Squirrelly bias

[dmag]

Im working on a Lafayette 250a integrated amp. The schematic link is for the kit version, KT250a. Same, but legible. The bias drifts some, not much, but enough. It always changes on startup. it can be .45 or it can be .50.All caps have been replaced, resistors have been checked for odd man out, those tied to the bias circuit replaced. All tubes have been swapped and/or replaced with others. The problem is most prevelant on the left channel. All pots have been cleaned. The balance and bias pots look like dual gang, but work individually, and i cannot source new ones. The left side feels a ever so slightly looser than the right. It seems extremely sensitive, as if just touching it with a screwdriver causes it to change. Ive opened them up and they are nice and clean. I had considered opening it up again, and tightening the contact a bit. Stock pots are 10k for bias and 5k for balance. Theres not much room to put in seperate pots. I have a pair of 2x10k i can try. Would the 10k section work for the 5k? Perhaps with a simple mod? I am giving this to a tube newbie friend, so i need to be sure it will be stable.
Oh, and Happy Thanksgiving
http://www.freeinfosociety.com/electronics/schemview.php?id=2382

 

[Alan0354]

Where are you measuring? R34, R35 or R70, R71?

You mean when you turn on, every time the reading change upon start up? You sure you don't have anything loose? Tap on the components in that area around R79 and R81 and see whether you find any sore spot that changes when tapped.

It is not uncommon to drift upon warming up. I don't see any particular thing that can drift. Take a look at R79 and R81( I don't know which one is left or right), the one that adjust the overall bias. Compare the two channel by measuring the wipers of R28 and R64. Note the voltage down and the corresponding bias reading. Next time you power up, if it is different, measure the points again. Confirm the voltage on the wiper of R28 and R64 change corresponding to the change in the bias. If so, then your bias circuit has a problem.

If the bias changed, but the reading on the wiper of R28 and R64 stay the same, you have something else wrong.

 

[BinaryMike]

It seems extremely sensitive, as if just touching it with a screwdriver causes it to change. Ive opened them up and they are nice and clean.

One of the more insidious possibilities here is parasitic oscillation. Have you checked with a wideband scope?

 

[dmag]

I have not, nor do I know how to do that. With instruction I would give it a try.

 

[Alan0354]

You mean you don't understand my post?

 

[BinaryMike]

I have not, nor do I know how to do that. With instruction I would give it a try.

If you have access to a scope with bandwidth of 100MHz or greater, and a 10X probe rated for at least 500~600V, then you could probe around the pins on the power and driver tubes with the vertical amplifier AC-coupled and set for high sensitivity. Oscillation may be visible as ballooning of the trace without actually touching a pin, and it might clear up or get worse (as determined by DC bias change) when you do touch a pin with the probe tip. These things can sometimes be difficult to pin down, so the immediate goal is simply to confirm oscillation or rule it out. There are some unreliable but occasionally useful techniques you can try if a scope isn't available, such as using radio receivers to pick up RF emission or probing with a floating piece of metal such as a screwdriver while monitoring DC bias. DC bias should NOT change just from adding a few pF to any node in a stable amplifier. An RF spectrum analyzer would be even better than a scope for this job.

 

[dmag]

Alan0354, im sorry, i did not see a notfication to your post, and was responding to binarymike's post. But yes, I do understand it. Company arrived much to early today to get to the bench. Binarymike, iirc my scope is 100mhz. ill check the probes. I did some reading on parasitic oscillation, and will do a bit more tomorrow. Hopefully, Saturday I will be able to look into both suggestions. Thanks to you both.

 

[sKiZo]

Might be a tube thing as well - try swapping sides to see if the problem follows?

 

[jharris]

Do you mean it's wrong/different every time you use the amp (even if it's fully warmed up), or just inconsistent as the amp is starting up?

Tubes take different amounts of time to "warm up" to their full/stable current, much longer than it takes them to start glowing and working. Because they'll be all over the place during this warmup, set bias when the amp has been running for a while. I give it at least 20 minutes, others have their own methods (sometimes WAY longer)... now, once you've done this, those bias settings won't necessarily make the current matched/correct in the warmup period briefly after the amp is turned on. But you can usually let it warm up and you'll find it evens out, as long as your settings were correct/matched when you made them.

 

[Alan0354]

Alan0354, im sorry, i did not see a notfication to your post, and was responding to binarymike's post. But yes, I do understand it. Company arrived much to early today to get to the bench. Binarymike, iirc my scope is 100mhz. ill check the probes. I did some reading on parasitic oscillation, and will do a bit more tomorrow. Hopefully, Saturday I will be able to look into both suggestions. Thanks to you both.

I would look for the more obvious things first. use a wood stick to tap around and see whether you can make the noise.