynamics is most
constant set useful during preliminary servovalve de- where
sign, or when attempting to change the
response of a given design. It is better, K, = pressure control servo-
Standard flow control servovalves are
and more accurate, for system design to valve static gain psi/ma
available in several sizes and with many
use empirical approximations of the Wn = 2 n f, apparent natural
internal design configurations. The value
measured servovalve response.
of servovalve sensitivity K depends upon frequency rad/sec
the rated flow and input current. Typi- < = apparent damping ratio
cally, for a 5 gpm valve at a rated 8 ma nondimensional
input current, K = 2.4 cis/ma.
PRESSURE CONTROL
The appropriate time constant for rep- SERVOVALVES
resenting servovalve dynamics will de-
pend largely upon the flow capacity of
the valve. Typical time constant approxi- TORQUE MOTOR
mations for Moog Type 30 servovalves
are given in the table below.
P
If it is necessary to represent servo- VALVE SPOOL
PRESSURE
valve dynamics through a wider fre- ACK
quency range, a second-order response
can be used, as:

where FIGURE 5
(J, = 27r f, apparent TO ACTUATOR FIGURE 6
natural frequency rad/sec
The controlled differential pressure
These servovalves provide a differ-
may be any rated maximum up to the
5 = apparent damping ratio ential pressure output in response to the
system pressure. For a 1000 psi rated
electrical input current. The static flow-
control pressure at 8 ma electrical input,
pressure curves for a typical pressure
K, = 125 psi/ma.
control servovalve are shown ill Figure 6.
A small droop, or decrease in the con-
With a blocked load, the apparent na-
trolled pressure with flow, does occur,
tural frequency for pressure control
even throughout the null region. This
servovalves is approximately 250 cps,
droop is usually small in pressure-control
and the damping ratio is about 0.3 to
nd Order servovalves; however, in some applica-
0.5. The actual blocked-load response
fn tions even a small droop can significantly
t; for a pressure-control servovalve depends
cpr alter the system response. In pressure-
somewhat on the entrapped oil volume
flow servovalves, droop is purposely in-
30 .0013 240 .5 of the load, so the load volume should be
troduced. Transfer functions for these
31 .0015 200 .5 noted with response data.
valves are discussed in the next section.
32 .0020 160 .55

34 .0023 140 6 It is convenient to measure the dy- When a pressure control servovalve is
35 .0029 110 .65 namic response of a pressure control required to supply flow to the load, the
servovalve by capping the load lines and blocked-load transfer function no longer

4
adequately describes servovalve re- P(S) -[K, i -K, ( 1 + 7s) Q] . . . . . . ....,. For most pressure-flow servovalves,
sponse. Instead, the output pressure is the dynamic response of each flow re-
determined by the concurrent values of
input current and load flow. A linearized
.~f-.-[l+ (?I) ,: (g)j lationship (i.e., flow to current, and flow
to load pressure) can be approximated by
a critically damped, second-order trans-
approximation to the overall servovalve where
fer function. In addition, it has been
dynamic relationship can be established
K, = blocked-load pressure found experimentally that these dynamic
by superposition. Thus, the pressure re-
sensitivity psi/ma responses are nearly equal. The assump-
sponse to input current and the pressure tion of identical dynamics further sim-
response to load flow can be considered K, = zero-load pressure droop psi/cis
plifies the overall transfer function, so
separable and non-interacting relation- 7 = droop time that the dynamic performance expressed
ships. The transfer fuctions which result constant about 0.016 sec. mathematically becomes:
from this linearization are easily me- Wn = apparent servovalve
chanized on analog computers, and sys- natural frequency about 200 cps Q(s)=(KI~-KK~P)
tem design based on these assumptions
< = apparent servovalve where
has proven to be valid for most cases.
damping ratio about 0.5 K, = servovalve sensitivity to
The response of a pressure control input current cis/ma
For the analysis of most physical sys-
servovalve to load flow at constant input
tems, the second-order bracketed term K, = servovalve sensitivity to
current can be measured by techniques
on the right can be replaced by a suitable load pressure cis/psi
described in Appendix II. The character-
first-order lag, thus simplifying the ex-
istic response is approximated by the W, = equivalent servovalve
pression still further.
following transfer function: natural frequency;

