Old carbon comp resistor drops to 1/20th of nominal?

[glebert]

Had a new "never saw that before" moment yesterday. Working on a friend's (formerly) all-original 1963 6G3 Deluxe and I was getting screen voltages higher than the plate voltage. Didn't know what the heck was going on, thought maybe I had miswired the doghouse somehow but then found the 1k two watt node resistor was measuring at 55 ohms. I don't think I've ever seen a resistor get more conductive. How does that even happen? I don't think it would be like when PCB becomes conductive because it gets carbonized, the resistor is intentionally a lump of carbon.

Anyway, I think I finally have the amp running right. What was supposed to be a 3 prong cord install and a quick health check (because he thought the amp was working OK), has led down may rabbit holes, goose chases, and dead-ends. Why do we do this again?

[xtian]

Why do we do this again?

Because a well cared for vintage amp is a thing of beauty?

[Chuck H]

many rabbit holes, goose chases,..

Anyone getting hungry?

Did you meassure the resistance out of circuit? I'd be concerned that the board could be conductive in that area.

[Helmholtz]

In rare cases CC resistors can arc inside when severely overloaded.
This causes the organic binder between the carbon particles to carbonize and results in low resistance.

[Chuck H]

In rare cases CC resistors can arc inside when severely overloaded.
This causes the organic binder between the carbon particles to carbonize and results in low resistance.

Like glebert I'd never seen this before. I'm wondering if this was a newer resistor or an original. Because I'm now wondering about the voltage rating for new manufactured carbon composition resistors as compared to vintage specs.

[glebert]

Like glebert I'd never seen this before. I'm wondering if this was a newer resistor or an original. Because I'm now wondering about the voltage rating for new manufactured carbon composition resistors as compared to vintage specs.

Original looking resistor, and does measure same out of circuit.

[Helmholtz]

Normally a screen resistor doesn't see much voltage (between its ends that is).
Things change with a shorting tube.Then voltage, current and especially dissipation can get high.

[The Dude]

I saw this once as a tech in a manufacturing setting. It was discovered that we had a batch of mislabeled resistors. I'm curious- was the color code on the resistor correct for its intended value?

[Chuck H]

I saw this once as a tech in a manufacturing setting. It was discovered that we had a batch of mislabeled resistors. I'm curious- was the color code on the resistor correct for its intended value?

It occurs to me that the stripe difference between a 1k and a 100 ohm resistor would be the third stripe being red or brown (respectively). And a red stripe might turn brown if the resistor was subjected to voltage such that it arced internally.?. But that would be opposite for detection purposes. So if it was a red stripe but it's brown now I guess there's no way to know what it started as.

[Mick Bailey]

Talking about wrongly marked resistors - I bought a pack of 5.6 ohm metal film resistors from an industry supplier and had no end of problems with the circuit I was using them in. It turned out they were 5.6k Ohm when I ended up measuring them. The packet, invoice and colour coding were all incorrect.

[Chuck H]

I've read about that sort of thing on only a couple of occasions here. That sounds awful. What I've seen for myself would be colors bands that are ambiguous. Like tan and brown being almost the same or orange and red being almost the same. With these mushy colors it's easy to misidentify a resistor that doesn't use both similar colors on the same resistor body.

[Tom Phillips]

I've gotten into the habit of measuring just about all resistors before I install them.

[Mick Bailey]

I now do that - it saves no end of lost time tracing a fault in a newly built circuit. Some of the synth stuff I build is quite involved to begin with and it pays to be extra careful. I never had a problem reading codes on older resistors with brown or tan bodies, but semi-translucent bands on blue or green metal film are a nuisance and ambiguous even under magnification. I'm thinking that human-readable codes are rather redundant in modern production. I jokingly said to a customer that he was probably the first human to touch his cheap Zoom multi-effects processor pedal when he pulled it out of the box.

[Greg Robinson]

The CRI (colour rendering index) of modern lighting types doesn't help either. CFL's and LED's frequently have CRI's in the low 80's, making it difficult or even impossible to distinguish colours. Even with near daylight levels of lux on my bench thanks to lots of overhead lighting, I often find myself taking things outside if I need to distinguish colour or detail. I do have a stash of halogens for lightbulb limiter duty, maybe I should get myself an articulating lamp for the bench with a halogen just for those tasks...?

[g1]

The CRI (colour rendering index) of modern lighting types doesn't help either. CFL's and LED's frequently have CRI's in the low 80's, making it difficult or even impossible to distinguish colours.

Sorry to go even further off topic, but some of the issues with LED light colour spectrum are just terrible. And they can slowly change without being that noticeable, unlike incandescent or other types that are either dead or alive.
https://nymag.com/strategist/article...stigation.html
https://www.washingtonpost.com/clima...ght-pollution/