Testing old resistors

[c.coyle]

When testing old carbon comp resistors in old equipment, my experience has been that more resistors drift high than low. Is that consistent with your experience?

I first test all resistors in-circuit. If it's high, it's high, and it automatically gets replaced. Then, if I'm feeling motivated, I lift a leg of the remaining resistors to see if any have drifted low. But usually I'm lazy and I just stop after the high test if everything is working OK.

 

[EngineerNate]

Makes sense. Resistors tend to fail open, not short. Even moisture intrusion wouldn't significantly reduce the resistance as water itself isn't very conductive, so the breakdown in the composition caused by the moisture would likely offset any resistance-reducing effects.

 

[avionic]

Resistors always go high. I have never in 45 years seen 'm go low.

 

[OMGCat!]

Resistors always go high. I have never in 45 years seen 'm go low.

I don't have nearly as much time in but I agree completely. I've only seen a range from a little high to open. Never saw one go low.

Seems like high value carbon comps are big offenders. Anything in the high k to Meg range seem to nearly always be wide open or some ridiculously high value.

 

[ChrisMarantz]

You should also consider that Carbon-Comp Resistors can get noisy even if they measure spot on.

A good Signal Tracer, like the Heath Kit IT12 can be very useful in finding noisy components - Chris

 

[dbxdx5]

You should also consider that Carbon-Comp Resistors can get noisy even if they measure spot on.

A good Signal Tracer, like the Heath Kit IT12 can be very useful in finding noisy components - Chris

Good to know. I recently picked up an IT12.

 

[Retrovert]

Use the procedure developed by the late Alan Douglas: drop the carbon-composite resistor in the trash can. If it makes a thunk or thud, it's definitely bad and should be replaced. If it makes a tink it's going bad and should be replaced, if it makes no sound it wasn't dropped it into the can hard enough.

Carbon composite resistors should all be replaced. Search for Vcr and you'll see why.

 

[triode17]

When testing old carbon comp resistors in old equipment, my experience has been that more resistors drift high than low. Is that consistent with your experience?

I first test all resistors in-circuit. If it's high, it's high, and it automatically gets replaced. Then, if I'm feeling motivated, I lift a leg of the remaining resistors to see if any have drifted low. But usually I'm lazy and I just stop after the high test if everything is working OK.

When I rebuilt my 66 Fender Vibro-Champ, all the CCs had drifted high. A few had DOUBLED in resistance. The good thing about this is, the circuits' limit themselves, the bad thing is, the amp. doesn't perform well. I used all metal films and the amp. never sounded better!

 

[Retrovert]

When I rebuilt my 66 Fender Vibro-Champ, all the CCs had drifted high. A few had DOUBLED in resistance. The good thing about this is, the circuits' limit themselves, the bad thing is, the amp. doesn't perform well. I used all metal films and the amp. never sounded better!

Word. True dat.

The CCs in old radios are notorious for such drift, yet the radio, like the guitar amplifier, still sort of works, more or less. A tribute to how robust the circuits are, and how easy it is to make a radio receiver or a guitar amplifier. Note that I did not say a good radio receiver or a good guitar amplifier, I just said a radio receiver or guitar amplifier. Tubes will basically work with horrible component choices or drifted values. Again, I wrote "work", not properly work in any way one wants.

Having a spot-on circuit delivers the original design goals, if not the original implementation results.

 

[Retrovert]

You should also consider that Carbon-Comp Resistors can get noisy even if they measure spot on.

Are inherently noisy.

All those grain junctions have a cost.

 

[Retrovert]

Makes sense. Resistors tend to fail open, not short. Even moisture intrusion wouldn't significantly reduce the resistance as water itself isn't very conductive, so the breakdown in the composition caused by the moisture would likely offset any resistance-reducing effects.

The reason is that carbon-composite resistor are made from grains composed of carbon black (conductor) and clay (insulator) which are pressed together. More carbon, lower resistance, less carbon, higher resistance. As the clay swells (hygroscopic) the properties change.

Moisture does not reduce the resistance, it is bound to the silica.

