0% found this document useful (0 votes)
2 views8 pages

Galvanometer Amplifiers for Oscilloscopes

This document discusses the design of galvanometer amplifiers with photo-converters for use with oscilloscopes. It summarizes previous work on such amplifiers and their limitations. The author then describes their own design of a galvanometer amplifier that achieves a bandwidth of 150-250 Hz while maintaining accuracy, using feedback and correction circuits. Key aspects of the input, output, feedback and correction circuits are shown in diagrams and equations. Experimental results validated the calculated dynamic characteristics of the improved amplifier design.

Uploaded by

Okta Dwiputra
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
2 views8 pages

Galvanometer Amplifiers for Oscilloscopes

This document discusses the design of galvanometer amplifiers with photo-converters for use with oscilloscopes. It summarizes previous work on such amplifiers and their limitations. The author then describes their own design of a galvanometer amplifier that achieves a bandwidth of 150-250 Hz while maintaining accuracy, using feedback and correction circuits. Key aspects of the input, output, feedback and correction circuits are shown in diagrams and equations. Experimental results validated the calculated dynamic characteristics of the improved amplifier design.

Uploaded by

Okta Dwiputra
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

E LECTRICAL ME ASUREMENTS

GALVANOMETER-TYPE AMPLIFIERS WITH PHOTO-CONVERTERS


FOR O S C I L L O S C O P E S

R. R. K h a r c h e n k o

Galvanometer amplifiers (GA) with photo-converters are widely used in electrical measurements; they have
passed an extensive period of development and have been dealt with in many specialized papers [1-5].
In the above papers, however, the problem of calculating the GA dynamic characteristics has not been
sufficiently developed. For instance, in the descriptions of the so-called photo-electro-optical amplifiers L~TI
(developed by B. P. Kozyrev), which, owing to the absence of a negative feedback possess a large sensitivity [1],
no data is given on their static or dynamic accuracy. In the description of the so-called photo-compensated
amplifiers of the Leningrad plant "Vibrator" (developed by B. A. Seliber and S. G. Rabinovich), which use
negative feedback and thus have greater accuracy with a smaller sensitivity [2], information on the input im-
pedance and the dynamic characteristics in various measuring ranges are lacking.
Above amplifiers are intended for use in circuits with slowly changing characteristics; hence, they have a
very limited bandwidth and are designed to be connected to normal low-frequency indicating instruments or
recorders.
Measurements in circuits with rapidly changing conditions are made by means of oscilloscopes. If the
sensitivity of the oscilloscope is insufficient, or a matching of the impedance of the measured object to the vibrator
isrequired, a dc electronic amplifier (DCA) isused with the oscilloscope. It is known that the increased sensitivity
by means of the IgCA leads to a rise in the error of the measuring channel due to drift.
In this connection the design of galvanometer amplifiers with a large bandwidth for a given accuracy
is of interest. On this subject two independent investigations were conducted in the USSR. The first one was
carried out by A. A. Nemura [3] who tested several GA models designed to work with vibrator I of oscilloscope
MPO-2. The models used vibrator VIII of the MPO-2 type as a galvanometer (in one of the versions) and vibrator
type GB-IV of a geodesic oscilloscope (in another version).
The main purpose of this analysis is the investigation of various correction circuits, which were designed
for obtaining the required amplitude-frequency characteristics with an optimum transfer process. The analysis
did not deal (it was probably outside the scope of investigations) with the effect of the resistance connected to
the input of the GA on the amount of feedback, which determines the static accuracy and the adjustment of the
correction. It should be noted that in these models the bandwidth exceeded 1000 cps, thus forcing the author to
consider in his theoretical analysis small time constants of all the components, and deal with a system described
by equations of the fourth and fifth degree.
The second investigation carried out by the present author dealt with amplifiers for vibrator VIII of MPO-2
type oscilloscopes. The low-frequency galvanometer with torsion suspensions of the "Vibrator" plant was used
in this instance as the instrument galvanometer. By means of feedback and corrections the working bandwidth
of the GA was raised to 150-250 cps with an optimum transfer process, It was shown experin-e ntally that in such
a bandwidth the small time constants of all the components need not be considered,and the entire system can be
treated as a system of the second order. As the result of these investigations the author developed a simple and
convenient method for calculating all the basic GA characteristics, and good agreement between the calculated
and experimental data was obtained. Below we give the basic techniques of this method.

117
=0,08~a Aa--'i~O56tlxn AU=0.42v Vibrator
11=35~ a n=o o ~ $ ,0 I~=lma ~ r e galvanomote~7---- ~ tvtm VIII
u:o rTXL2 ,nl.

Fig- 1. 1) Input circuit; 2) galvanometer; 3) photo- Fig. 2.


converter; 4) correcting amplifier; 5) feedback circuit.

