• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Tube amp worsens after 6 hours or so

nooboy

nooboy

Active Member
Hello. So here's the thing: I have an antique audio (ASL) combo driving Dynamo a-25 speakers. 2x25 watts. Kt-88 shuguang tubes, and buggle boys. Anyway... I love the set up, really, but after about 7-8 hours at 12 o'clock volume, it starts sounding like shite. Muddy, like every record I put on was full of dust. Yet, bias meter a-OK. I can live with it, as I'm usually a 5 hour tops listener, but I wonder what's happening. Static build-up? Overheating? The problem has always been there, at least 4 years.
 
Register to hide this ad
does it seem to be getting overly hot? The tubes aren't starting to glow a dull red are they? Kinda sounds like a component getting hot and drifting off value or failing. Could even be a marginal tube getting good and hot and developing some internal leakage .
 
Gadget: Maybe, but the overall sound is affected, not just one canal. I'm grasping for straws, thinking it might be a common thing with my kind of pseudo early 70's set-up. Might be the phono stage, but my instincts point to the combo amp.
Whoaru: 12 , in triode mode (I should have said that), on my kind of speakers, is pretty reasonable.
The both of you: thanks.
 
The likeliest common mode problem would be the HV power supply slowly sagging in the power amp, comfirmable by SAFELY monitoring voltages during play session.
Otherwise try the Gold Lion KT88 new issue tubes from Jim McShane if you really like the amp.
 
Last edited:
gadget73 said:
Problems in both channels tend to point towards power supply issues.
Click to expand...

If the bias supply is shared a coupling cap or caps being bad can draw down the bias .Or a bad cap in the bias supply , but Getting worse as it gets hot , is most likely a coupling cap
 
Coupling cap. Got it. i imagined something like static buildup. As it has been a stable problem, I'd be tempted to let it be (I rarely keep my system on for more than six hours). Hopefully, it won't get worse. Thanks.
 
Or a bad carbon composite resistor which drifts with temperature.

Not only does the value change with age, but they often have weird behavior when they get hot. Especially if they've ever gotten damp.
 
If it was made in the last 20 years, I'll take a wild guess and say it probably doesn't have CC resistors in it. They aren't real mainstream anymore. Some old ones held up OK but by and large they just aren't amazing, nor would I install new ones in anything. I also don't have any special feelings about keeping them, other than in some cases its a lot of trouble to replace the things.
 
gadget73 said:
If it was made in the last 20 years, I'll take a wild guess and say it probably doesn't have CC resistors in it. They aren't real mainstream anymore. Some old ones held up OK but by and large they just aren't amazing, nor would I install new ones in anything. I also don't have any special feelings about keeping them, other than in some cases its a lot of trouble to replace the things.
Click to expand...

They aren't mainstream, but they were used for grid and screen stoppers because of a belief that inductance matters.

Audioidolatry says that carbon comps are superior for this purpose, but the fact remains that they are an inferior technology that likely should vanish from audio. The only time CC resistors are technically useful is in grid or screen stopper resistors because the value isn't highly critical, so small amounts of drift are somewhat ok, and the non-inductive nature means that they won't form LC circuits with tubes. (For those of you unfamiliar with this search for "Miller Capacitance".) That's really the issue.

But it truly isn't clear how bad the spiral cut metal film is for this purpose. A couple of pH just don't matter, especially when the wiring and PCB have far more inductance. Does this matter when dealing with RF circuits in the 400+MHz range? Absolutely, it's a big deal. Worse at GHz! But at audio frequencies? Not so much.
 
I remember looking into that once, and the inductance values are indeed quite small. Not nearly enough to make any difference at 20kc. RF work is definitely it's own universe though. Even in fairly pedestrian stuff like AM and FM radios you run into goofy things like "gimmick" capacitors and very particular lead placements to keep things behaving. I remember playing hell with a 1940s Westinghouse tabletop AM radio that decided it was going to oscillate after replacing a paper cap with a film. I guess I didn't get it positioned just so and it went absolutely nuts. Took a while of pushing and prodding parts around to find that, and fixing it was simply a matter of re-bending the lead a bit to move the part around to a happy place. I've had that happen with other tuners too, usually the same fix is applied once you find the problem part.
 
gadget73 said:
I remember playing hell with a 1940s Westinghouse tabletop AM radio that decided it was going to oscillate after replacing a paper cap with a film. I guess I didn't get it positioned just so and it went absolutely nuts. Took a while of pushing and prodding parts around to find that, and fixing it was simply a matter of re-bending the lead a bit to move the part around to a happy place. I've had that happen with other tuners too, usually the same fix is applied once you find the problem part.
Click to expand...

You have my sympathy on that. Debugging poorly designed circuits is awful because every piece of it wants to become a giant oscillator.

The problem arose because the circuit is fundamentally unstable and oscillates. The leads act as antennas for both broadcast and reception.

What you were seeing was likely a high-frequency noise carrier modulated with a lower frequency. That noise is then captured by wires throughout the preamplifier or amplifier and demodulated to be heard as noise and further instability and oscillation.

This is why grid stoppers (among other types of stoppers) are crucial to make circuits stable, and why Miller Capacitance matters as it creates RC and LC oscillators out of every tube in the path.
 
I'm guessing the higher ESR of the old paper cap acted as a grid stopper or something. Its an overly-crowded AA6 type set. I've had more trouble out of "simple" radios than complicated beasts. I suspect its just because the "simple" ones use marginal engineering to make it go for the least coin. The fancy ones are actually built in a sane way and tend to behave themselves.
 
gadget73 said:
I'm guessing the higher ESR of the old paper cap acted as a grid stopper or something.
Click to expand...

The ESR would only be a few ohms, and the grid stopper would be 40K. I suspect it was lead length or position.

gadget73 said:
Its an overly-crowded AA6 type set. I've had more trouble out of "simple" radios than complicated beasts. I suspect its just because the "simple" ones use marginal engineering to make it go for the least coin. The fancy ones are actually built in a sane way and tend to behave themselves.
Click to expand...

Yeah, that is true. The el-cheapo American Sixes leave out stabilization resistors, snubber caps, and enough filtering caps. You can easily undo those design decisions and make a much more stable radio.
 
That particular one though is just too tight and frankly not worth the trouble. Its one of those Westinghouse Little Jewel "refrigerator" radios. Chock full of octal tubes and packed into a chassis that it barely fits on. Its a good performer as it is, but I just don't have the ambition to try and make it better for the rare occasions I use it. Honestly I was amazed that it was a 6 tube set when I got it. I'd have figured a fiver with no RF stage with the chassis the size that it is.
 
You must log in or register to reply here.
Back
Top Bottom