PRESSURE-FLOW critically damped rad/sec

C O N T R O L (PQ)
The static flow-pressure characteristics
SERVOVALVES
for pressure-flow servovalves exhibit a
nearly linear relationship between flow,
current, and pressure as shown in Fig-
where TORQUE MOTOR ure 8. A wide variation in the sensitivi-
ties K, and K, is possible by appropri-
K, = static droop characteristic psi/cis
ate selection of internal valve parameters.
7 = equivalent droop time Normally, these constants are selected
constant set to suit the individual requirements of
specific servo systems. Typical values
might be K, = 3 . 0 cis/ma a n d K, =
0.02 cis/psi. The equivalent natural fre-
The droop constant K, represents the
quency for these pressure-flow control
slope of the flow-load curves of Figure 6.
servovalves is generally about 100 cps.
Values for K, generally fall in the range
20 to 50 psi/cis. Specially designed TO ACTUATOR FIGURE 7

spools can be utilized in pressure control

These servovalves combine the func-
valves to reduce the droop to almost
tions of pressure and flow control to pro-
zero. The equivalent droop time constant
vide characteristics which contribute
has a significant effect on the stability of
effective damping in highly-resonant
many servo systems. Typical values cor- loaded servo systems. Flow from these
respond to a corner frequency near 10 servovalves is determined not only by
CPS, or 0.016 second. The second-order the electrical input signal, but also by the
response of the droop transfer function is differential load pressure. For a linear
a high frequency effect, usually having a transfer function approximation to dy-
natural frequency near 200 cps. namic response, it may again be as-
sumed that principles of superposition
For system design, a simplified overall prevail. With this assumption, flow from
transfer function which includes both the servovalve may be considered sep-
blocked load and flow droop effects can arately dependent upon input current and
be used. This would be: load pressure.

5
 the amplitude of the load pressure
variations approach zero, the pressure
feedback time constant becomes shorter.
Theoretically this could lead to system limit
cycle oscillations; however, in practice this
seldom occurs due to finite frictions in the
actuator and load. A conservative design
approach is to size the DPF filter for
adequate pressure feedback at a
reasonably low amplitude of load pressure
variations; for instance, equal to 1/10
supply pressure. This allows a linear
The values of K1, K2, and Ƭ can be set analysis of system performance using the
throughout wide ranges by choice of relationships given previously, and gives
internal servovalve parameters. Most good agreement with actual system
practically, these values are selected for dynamic response.
each specific system application and will
be influenced by various system
requirements (e.g., system frequency or
transient response, static and quasi-static
system stiffness, system accuracies, etc.).
These servovalves function as Typical values of K1, and K2 would be
pressureflow control valves under dynamic similar to those mentioned for the
conditions, but act as pure flow control pressure-flow control servovalves. Values
servovalves in the steady state, or for Ƭ generally correspond to a corner
statically. As such, the benefits of pressure frequency of about 1/3 the load resonant
feedback are realized as damping for a frequency. For instance, with a 10 cps
resonant load, but statically the system resonant frequency load, Ƭ would be set to
retains the high resolution and stiffness approximately:
characteristics obtained with a flow control
servovalve.

The dynamic performance of DPF

servovalves is best represented for system
analysis by the following expression, which The equivalent natural frequency of DPF
presumes linearity by superposition: servovalves is generally about 100 cps.