Resistors do drop in value. The cause is high current paths create conductive channels by fusing grains together. The MOV does this, and the pathways drop in resistance until until it eventually shorts out (low enough resistance to pass high current), heat up, and starts a fire. I've elsewhere explained that in detail.

 

[gadget73]

I've also never had one drift low. I've also found that the obnoxiously noisy CC resistors have also been off value. I won't say its impossible for one to be noisy and on value, or to drift low but that hasn't been my experience thus far. I'm also not opposed to just changing them, especially if I get into a piece of equipment with a bunch of off value parts. At some point it ends up taking longer to check and replace things individually than it does to just shotgun the whole mess, especially if you're doing caps already.

Depending on the gear, noise may not matter. I left a bunch of original but on-value parts in my DC power supply because it just didn't matter.

 

[Retrovert]

The original resisters were very noisy when new. So noisy that wire-wound was often used at ten times the cost. The "cracked carbon", which really was a vapor deposition process, was touted by Mullard for the 5-20 as it removed much of the traditional resistor noise. Such noise was, as I've shown from quotes by Hafler, one of the issues for the PAS.

I would not leave carbon composites in place in a power supply because of the potential for drift. The phenolic coating was never that good when new, and it degrades over time, and with heating, allowing ingress of moisture. Asking it to survive sixty years and still function is asking a lot.

I agree with the statement that it takes more time to verify function and selectively replace than to do the complete overhaul. The other advantage of a complete rebuild is replacing the degraded phenolic sockets with porcelain ones. Polish the chassis while everything is gone and it looks really nice. Everyone's crazy for a sharp and shiny amp...

 

[gadget73]

Noise being a relative term. When it starts crackling and popping through the speaker in a very obvious "well that ain't right" sort of manner, the part has deteriorated beyond whatever noise it would have made when brand new.

This is a bench supply with variable output. I don't use it terribly often. It can be imperfect and still get the job done.

 

[Retrovert]

The standard carbon-composite noise is Johnson-Nyquist noise aka waterfall noise. Sounds like running water. Even when brand new it is plagued with that.

This is why cracked carbon was used in the 5-20. Amazing invention for the time. Cost about three to five times more than the composite type, but still about about half of what a wire wound cost (10x the CC).

 

[EngineerNate]

The reason is that carbon-composite resistor are made from grains composed of carbon black (conductor) and clay (insulator) which are pressed together. More carbon, lower resistance, less carbon, higher resistance. As the clay swells (hygroscopic) the properties change.

Moisture does not reduce the resistance, it is bound to the silica.

Resistors do drop in value. The cause is high current paths create conductive channels by fusing grains together. The MOV does this, and the pathways drop in resistance until until it eventually shorts out (low enough resistance to pass high current), heat up, and starts a fire. I've elsewhere explained that in detail.

Today I learned something new!

I guess I was mostly thinking of drift under regular operating conditions as opposed to serious overcurrents.

 

[Retrovert]

Electricity finds the lowest resistance path and uses it. Carbon-composite resistors are awful things. Give them all the death penalty and replace with shiny, new metal films.

 

[EngineerNate]

Electricity finds the lowest resistance path and uses it. Carbon-composite resistors are awful things. Give them all the death penalty and replace with shiny, new metal films.

We need an "I'm with him" smiley.

 

[dbxdx5]

You should also consider that Carbon-Comp Resistors can get noisy even if they measure spot on.

A good Signal Tracer, like the Heath Kit IT12 can be very useful in finding noisy components - Chris

I'm new to signal tracers in general and the Heathkit IT12 specifically. I believe I read that there's a risk of damage to transistors when using these. What precautions should I take?

 

[smurfer77]

[EDIT: there are some mechanisms for decreasing resistance of resistors over time, but they are not too common from my point of view. read posts below for more info. The statement I made below is not strictly correct but i leave it for clarity/history.]

There is no failure mode for resistors to drift low. The core conductor becomes compromised by a number of possible mechanisms and this never results in increased conductivity, only reduced conductivity (there are some materials which increase conductivity under some of scenarios but they are not used for making resistors) and hence increased resistivity.