24
\$\begingroup\$

Recently, I've been working to repair a broken 1980 "Arp Solus" synthesizer that I acquired from a friend.

I have, however, found myself confronted by an unfamiliar arrangement: some kind of resistor fixed (glued? melted?) firmly to the top of a nearby IC.

Resistor stuck on top transistor array

The IC beneath is a CA8036 General Purpose Transistor Array. The resistor is axial, matte black and cylindrical, with no notches or contours, and is labeled "1.87 kOhms", "+/- 3%", "KRLP IC", and "8047". In the schematic, it is drawn as a standard 1.87k resistor but is also marked "3% T.C."

Resistor as notated in schematic

My initial though was that "T.C." stood for temperature coefficient, and the resistor was placed such that it would compensate for the changing behavior of the transistors as they began to heat up, perhaps keeping the oscillators in tune. But a 3% temperature coefficient would be 30,000 ppm/C which seems impossible.

What kind of resistor is this? Why is it stuck to the IC? And to what does "3% T.C." actually refer?

Thanks!

Additional info if needed:

  • Here's the link to the service manual, which includes the full schematic.
  • The arrangement occurs twice in the circuit, one located in each of two voltage-controlled oscillators.
  • Neither VCO is functioning.
  • The synth has apparently undergone one other previous repair, somewhere between 15-20 years ago.
\$\endgroup\$
4
  • \$\begingroup\$ Perhaps a larger portion of shematics could reveal the function of the resitor and IC \$\endgroup\$
    – Marko Buršič
    Jan 12, 2016 at 23:45
  • \$\begingroup\$ Schematic I'm working from: synthfool.com/docs/Arp/ArpSolus/arp_solus2.jpg Resistor is along the sparse bridge towards the upper middle of the schematic, between Z4,pin1 and Z11,pin5. \$\endgroup\$
    – bleep
    Jan 12, 2016 at 23:46
  • 1
    \$\begingroup\$ It is very likely a 3300ppm/C "tempco" resistor with 1.87 nominal resistance at 25° \$\endgroup\$
    – Nils Pipenbrinck
    Jan 13, 2016 at 0:05
  • \$\begingroup\$ A standard exponential current source, used to drive a VCO, has a coarsely 3300ppm/C temperature dependence in the relation between \$V_{BE}\$ and the current. The usual trick is to use a 3300ppm/C tempco in the voltage divider/opamp that sums the control voltages, canceling the dependence. Although the schematic looks like that this is not exactly the standard arrangement, nonetheless I'm 99% sure that's what the part is and that's what it's doing. \$\endgroup\$
    – Timo
    Jan 13, 2016 at 17:49

5 Answers 5

Reset to default
24
\$\begingroup\$

It would appear to be a thermistor, for which a 3% temperature coefficient (usually referred to as "alpha" in the datasheet) is not unusual. 1.8k is probably the resistance at 25C.

It's probably a temperature-feedback mechanism, either to protect the transistor array from overheating, or to correct the behaviour of the circuit with respect to the temperature dependence of the transistor array. As BJTs get hotter, they conduct better.

Vishay intro to NTCs. Yours may be positive or negative, who knows?

\$\endgroup\$
2
  • 1
    \$\begingroup\$ You're right. It is temperature compensation. It sits directly at the tail of a so called expo-converter that does the 1v/octave scaling. The circuit is somewhat crude but the transistor in the feedback of the temperature sensed opamp does exactly that. \$\endgroup\$
    – Nils Pipenbrinck
    Jan 13, 2016 at 0:02
  • 2
    \$\begingroup\$ As synth enthusiast/hobbyist builder, I can verify that exactly this use is standard in synthesizer VCO's, including the specific transistor array. The 3% TC does seem to be 10 times too big though, 3300ppm/C is standard. Maybe it's mismarked on the schematic? \$\endgroup\$
    – Timo
    Jan 13, 2016 at 16:47
8
\$\begingroup\$

This is probably (as you initially suspected) a temperature-sensitive resistor, or a thermistor. These come in positive and negative temperature coefficients, and can indeed have coefficients in the 3% per degree C range. Thermistors are not all that linear with temperature, and usually are specified with a curve and a "K" value, rather than a %TC number.

It could be a heater, but given the small size I suspect that it is a sensor instead.

[edit] Looking at the schematic, it's not a heater. It is doing temperature compensation of offset or control voltage, perhaps to keep the frequencies stable?

\$\endgroup\$
1
  • \$\begingroup\$ I'd buy that... \$\endgroup\$
    – Daniel
    Jan 12, 2016 at 23:55
5
\$\begingroup\$

I have not looked at the schematic, but it's a temperature compensation thermistor for a log/antilog circuit. The chip is a transistor array (I used them for instrumentation back in the old days), and the thermistor is supposed to be thermally coupled to the chip. Below you'll see a typical schematic of such as circuit (taken from here) with a ~0.3% thermistor.

enter image description here

(Much) more on log/antilog amplifiers can be found here and here including the math behind why the coefficient has to be around +3400ppm/K at room temperature.

The nearby dual op-amp (RC4558) is probably associated with this circuit. Typically in a music synthesizer it would be used to create a VCO with an exponential response of frequency with respect to control voltage. This is really old stuff (1970's maybe judging by the photo). Not much goes wrong with it, fortunately, and most of what goes wrong should be easily fixed. I'd suspect power supplies first, then the RC4558s. Don't fool with the thermistor/CA3086 unnecessarily-- they are probably the hardest parts to get.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks, Spehro, lot of good info there. The synth is from 1980. The power supplies are working fine and it even makes noise when you turn the filter resonance to the point where it self-oscillates. It just seems the oscillators are both out. Last night I clipped the leads to one of the thermistors and tested the resistance and it doesn't seem to be conducting at all. But that wouldn't be enough to kill it. At this point the only suspect parts are the CA3086 and the surrounding FETs \$\endgroup\$
    – bleep
    Jan 13, 2016 at 18:18
2
\$\begingroup\$

Given the high values of the resistors around it, it is not being used as a heater, but rather as a sensor. It appears to be temperature-compensating the log-to-linear converter, which is "upstream" from the actual VCO. For example, take a look at the one built around Z11 and associated components on the upper right corner of the first schematic page.

\$\endgroup\$
0
\$\begingroup\$

It is most likely the heater for the temperature control (TC) being implemented by one of the spare transistors in that IC.

Edit:

Having a look at the full schematic, it is apparent that it's not an active control, but appears to be working as a heater of some sort for the transistor array Z12.

I can't explain why it runs off of a control voltage!

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.