Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
This inv(n~ioll relates to al)paratus for, an(l a metho~
of, measurlng teml)eratllre rise in a.c. windings unde~ load
conditions using the method of electrical resistance change.
Apparatus for measuring temperature rise in a.c.
windings under load llse a method of electrical resistance
change. An exalnple is disclosed in U.S. Patent No. 2,578,455
(Seely) which discloses apparatus in the form of an electrical
circuit (Fig. 1) which measures the temperature rise in a.c.
windings under load conditions using the method of electrical
resistance change. A sample 10, for example a transformerj is
powered off the line 12 through the capacitor 14. The ohmmeter
circuit consists of the secondary winding 16 of the, potential
transformer 18, switch 20, interconnecting wires and the sample
10. An ohmmeter 24, equipped with its own d.c. source, is
isolated from the a.c. line by the capacitor 14. The a.c.
voltage existing across the sample 10 is compensated by
secondary voltage developed by the potential transformer 18 with
a 1:1 ratio. The residual a.c. voltage across the input 24a of
ohmmeter 24 is further filtered by capacitor 26. With the
switch 20 in "2" or "Tare" position the ohmmeter 24 will
indicate the resistance of the interconnecting leads and
secondary 16 of transformer 18. In this position the switch 20
is open and the transformer 18 is not energized, respectively.
In the "1" or "Measure" position of switch 20 both the
sample 10 and transformer 18 are energized. The necessary
7.~3
isolation o~ the d.c. si~llal generaLe(l by the ohmllleter 24 from
the a.c. line is accolnplisllc(l l)y the capacitor 14. This
arrangement prevents the ohmmeter 2~ from measurirlg other
parallel resistance associated with the input line. In the "1"
position the ohmmeter 2~ will indicate total resistance of the
sample 10, the secondary 16 of the potential transformer 18 and
the interconnecting leads. The resistance of the sample 10 is
then calculated by subtracting the resistance of interconnecting
leads and resistance of the secondary winding of the potential
transformer 18 from the total resistance obtained while the
switch ~0 was in the "1" position.
With this apparatus two separate measurements are
required in order to cancel the effect of the resistance of
interconnecting leads and the resistance of the potential
transformer. In practice, two readings are required to be taken
at a first temperature, conveniently room temperature, and a
reading is required to be taken at the other temperature T, that
is, when the windings are under load. The temperature rise is
calculated using equation (l)
TRISE = (RH/Rc ~ l) (~ + t) (l)
where: RC is the resistance of the windings at the beginning
of the test;
31'i'~
Rll is the resistance of the win(lillgs a~ the elld of
the test;
t is ambient rooln temperature, an(l
K is a constant equal to 234.5 for copper of 100%
conductivity
In addition to the disadvantages of requiring multiple
readings followed by subtractions in order to arrive at sample
resistance and also the necessity of measuring and noting the
sample resistance when cold, the apparatus as shown in Fig. l
needs auxiliary components to form a practical system. If a
permanent record is to be provided an additional shunt must be
connected in series with the sample. The voltage drop across
the shunt, directly proportional to the d.c. measuring current,
is then applied to the recorder. The existence of additional
resistances in the circuit causes degradation of the accuracy
and invalidates to some degree the fundamental assumption that
the measuring current depends on the value and changes of the
sample resistance only. This is especially important in
relation to the windings of large transformers or motors
exhibiting low resistances, say 1 or 25¦ , and becomes a factor
limiting practical use. The ohmmeter 24 in Fig. 1 includes its
own d.c. supply needed for the resistance measurement. This
direct current produces an undesirable heating effect of the
"sample" i.e. the a.c. winding in which the temperature rise
requires to be known. In order to minimize this influence, the
~irect currellt sho-ll<l be kel)t low; llowevcr tllis pro(luces a small
voltage drop across the sample and therefore degrades ~he
accuracy of the sample resistance rneasurements especially with
windings having low resistance values. Therefore, tile ohmmeter
sampling current is a cornpromise between two mutually
conflicting factors. This means that the direct current has to
be both adjustable and metered. The described above method uses
means of an absolute resistance measurement at the beginning of
the test. This approach contributes to additional errors which
are self-cancelling when a ratio measurement as per method
described below, is used. In addition, some tests take only up
to ZO seconds to complete and it is essential for a permanent
record of the test to be obtained in order to ascertain exactly
when the transformer is destroyed or when the thermal protection
mechanism, if any, is effective. Permanent recording is not
available with the SEELY apparatus and as such is considered to
be a further disadvantage associated with this device.
The above mentioned limitations and disadvantages
degrade the accuracy of the measurement and increase both the
operators time and the possibility of human error during testing
procedures.
An object of the present invention is to obviate or
mitigate the above mentioned disadvantages.
