Super-Drifty Silver Mica Capacitors

Discussion in '"Boat Anchor" & Classic Equipment' started by KW4H, Oct 18, 2021.

ad: L-HROutlet
ad: l-rl
ad: Left-2
ad: Left-3
ad: abrind-2
ad: L-MFJ
ad: L-Geochron
  1. VE3EKJ

    VE3EKJ Ham Member QRZ Page

    I'm an old guy who was building tube receivers when he was not even 11 years old. While temperature compensating caps were common I have never seen or heard of resistors as such!

    If they were around they must have been an extremely rare bird. When I got older my career was selling electronic parts to manufacturers, working with engineers as well as buyers.

    Still, I'm not too old yet to learn a new thing! Could you be kind enough to give me a link or a reference to such resistors?

    Bill V E 3 E K J
  2. KW4H

    KW4H Platinum Subscriber Platinum Subscriber QRZ Page

    I'm curious as well -- electronics is a field where I should never stop learning. Did some online searches and learned that temperature compensating resistors have SIX color bands, with the sixth being a temperature coefficient ranging from 25 to 100 ppm. I've never run across these in any circuits I've worked with over the past 40 years -- at least I don't think so. This is very interesting...

    Steve, KW4H
    PY2RAF and HA3FLT like this.
  3. K0OKS

    K0OKS Ham Member QRZ Page

    While temperature compensating resistors do exist I have never used them.

    I was referring to temperature coefficient of run of the mill resistors.

    Every part has some temperature coefficient where its nominal value varies with respect to temperature. In critical circuits parts are selected with differing temperature coefficients in an effort to reduce any overall change in the circuit with change in temperature. If one component drifts down in value when it gets hot sometimes a designer can use another component that drifts up in value when it gets hot, such that the overall circuit behaves the same when hot or cold.

    This info is in the Handbooks as well. It’s definitely in the older ones, but I’m pretty sure it’s in the section on basic components in the newer Handbooks as well.

    Modern components have much less drift, and modern circuits are less susceptible to drift because they use crystals and ICs that package a bunch of components into one thermal component that is often laser trimmed at the factory for consistency.
  4. SM0AOM

    SM0AOM Ham Member QRZ Page

    The problem is just that the resistance values around LC frequency determining circuits do not enter the frequency equations. They only influence the circuit damping and DC operating points, which have only second or third order effects on tuned frequencies.

    When looking for causes of frequency drift in tuned circuits, capacitors come first and then inductors. This was once taught to me in great detail in "Circuit Design part 2".

    W1BR, PY2RAF and N2EY like this.
  5. K0OKS

    K0OKS Ham Member QRZ Page

    I completely agree. I guess it also depends on the degree of drift you are trying to eliminate.

    Since we are talking about (mainly) LC circuits it does figure that that the L and the C would be there just important, with the C obviously much more prone to temperature effects.

    And of course capacitors have temperature coefficients too… even inductors…
  6. WA1GXC

    WA1GXC Ham Member QRZ Page


    I might be wrong, have not seen your source--but came upon an Internet information page that suggested what you just stated--resistors with a "Temp. Coefficient" color-band marking.
    I believe this is wrong. What the writer assumed was T.C. is actually a reliability rating and is independent of real-time characteristics.

    There is a system of resistor color-code markings for Military-spec qualified resistors called "R C R" series--they're not common but widely available and do get into the surplus pipeline.

    The last color band is a predictive characteristic for reliability, and shows the typical resistance value drift with age--e.g. 3% after 1000 hours in-service, etc. -- it's not temp. coeff.

    Would be interested to see any further authoritative info on "temp coefficient" components.

    N2EY and KA0HCP like this.
  7. W1BR

    W1BR Ham Member QRZ Page

    Polystyrene caps have good RF characteristics.
  8. KM1H

    KM1H Ham Member QRZ Page

    What he said...period. I dont know where some of these other ideas come from when dealing with simple HF circuits.

    If you want to get to 100-300 GHz and beyond it is an all new world as I experienced with a part of the F-35 avionics. Even fractal antennas are used:D

    KW4H likes this.
  9. KW4H

    KW4H Platinum Subscriber Platinum Subscriber QRZ Page

    Thanks for the clarification, Carl. What you and Karl-Arne said mimics my training. And replacing that eBay-purchased mica absolutely ended the drift. The radio has other problems, but the drift entirely went away. Only thing I haven't done yet is compare it to the much cheaper ceramic C0G; that will have to wait as I'm digging through other issues in the radio and don't want to create a disturbance in the force. :p

  10. WA1GXC

    WA1GXC Ham Member QRZ Page

    Sorta yes, sorta no.

    Too much information Dep't.

    I was on the right path, but gave an incorrect rendering of the facts, from memory.
    Follows is an Allen-Bradley Mil-spec molded resistor catalog link. You can slog thru it, but I'll summarize.

    "R C R " final color-band marking is reliability - "MTBF"--Predicted Mean Time Between Failure---Statistical mean failure rate per 1000 hours in-service.

    Brown--designator M--1% failure
    Red--designator P--o.1% failure
    Orange--designator R--0.01% failure
    Yellow--designator S--0.001% failure

    There is a temperature "Characteristic", most commonly "G", but appears to represent maximum ambient temp. in service and max resistance drift at that limit.
    "Temperature Stability--Between 0 deg. C and 85 deg. C, Allen-Bradley hot-molded carbon composition resistors exhibit a very low temperature characteristic, typically less than 2 percent deviation
    from room temperature values, less in low resistance values."

    See page 5 for five color-band marking system

    The MIL specification documents are older MIL-R-11 and more modern MIL-R-39008

    "When I'm wrong, I say I'm wrong."
    --Actor Jerry Ohrbach, "Dirty Dancing"

    Consult your attorney.

    Last edited: Oct 24, 2021

Share This Page