The frequency sensitive filter

which operates on the feedback pressure
is created by hydraulic orifices and
capacitive elements. As such, the actual
filter response is nonlinear, reflecting the
square root relationship of hydraulic
where orifices. The result is an apparent change
in the filter time constant, Ƭ with the
K1 = servovalve sensitivity to amplitude of the load pressure. This
input current cis/ma These servovalves represent a further
nonlinearity gives a longer time constant
with increasing amplitude of the load extension of the pressure-feedback
K2 = magnitude of the dynamic pressure variation. A longer time constant technique for damping resonant-loaded
pressure feedback cis/psi allows more effective pressure feedback; servosystems. In PQ servovalves the
hence, more system damping. As
pressure feedback is proportional, so acts
for static, as well as dynamic, load forces.
Under static loading, the pressure
feedback produces a servoloop error which
must be offset by a corresponding posi- K, = proportional pressure and negative input currents produces a
tion signal. The effect is that of apparent feedback sensitivity cis/psi corresponding rate of change of flow
compliance of the servoactuator; that is, from the servovalve. In transfer function
KS = positive pressure
the actuator position will change propor- notation, this response is:
feedback sensitivity cis/psi
tional to load force. The DPF servovalve
r = time constant of
+ (s) = $-
where
washes-out the static pressure feedback,
low-pass filter set K = acceleration switching rnJ/seG
and essentially removes the position
error due to actuator compliance. w” = equivalent natural servovalve gain time unbalance
frequency of servovalve; it = time unbalance of the
In most resonant-loaded servosystems,
critically damped rad/sec current input nondimensional
the position feedback signal is derived
from the actuator. This means that the Note that if K2 = KS, then the SLEW This relationship is valid for system an-
physical compliance of the structure, servovalve becomes equivalent to the alysis throughout the frequency range to
which is contributing to the basic reson- D P F s e r v o v a l v e . H o w e v e r , i f K3 > K2, near the switching frequency, which is
then SLEW correction becomes effective. generally 150 cps or higher. It must be
ance problem, is outside the position
pointed out, however, that servo perform-
servoloop. This compliance causes errors Normally K2 is set to give the desired
ance for signal information which ap-
in position of the load under static load- load damping, and will have a value COT-
proaches the switching frequency is lim-
holding conditions, the same as caused responding to that used in either the
ited by the sampled-data nature of the
by actuator compliance. SLEW servo- PQ or DPF servovalves. The value of K3
system. Typical values for K would be
valves include an additional static pres- is generally about 1.5 K2 but should be from 100 to 1000 ins/secz.
sure feedback effect to correct for load set for the actual compliance of the
The s in the denominator of the trans-
position errors caused by structural posi- servosystem. The corner
frequency, f,,
fer function indicates that the servovalve
1
tion errors. of the low-pass filter (where r = 2?rf 1 acts as an integrator. This is true under
c
most conditions of operation. A non-
This is accomplished by combining is generally about one-tenth the load
linearity which occurs is apparent when
two pressure feedbacks within the SLEW resonant frequency. Actual values for
the servovalve is required to pass flow
servovalve; one, a conventional propor- the pressure-feedback parameters are to the load. Due to flow-reaction forces
tional pressure feedback which has nega- best established through analog com- acting on the valve spool, the true in-
tive polarity for damping of the load res- puter study of the complete actuation tegral effect created by the spool is up-
onance; and the other, a positive pres- system. set. This means that in the steady-state
sure feedback which causes the actuator a finite electrical signal is required to
sustain valve flow. A constant input
ACCELERATION
to extend under static loading to correct
velocity to an acceleration switching
for position errors from both actuator SWITCHING (AS)
servo will, therefore, create a small, but
and load compliant effects. The positive SERVOVALVES
finite, steady error.
pressure feedback path includes a low-
pass filter, so that it is effective only With constant-amplitude sinusoidal in-
under quasi-static loading conditions. TORQUE MOTOR puts to a switching amplifier and AS
servovalve supplying flow output, this
The following transfer function can be nonlinearity appears as a break in the
used to represent the SLEW servovalve: slope of the amplitude ratio of the servo-
valve response. Below the break, the
SPOOL response exhibits a zero-slope or propor-
tional amplitude ratio. The frequency of
this break point is dependent upon in-
ternal valve parameters and the ampli-
FIGURE 11
t u d e o f t h e s i n u s o i d a l i n p u t . I n well-
In these servovalves, the conventional made servovalves, the break frequency is
proportional input current to the first generally below 1 cps; so, for system
where
stage is replaced by an alternating switch- stability and dynamic response consider-
K, = flow sensitivity to ing action. Control of the relative on ations, this effect is unimportant and
input current cis/ma and off time durations of the positive can be ignored.