According to the invention there is provided apparatus
for measuring the change in the temperature of an a.c. winding
under loacl conditiolls coml)risillg, alternatillg power supply nleans
Eor energising the a.c. windillg, thc a.c. winding being
connected between a current supply line and a current return
line, measurement nleans connected in parallel with the power
supply means, said measurement means having constant direct
current supply means and feeding a constant current into one end
of the a.c. winding, voltage detection means for detecting the
voltage difference developed across the a.c. winding by the
constant direct current, signal processing and amplifying means
for processing the detected voltage difference and providing an
output voltage signal corresponding thereto, said output voltage
signal being a function of the resistance of the a.c. winding at
the winding temperature, the magnitude of the constant current
being adjustable, such that when the no-load output voltage
signal is set to a predetermined value at ambient temperature,
the output voltage under predetermined load indicates the ratio
of the resistance of the a.c. winding at the higher winding
temperature to the resistance of the a.c. winding at the ambient
temperature, said ratio being adapted to indicate the
temperature rise of the a.c. winding.
~ urther, according to the present invention there is
provided a method of measuring the temperature change of an a.c.
winding under load conditions using a resistance change in the
a.c. winding comprising:
supplying an a.c. electrical signal to the a.c. winding;
supplyirlg a d.c. Incasurelllerlt signal to thc winding;
the d.c. signal havillg a reference voltage and a
constant current;
detecting the voltage difference developed across the
a.c. winding caused by the flow of constant current therethrough;
processing the detected voltage difference to give an
output voltage signal, said output voltage signal being a
function of the resistance of the a.c. winding at the winding
temperature;
selecting the values of circuit components such that
resistances other than the resistance of the sample are
cancelled so that the output voltage is a direct function of the
electrical resistance of the a.c. winding only; and
calibrating the measuring apparatus at a no-load
condition by adjusting the constant current to set the output
voltage signal to a predetermined value for ambient temperature,
said output signal under load then varying in accordance with
the ratio of the electrical resistance of the a.c. winding at
the higher winding temperature to the electrical resistance of
the a.c. winding at the ambient temperature; and calculating the
temperature of the a.c. winding using this ratio.
An embodiment of the present invention will now be
described by way of example with reference to the accompanying
drawings in which:-
FIG. 1 is circuit diagram of a measuring apparatus
known in the prior art;
FIG. 2 is a complete circuit diagram of the power
supply circuitry all(l the measuring circuitry of the apparatus
according to the present invention;
FIG. 3 is a circuit diagram in more detail of a part of
the measuring circuit of the digram of Fig. Z;
FIG. 4 is a schematic diagram of a part of the
circuitry of Fig. 2;
FIG. S is a schematic diagram of the circuit of Fig. 4
for use in circuit analysis;
FIG. 6 is a detailed circuit diagram of part of the
circuits of Fig. 2 and Fig. 3.
Referring now to the drawings and in particular to Fig.
2, the circuit shown consists of two parts, a power supply
circuit 28 for energising the a.c. winding 30, and a measuring
circuit 29 for measuring the resistance of the winding and the
change of resistance of the winding with temperature. The a.c.
winding, or sample 30, according to the present invention is
energised by a 120V/120V isolating transformer 32 and auto
transformer 34. The resulting a.c. voltage from the setting of
the autotransformer 34 is applied to the sample 30 and is
monitored on a digital panel meter (DPM) 36 connected via an
a.c. voltage transducer 38. The current through the sample 30
is also monitored on a similar digital panel meter (DPM) 40.
Two ranges of current are available for passing through the
sample: 0 to 2 amps and 0 to 15 amps using two separate current
transd~lcers ~12 and ~tq respectively. These a.c. current
transducers 42, 4~ are conllected in series. When the switch 45
is positioned in the 0 to 15 amp range it activates a relay 48
the contacts 47 of which are then closed causing shunting of the
input of the 0 to 2 amp current transducer 42 thereby protecting
it against overload. All three transducers 38, 42 and 44
described above provide output signals in the form of d.c.
current from 0 to lmA. This feature makes scaling very easy by
means of a single resistor or a single trimpot.
The part of the circuit 29 shown in Fig. 2 relates to
measurements of a sample comprises ampliEiers 46, 48, 50 and 52,
a ratio display digital meter 54 a Darlington pair 56 and
associated circuit components as shown. The voltage across the
sample 30 is fed to the inverting and non-inverting inputs 58
and 60 respectively, of amplifier 48 which additionally
functions as a low pass filter. Amplifier 50 provides
adjustable gain. A built-in strip chart recorder 62 connected
to the output 53 of amplifier 52 provides a permanent record of
each test and since a chart speed is constant, the plot obtained
gives an exact presentation of the copper resistance change of
the winding during the test.
The temperature rise through the sample 30 may be
represented by equation (1) from which it can be seen that a
knowledge of the ratio of RH/~C is sufficient to enable
calculations to be performed without the equivalent of having to
-- 8
l7~
know tlle speci~ic values ol either ~ or Rc.
~ eferring now to Fig. 3, the amplifier 46 controls and
stahilizes actively the current Elow throug}l the sample 30 via
the Darlington pair 56. The potentiometer R8 establishes a
reference potential at the non-inverting input 60 of the
operational amplifier 46. The emitter 66 of the Darlington pair
56 is driven such that a point V2 is kept at this reference
voltage. Since V2 is kept constant, a constant current flows
through the sample 30 and through resistors Rl and R2 and
also the parallel combination of R3 and R4. Therefore the
current through the sample may be written as:
sample (d.c.) = Vl
2 (aR4R3/(R3 ~ aR4 ) ) (2)
where "a" is the fraction of R4 not shunted. The multiturn
potentiometer R7 is used to set t~le value of current through the
sample 30 in the range from a few mA up to 200 mA.