_7 #, & t"

S,applV

LI'I~
-------ody = First circuit Second circuit

Fig. 3. 1) Photovaristor, 2) illu- Fig. 4.


minated area.

A block schematic of the galvanometer amplifier with a photo-converter and correcting unit with feedback
is shown in Fig. 1.
The gain of the system without feedback is given by the expression

K= -~=,Sg Kp Samp (I)

where n, Sg, Kp, and Sa, shown in Fig. 1, are gains (transfer factors) of various units.
The gain of the system with feedback is

K
Kfb - I + Ky' (2)

where .y = I'x/Iz is the feedback factor.


The relative variations (instability) of Kfb are determined from the formula

AKfb - _ ~ gK Ky av_
Km I + Ky K I + KY y

In order to obtain a highly stable amplifier two conditions must be fulfilled:


1) stability of parameter 7 must be ensured; this condition requires a special input circuit arrangement
(see below).
2) an ensured relation of parameters 1 ~ KT;this condition is easily attained in galvanometer amplifiers.
Let us examine the characteristics of separate units in the models we tested.

i18
Output and feedback circuits. The output and
N'r9.
.P feedback circuitsare interconnected; that is why they
must be considered together (Fig. 2).

The technica! requirements of the GA under


consideration are the following:

1) its output variable consists of current Iz (the


GA is connected to vibrator VIII of the MPO-2 oscillo-
scope type); hence, a current feedback is used.

2) the output impedance of the measured object


connected to the GA input may vary between 0 and
m ohms; the variations of the feedback factor y caused
Vibrator type VIII by this difference in the impedance must not exceed
Fig. 5. 1%. It is obvious that if this condition holds all the
dynamic characteristics of the system become fixed
and almost independent of the value of the input
impedance.
t,
An input circuit which provides these conditions
is shown in Fig. 2.
y___,
1.0!
. . . . . . ----"~'~- ---'-" " ~ .....
The relations between the parameters are the
f rsrg _
a8 , I following: R >> rg, r s, rib; r e = 100rs+rg ; the
,' ..... -Operating frequency rang e - ~ cps
a,l I I i I I I ,
0 ~o 8o IZO fSo f cps value of r s is calculated to obtain the required degree
Fig. 6.
of damping in the galvanometer 3 = Rk w h e r e Rk
rg+rs
is the critical impedance. The value of 8 is chosen
in turn so as to ensure an optimum transient response
- rdc lure see-- (see below). The nominal input voltage of the GA
i n this circuit is determined by a simple expression
U1 ~- reIv The input circuit transfer constant is
..~m sec

A A A A A A _(g_ rs
II II II II II II n= -- -.
I1
Fig. 7. 11 is the recording made with rs +rg
a type IV vibrator; 12 is a recording and the transfer constant of the feedback circuit is
made with a type VIII vibrator.
v
I1 rib
Y:- 12 R

Photo-converter, A differential photovaristor (PV) type PbS with dimensions a = 2 m m and b = 8 m m was
used as a photo-converter and connected in a bridge circuit (Fig. 3)o The bandwidth of the photovaristor (PV)
exceeds 1000 cps, according to special tests, and hence in calculating the dynamic characteristics of the GA~the
photo-converter can be considered as a unit providing no lag.

The relative change in the resistance of each half of the PV when the illuminated area is displaced by ZXa
amounts to

k--d -- 1
X= AR _ kl .Aa '
R kd
-tel
-al + a7 (3)

119
where kd and k! are the dark and light resistivity of the PV. The variation of the output voltage in the bridge
circuit corresponding to k (Fig. 3) is

AUfUp 2~
I+~ ' (4)

where Up is the voltage across each half of the PV for


r
AR= 0;
R'

the maximum value of Up is determined by the heating of the PV.


From (3) and (4) the sensitivity of the photovaristor can be obtained, Kp = AU/Aa; in our models the value
of Kp was set at (5 to 8) v/mm.
Co~eetin 8 amplifier. Methods of correcting linear converters have been developed in detail in the theo~/
of automatic control where they are widely used, The problem of correcting the dynamic characteristics of a
moving coil instrument has been dealt with in [6].
In the present work they have been applied to the circuit in Fig. 1,
Let the correcting amplifier of this circuit be replaced by an aperiodic de amplifier.
The transfer constant of such a system without feedback will be, according to (1),

K(p)= K
x§ 2~p+% og p (s)

where 8 is the degree of damping; w0 - the natural angular frequency of the galvanometer.

The transfer function of the system with feedback will b%aecording to (2),

Kqp) .