7
APPENDIX I
ANALYTIC inary design of a new valve 1. An ideal current source (infinite
ANALYSIS OF configuration, or when attempting to impedance) is used.
SERVOVALVE alter response of a given design by 2. Negligible load pressure exists.
DYNAMICS parameter variation. Analysis also 3. All nonlinearities can either be
contributes to a clearer understanding approximated by linear dynamic
It is possible to derive meaningful of servovalve operation. effects, or can be neglected.
transfer functions for electrohydraulic 4. The armature/flapper can be rep-
servovalves, and several papers have Rather elaborate analyses of resented as a simple lumped-
reported such work (ref). Unfortunately, servovalve dynamic response have parameter system.
servovalves are complex devices and been performed at Moog, including 5. Perturbation conditions can be
have many nonlinear characteristics computer studies which involve several applied to the hydraulic amplifier
which are significant in their operation. nonlinear effects, and up to eight orifice characteristics.
These nonlinearities include: electrical dynamic orders (excluding any load 6. Fluid compressibility and viscosity
hysteresis of the torque motor, change dynamics). Unfortunately, these effects are negligible.
in torque-motor output with complex analyses have not contributed 7. Motions of the flapper are small
displacement, change in orifice fluid- significantly to servovalve design due to with respect to spool motion.
impedance with flow and with fluid uncertainties and inaccuracies 8. The forces necessary to move the
characteristics, change in orifice associated with the higher-order effects. spool are small with respect to the
discharge coefficient with pressure driving force available.
ratio, sliding friction of the spool, and These analyses have been extremely
others. useful when reduced to their simpler The last assumption implies that the
form. A very adequate transfer function differential pressure across the spool is
Many servovalve parts are small so representation for the basic Type 30
have a shape which is analytically non- mechanical feedback servovalve is
ideal. For example, fixed inlet orifices given in Figure 12. This simplified
are often 0.006 to 0.008 inch in representation results from the following
diameter. Ideally, the length of the assumptions:
orifice would be small with respect to
its diameter to avoid both laminar and
sharp-edge orifice effects; however,
this becomes physically impractical
with small orifices due to lack of
strength for differential pressure
loading, and lack of material for
adequate life with fluid erosion.
Therefore, the practical design from the TORQUE
SUMMATION
performance standpoint is not
necessarily the ideal design from the i TORQUE
ARMATURE-FLAPPER
1
Kr
xf HYDRAULIC ∆Q SPOOL
xs SPOOL FLOW Qi
analytical standpoint. MOTOR
KI
- 2
æ 2ζ ö æ s ö
1+çç ÷÷s +çç
AMPLIFIER
K2
1 GAIN
K3
èωn èωn Ass
Experience has shown that these non-
linear and non-ideal characteristics
FEEDBACK WIRE
limit the usefulness of theoretical Kw
analysis of servovalve dynamics in
systems design. Instead, the more
meaningful approach is to approximate
measured servovalve response with
SIMPLIFIED SERVOVALVE BLOCK DIAGRAM
suitable transfer functions, as
discussed in the body of this technical
bulletin. FIGURE 12