In order to facilitate understanding the operation of
the circuits shown in Fig. 2 and Fig. 3, analysis of the circuit
is described as follows with reference to Figs. 4 and 5:
The resistance of the sample 30 may be represen-ted by
the following equation:
RSAMPLE RWINDING RsUT RWIRING RsUT Rx (3)
e SUT Rsample under test
7~
Since RX may bc as large as Rsur it cannot be ignored.
Referring to Fig. 5 the following e~luations are presented.
Vl = (Rl + Ry)i (4)
and V2 = (RsuT + RX y
V2 2Vl (RSUT + RX + Ry + 2Rl - 2Rl -2Ry)i (6)
X Ry
then V2 - 2Vl = RSUTi
Referring to Fig. 6
Vl - V + Vl - VO = O (8)
R13 1/sC5
where s = ~ + j~r
/
C - ~ _
- 10 -
~ 3 ~ 3
and also tlle swllmatioll o~ the currellts at node 72 may be written
as follows:
Vl _ V2 + Vl - Vl + Vl + Vl vo O ( 9 )
Rll 1 R13 l/5c R12
where V is the voltage at node 72
Equation (8~ reduces to
Vl - V + (Vl - Vo) R13SC5 = o
Vl (Vl Vo) R13SC5 (10)
and equation (9) can be transformed as follows:
(Vl _ V )R R + V1-V )R R + VlsC R R R (Vl V )R R
Solving for Vl gives:
12 13 11 12 Rl1R13+ sC3RllR12R13) = VlRl2Rl34 VlRllRl2+ V R R
Vl V2R12R13 + VlRllR12 + Vo~llR13 (11)
R12R13' RllR12+ RllR13+ 5C3RllRl2 13
Now by substituting Vl in the last relation from (10) it can
be shown that
Vl + (Vl-Vo)Rl35C5 = V2R12R13 + VlRllR12 + VoRllR13 (12)
R12R13+ RllR12~ RllR13+ 5C3RllRl2Rl3
and solving the equation (12) for VO gives
1 + S2 - 12[V2 ~ Vl(sC3Rll + 1)] (13)
3 5 llR12R13+ Sc5(RllRl2+ RllR13+ R12R13) + R
Now assu~ing that Rll = R12, at 4r = O, then
-VO - -Vl + R12(V2 Vl) (14)
Rll
3~
or VO V2 2Vl (15)
and i~ equations (7) and (15) are equated tllen the following
relationsllip is yieldecl
RSuTi = -VO (16)
The sample current is set at the beginning of the test so
that a suitable voltage drop across the winding under test is
obtained. Under a no-load condition the ratiometer reading is
adjusted so that the signal on the ratiometer is set to l.OO at
the ambient temperature; then as RSuT changes with increasing
temperature, then VO represents the ratio RH/RC. In
addition, a ratio measurement is inherently more accurate than
absolute measurements, giving half of the errors possible from
absolute measures; . It should be noted that K and t are constant
for a given measurement.
Capacitor C5 and C3 provide additional filtering of
the input signal (Figs. 2, 3, 6). The amplifier 48 is
additionally an active low pass filter. The amplifier 50 provides
adjustable gain (by potentiometer Rl5) from 2 to l with R15
shorted. This additional gain is used to set the digital readout
to one on the ratio meter 54, at the beginning of the test. As
the sample progessively heats up during the test, the display on
the ratiometer 54 continues to display the ratio RH to Rc.
An important advantage of the present invention is that
the initial "TARE" measurement of the resistance of the leads at
~ 3
ambient temperatllre is elinlinated ancl any errors due to the
resistance of the Leads charlging during the test is also
eliminated. This provides time-saving simplicity o use and also
decreases the possibility of operator error. Furthermore the
direct current through the sample is regulated, controlled and
known therefore the influence of the heating effect is predictable
and may be accounted for if this value was considered
appreciable. In addition the direct display of the RH/RC
ratio simplifies calculations needed to be performed to compute
either the high tempera-ture or the temperature rise as given in
equation (1).
It should be appreciated that changes may be made to the
circuit details of the example shown without affecting the scope
of the invention. The reading of the output voltage on the
ratiometer need not be set exactly to unity at the no-load
condition although unity is the most convenient value; any reading
is sufficient, the reading under load conditions represents the
ratio RH/RC, and this value itself is a function of the output
voltage. A microprocessor circuit could be used for this circuit
which could incorporate algorithms to perform routine
calculations, and thereby indicate directly the temperature rise
in the a.c. winding in degrees Celsius. It is considered that a
man skilled in the art could adopt the inventive concepts
disclosed herein to be implemented in a microprocessor controlled
device, such adoption being routine and straightforward without
any inventive merit being required.
Tllus thcre is provide(l a novel circuit arrangement which
overco~nes the proble~ns yresent in the prior art of measuring the
temperature of energized a.c. windings present in the prior art.