K I
I+N
V 1+ ~ r p + [[ 1 p~
!+ '

ooVl+ v
I
-- Kfb I -t- -'TP~- ~ p 2 (6)
~ (~o)

where N = KT.
W e have thus obtained a second-order equation with equivalent parameters:

for the gain


K
/~tb- I+N'
("0

120
natural frequency

-'o = OoVi-+ N (8)

and degree of damping

VI+N (9)

It will be seen that a fixed feedback increases the natural frequency of the circuit by a factor of d'l + N;
this amounts to a correctio~but it is achieved with a loss of 1 + N. Such a correction is not advantageous, and
it is not the aim but the consequence of the feedback, which is intended for stabilizing the gain of the system.

In order to obtain an optimum transient response for this system, it is necessary to fulfill the following
conditions:15' = 0,6 to 0.8; for this purpose according to (9) the galvanometer damping should be 15 = 15"4-1 + N =
= (0.6 to 0.8) ~ + N. If this cannot be achieved or is inexpedient, the fixed feedback should be supplemented
by a selective negative feedback.

Let us now examine the special correction (widening of the working bandwidth) which is not accompanied
by loss of sensitivity.
In the circuit of Fig, 1 let us construct the correcting amplifier in such a manner that its transfer function
is expressed by equation

l+Tp (lo)
Samp (P) =Sampl+ mTp J

where m < 1.

It follows from (10) that the gain of the amplifier rises with frequency from its initial value of S to a
amp
certain maximum value S a m z l / m . The transfer function of the system without feedback will now be
t"

1 l+~p
K~p)=K
1 + 2~p~ I lq-m'rl~ (11)
o~o r p~

If 15 > 1 it is possible to represent the denominator of the second-order polynomial with respect to p as the
product of two factors:

215 1
1 -]- ~ . p - l - y K p - = ( 1 - k a p ) ( l + b p ) , (12)
tuo tu o
where

--a-.I-b: ~ = ab.
(% (002 (12a)

Let the parameters of the circuit be chosen in such a manner that the equality b = r holds, where b is the
smaller root of (12).

Then we can c a n c e l similar factors in the numerator and denominator of (11) and obtain

1 I
K (p)=K . . . . . . . . . =
1+ap 1-4-mTp
=K
1
(13)
213~
1 q- -(~- P + ~( ~ , 1) ~ p-"

121
where

21~K 1
(.)o)-----~--=a-kmx; (~o)--~=am'~. (13a)

Combining (12a) and (13a) we have

(~176 ~ - V mT----. (14)


(0 o

Thus, as the result of the above correction we have again obtained a second-order system (13) with equivalent
parameters (w0)k and Bk (subscript k denotes the correction system); the natural frequency has now risen by
without any loss in sensitivity.
When a negative feedback is applied to this circuit we again obtain a second-order system (6), (7), (8),
and (9) with parameters:

K
(K'~)fb - I--FN ' (1,5)

(~ = ~ V T ~ --]/~Lk N, (16)

K~-- t~,~
~i+~ (17)

It will be seen that the combined effect of the special correction and negative feedback raises the natural
frequency by a factor of 1/4~m. ~ with a loss of gain of only 1 + N. Thus, the special correction provides
a gain in sensitivity of 1/m.
From combining (12a) and (13a)it is possible to find the design formula which relates Bk and B:

1~,~= ~,(l-l-m) + ]/"13u--I(l--m) (18)

In designing a system it is necessary to take t3' k = 0.6 to 0.8;next,from the value of N determine the value
of Bk from formula (17) and from the latter and formula (~8) calculate •.
Such a system will possess a mfficient reserve of stability even if some of the small time constants are
ignored.
Above theory shows [Link] correcting galvanometer amplifierbit is important to find a correcting circuit
with a transfer function (10).
We have found and investigated several such circuits, two of which are shown in Fig.4.
For the first circuit
Samp_~ ~Ra

x = C~/?~ ;
Rt+Ra
m--RI+Ra..t-R~(IX-F1) ;
for the second circuit with condition rz ~ Ra

~tRam , v-----Crl; m - - - - -F~


-.
Samff- R a + R i r, +r~

On the basis of above considerations the author designed, constructed,and tested out a model of a galvanometer
amplifier.
The full schematic of the model is given in Fig. 5. The gains (transfer constants) of all the stages and the
values of the signal along the channel for a nominal value of the input and output signals are also given in the
circuit of Fig. 1.
Below we give the technical data of the model.
The galvanometer has the following characteristics, the coil resistance is r_ = 22 ohms; the critical
impedance Rk = 342 ohms; sensitivity in the assembly (the distance between the mirror and the surface of the
photovaristors l = 52 ram) Sg = 0.7" 10 s ram/amp; natural frequency f0 = 1.7 cps" The degree of damping of the
galvanometer /5 = Rk = 2.
~g+r s