The analytic representation of servo-

valve dynamics is useful during prelim-
negligible during dynamic conditions. If so, of the torque motor, and is individually
then spool mass, friction, flow forces, and adjusted in each servovalve to meet
other spool force effects can be neglected. At prescribed dynamic response limits. The
first this assumption may seem unreasonable; damping force on the armature/flapper is
but it can be shown to be quite valid, and the likewise a composite effect. Here, it is known
simplification which results more than justifies from experience that the equivalent ζ is about Any of the loop gain parameters can be
its use. 0.4. altered to change servovalve response. For
example, the following changes would
The simplified block diagram is a third order The hydraulic-amplifier orifice bridge reduces increase internal servovalve loop gain: (1)
system consisting of the armature/ flapper to a simple gain term with the assumptions smaller spool diameter, (2) larger nozzle
mass, damping and stiffness, together with listed earlier. This gain is the differential flow diameter, (3) higher nozzle pressure drop, (4)
the flow-integration effect of the spool. The unbalance between opposite arms of the higher torque motor charge level. The higher
spool, in this case, is analogous to the piston bridge, per increment of flapper motion. torque motor charge gives a lower kf which
of a simple position servoloop. increases loop gain, but this also lowers the
Internal loop gain of the servovalve is natural frequency of the first stage.
The rotational mass of the armature/ flapper is determined by the following parameters. Unfortunately, the directions of these two
quite easy to calculate. The effective stiffness effects are not compatible in that higher loop
of the armature/flapper is a composite of gain cannot be used with a lower natural
several effects, the most important of which frequency first stage. Therefore, an optimum
are the centering effect of the flexure tube, charge level exists which produces maximum
and the decentering effect of the permanent The hydraulic amplifier flow gain, K2, can be loop gain for the stability margin desired.
magnet flux. The latter is set by charge level related to nozzle parameters by the following:
 I

APPENDIX I I e.g., flow to the load or pressure in the load” is easily obtained with low compli-
load lines. Measured servovalve output ance, capped load lines.
MEASUREMENT IS dependent upon the nature of the
In servovalves which utilize load pres-
OF SERVOVALVE load, so it is important for consistent and
sure feedback (i.e., pressure control, pres-
DYNAMIC useful response informatron to maintain
RESPONSE sure flow, dynamic pressure feedback and
precisely known loading conditions. The
SLEW servovalves), the output is deter-
Moog has developed hrghly specialized philosophy of dynamic testing at Moog
mined not only by the coil current, but
test equipment for production measure- is to attempt to maintain an ideal load.
also by the action of the pressure feed-
ment of servovalve dynamic response. In this way, the servovalve response is back. The principle of superposition may
This appendix describes this equipment most completely isolated with the valve be employed to individually measure the
and expiains the design philosophy which operating as a single component. Prac- servovalve response to current with no
insures accurate and consistent dynamic tically, this philosophy is well suited to load, and the response to load variations
response measurements. servovalve dynamics, for the measured with zero or constant current. Tech-
Servovalve dynamic response is de- information is applicable to all systems niques for the latter test represent an
fined as the relationship of output to in- and is consistent if measured at different extension of those described here in that
put with all other operational variables times and with different pieces of test the appropriate response is measured
held constant. This relationship is con- apparatus. while a forced hydraulic load is applied
veniently descrrbed in terms of the ampli- to the servovalve under test.
The ideal load for a flow control servo-
tude ratio and phase angle of the output
valve would be massless and frictionless,
in response to a sinusoidal input of vary-
ing frequency. The input to Moog servo-
presenting absolutely no obstruction to DYNAMIC TEST
flow from the servovalve. In practice, EQUIPMENT
valves is usually considered to be the
this “ n o l o a d ” operation can be ap-
drfferential current between the two A simplified schematic for production
proached with sufficient purity to assure
motor coils. The two COIIS may be con- dynamic test equipment is presented in
no detectable influence due to loading.
nected in series aiding; or the valve may Figure 13. Photographs of a control con-
The mechanization of this loading condi-
be supplied with a single two-lead coil, in sole and test actuator appear in Figures
bon is described In detail later.
which case the input becomes the abso- 14 and 15, respectively.
lute current in the coil. For a pressure control servovalve, the The servovalve under test is driven
Servovalve output is normally con- ideal load would have zero flow and zero with sinusoidal input current from the
sidered the primary controlled variable; compliance. This so called “blocked electronic oscillator. The amplifier cir-

