10K screen grid resistor?

[sfriedman92]

I have a Savage Macht 12 (parallel SE 6V6s) that needed a new lamp holder. When replacing the lamp I noticed what I thought was one discolored 100 ohm screen grid resistor to both output tubes.
After pulling the resistor out, I measured it at 10.01K, which seems high (I would think 470R or 1K for 6V6s) but unlikely to be the result of drift due to heat damage. Anyone know if the 10K was a design choice? I'm thinking about just replacing the current 10K with a new 3W 10K. I can measure idle voltages if that's helpful.

I've called and left messages with Savage but can't get a hold of them.

Thanks!

[FUCHSAUDIO]

10-K seems a doubtful proper value. I've seen 100 ohms up to 2-K on screens, depending on amp model, power tube type etc.
It's possible both of them burned and went to 10-K, but that would be odd.

[roberto]

It could be used as an "effect" on the power amp, causing strong compression when the secondary emission takes place.
Another reason could be that they developed the amp using tubes with weak screen grids, so they protected them this way.

[sfriedman92]

Thanks for the thoughts! I was surprised by the 10K as well but seems unlikely to have drifted to that very specific 10K value. I don't think a 470Ω will hurt anything so I'll try it and see how it sounds.
If I hear back from Savage, I'll let y'all know.

[Stevem]

If was a carbon comp that burned up and not a wire wound, or a big metal film then they can go all over the place in value.

[pdf64]

… I don't think a 470Ω will hurt anything …

Consider the likelihood that screen grid voltage would increase, which will cause anode current to follow suit.

[sfriedman92]

Still haven't heard from Savage. For reference, here's the resistor. Discoloration isn't too bad but the image doesn't pick up the brownish tint on the component that well.

[Reeltarded]

It is a 100R resistor. brown black brown.

[ViperDoc]

Taking bets on whether that was "overheated to a near-perfect 10K", or.... most likely a rare factory mislabel? Replace with a fresh 100R.

I experimented on an Express build with a 2W PEC 10KL pot in series with a resistor and found more of a noticeable rise in compression/transient loss with lowered screen voltage than I expected, even up to 2K. Maybe that would be less noticeable on lower gain amps, never tried it on another circuit.

[rooster]

Fender 5E3 Tweed Deluxe used a 5K screen resistor and 10K was used in earlier models, cathode biased 6V6 pair. He's copying a Fender circuit although the parallel 6V6 is not a Fender design exactly. And, like noted in the post before this one, a larger screen resistance creates more sag, note the Express and various Peavey amps over the years.

[sfriedman92]

@rooster maybe I'm misunderstanding but looking at Rob Robinette's schematic for a a tweed 5E3, I don't see any screen grid resistors: https://robrobinette.com/How_The_5E3_Deluxe_Works.htm. Correct me if I'm wrong (sincerely not trying to be a PITA) but he seems to say as much:

Although the 5E3 doesn't have them, screen grid resistors are used in most amps to prevent tube damage from excessive screen grid current ...

There's a 5K dropping resistor in the power supply as part of the power supply filtering network. Is that the 5K resistor you're referring to? Makes me think I should double check the power supply section since the amp doesn't have a choke. Could the resistor I'm looking at could be a dropping resistor connected directly to pin 4 of one of the 6V6s? The amp is point-to-point in the power supply section so it's a little more difficult to trace everything out.

I'll take a look tomorrow and report back

[timrobertson100]

That sounds plausible indeed.
Any chance they were both on pin 4 and then paralleled (i.e. direct wire between pin 4 on both 6v6s) as well which would result in a 5K?

When you have it open suggest taking some photos of the section.

[sfriedman92]

@rooster you nailed it. Thank you. I assumed this resistor was a screen grid. It's just a dropping resistor in the DC power supply filtering network. There are no dedicated screen grid resistors and the layout will make it difficult to add. My conclusion is the resistor is supposed to be 10K and has either a little heat damage or the color band changed over the ~18 years of life. I'll temp it back in to check the bias to make sure it's not way off with the 10K and probably bump up to a 5W 10K since I'm ordering parts anyway.

Thanks for the help y'all