The feedback factor of the system N = K7 ~ 400. The parameters of the correcting system are: r = 27.9 msec;
m = 0.02. The value of/3 k calculated from (18) is 13.2; the value of B' k (the equivalent degree of damping of the
system) calculated from (17) is 0.66.
Tests of the GA model showed that its static amplitude characteristic is linear over the whole range. Any
variations in the slope of this characteristic (after prolonged operation) could not be observed on grade 0,5 instru-
ments connected to the input and output of the GA. Neither was there any zero drift observable with a normal
supply voltage. With supply voltage variations of :E 10% from the normal, the zero drift did not exceed 0.25~ of
the full scale reading. The amplitude-frequency characteristic of the GA model at a nominal supply voltage
(127 v) is given in Fig. 6, which shows that the operating frequency range amounts to some 150 cps. Fig. 7 shows
an oscillogram of a transient process when the input current was abruptly switched on; the decay time of the system
amounts to 10 msec.
Careful testing of the model showed two peculiar defects in above system: a large sensitivity to shocks
and vibrations; and a noticeable effect of ambient humidity.
The first defect is probably due to unsatisfactory balancing of the galvanometer moving coil (additional
experiments should be carried out in this connection~ and vibration-proofing of the body of the galvanometer-
amplifier should be developed,
The second defect is due to the physical properties of the lead sulfite photovaristors used. In order to
eliminate this defect it is necessary to protect the varistors with a moistureproof transparent film of lacquer.
On the basis of the work carried out it was possible to suggest that the instrument-making industry should
manufacture a series of multichannel miniature galvanometer-amplifiers for oscilloscopes. In these GA all the
components should be miniaturized;thus, the galvanometer should be of the GB type (units of the geophysical
series), the photovaristors of a special make (mounted on a single plate), and the correcting amplifier transistorized.
For the p0xpose of studying the possibility of the last suggestion, the electrical measurements department
of the Moscow Eleetrotechnical Institute made a special study, with the participation of the author, of a transistorized
de amplifier (single channel) and the replacement by it of the correcting tube amplifier in the circuit shown in
Fig. 1. The results of this study are reported in [7].
Engineer S. G. Golub' participated in the experimental part of the work,

LITERATURE CITED
[1] B. P. Kozyrev, Uspekhi Pizicheskikh Nauk 4__44, 2 (1951).

123
[2] B. A. Seliber and S. G. Rabinovich, Avtomatika i Telemekhanika 8 (1956).*

[3] A. A. Nemttta, Author's Report of Dissertation [in Russian] (Kuibyshev Industrial Institute, 1956).
[4] Toulin, J. ScL Instr. No. 10 (1955).
[5] Branderburger, ATM No. 228 (1955).

[6] R. tL Kharchenko, Priborostroenie No. 2 (1956).

[7] I. N. Gratsianskii, Izmeritel'naya Tekhnika No. 2 (1960).*

MINIATURE TRANSISTORIZED GALVANOMETRIC AMPLIFIER

I. N. G r a t s i a n s k i i

Galvanometric amplifiers are beginning to be used in measurements as attachments for raising sensitivity;
for instance, with vibration oscilloscopes.
Galvanometric amplifiers (GA),with their relatively high input resistance, can be made, for equal sensitivity,
more accurate than other types of amplifiers.
A number of papers deal with the general questions of their theory and describe various types of such
amplifiers [1-3].
R. R. Kharchenko [4] examined the relation of the dynamic characteristics of a definite structural system
with its parameters, which determine the specification data of a GA. He also describes the results of testing
a GA with a correcting amplifier which used thermionic tubes.
When a large number of parameters which vary with time are measured simultaneously, multichannel
converters and instrument amplifiers are used.
In the present article a GA is described which employs miniature componentsj namely, a miniature
galvanometer GB* * (from the geophysical series of units) developed by E. S. Borisevtch [5], lead sulfite photo-
varistors of a special construction (V. I. Lenin All-Union Electrotechnical Institute) and a transistorized dc
amplifier (Fig. 1).
The basic unit of the GA is the moving-coil galvanometer G (Fig. 2). A beam of light is directed by the
galvanometer mirror from the source of light (L = incandescent lam~a) to the differential photovaristor PV which
is connected to the input of a balanced amplifier.
In the middle position of the beam, which corrresponds to 11 = 0,both halves of the photovaristor are equally
illuminated and have the same resistance. The output current 12 = 0.
In order to stabilize the GA and extend the operating frequency range, special corrections are introduced
(RkCk) with a strong negative feedback [4].
The two-stage transistor de amplifier (T1T' 1, T2T'2) type OE-OK is based on a push-pull parallel balanced
circuit.

* See English translation.


* * These galvanometers,in addition to other qualities,are less sensitive to shocks and vibrations (as compared

with normal types) since,owing to their narrow, long coil, they are better balanced. The problem of vibration-
resistance in galvanometric amplifiers as a whoie has not been investigated and requires further study.

124

You might also like