FIGURE 13

10
cuitry which supplies the input current The position transducer signal is utilized In each case, the amplitude of the
to the servovalve operates with a high for continuous centerrng of the piston. summed signal is related to the phase
degree of current feedback. Operation This position signal is supplied to the angle between the signals, regardless of
with current feedback is essential for servovalve by a low pass, low gain, nega- the signal frequency.
precise testing for several reasons, in- tive feedback loop so the average ac-
Circuitry for performing the phase
cluding: tuator position is maintained near center.
measuring function utilizing these re
1. The dynamic effect of servovalve Electrical signals from the piston vel- lationships is included in the test equip-
coil inductance, which would other- ocity transducer, which indicate output ment. Following the amplitude ratio
wise upset the proportional rela- flow, are passed through an isolating measurement, the operator depresses the
tionship of coil current to com- amplifier, then to a logarithmic poten- appropriate phase selector button. The
mand input voltage, is virtually tiometer. The potentiometer dial is cali- amplitude of the oscilloscope display is
eliminated. brated directly in decibels (db). Using then adjusted to the reference amplitude
2. Purity of the sinusoidal input cur- the oscilloscope as an error detector, the by rotating the phase dial. Rotation of
rent is obtained throughout the operator is able to compare the electrical this dial adjusts the gain of the sum-
response range (to several hun- output signal at any test frequency with difference amplifier. The dial is cali-
dred cps). a reference signal amplitude which is brated directly in degrees of phase angle
proportional to the input current. The which can be recorded by the operator.
3. Calibrations of input current per
db potentiometer can then be adjusted
unit input voltage are unaffected
for zero error. The amplitude ratio of
by changing from valve-to-valve,
servovalve output to input current is di-
regardless of torque motor coil
rectly indicated by the calibration of the
resistance.
db dial.
4. The long term amplifier stability
and accuracy associated with high The phase angle of servovalve re-
feedback are attained. sponse is determined by appropriate
summation of the output flow and the
input current signals. With sinusoidal
As indicated in Figure 13, the torque
current and flow signals, the amplitude
motor coils are connected in series aiding
of the summation signal is related to
in this test equipment. Servovalve op-
phase angle. This is seen by the follow-
eration in this manner is identical to
ing:
operation with push-pull differential cur.
rent, with or without quiescent coil cur-
rent. The series coil connection is used
Input current signal et =Asin wt
only for simplicity of the electrical cir-
cuitry. Output flow signal ez=Bsin( wt+e)

For dynamic flow tests, servovalve out-

put is plumbed to a test actuator. The where A and B are constants
actuator has a lap-fit, extremely low 8 is the relative phase angle
mass piston which has no end seals
If these signals are subtracted;
or rod-end bearing surfaces. Instead,
small rods extend from each face of the
piston to carry, respectively, the core el - e2= PAsin ($)][cos(#-$)]
FIGURE 14
for a variable reluctance position trans-
ducer and the core for a translational
motion velocity generator. Both trans- (utilized for phase angles
ducers are immersed in oil so that no from 0 to 900)
external piston seals are required.

The velocity transducer generates a and by adding the signals:

voltage which is proportional at each in-

stant to the flow from the servovalve. el 1e2= pAcos($][sinkt-$$]
This indication of output flow is not lim-
ited by resolution of the electrical sensor
nor is the signal reduced in amplitude (utilized for phase angles
for small excursions of the test actuator. from 90 to 270”)

11