Marshall TSL:100 bias drift problem

[Enzo]

Well, that aside, we would already know we didn't have a bunch of hum in the speaker. SO we would find some several volts of DC and nothing audible to explain it. Implying RF. I suppose you would be right, if we had a bunch of hum, we would not be able to detect simultaneous RF with this. In other words the context answers your concern.

[lowell]

So at power on, had nothing. After about 40 minutes, the hum is loud, and there is about 60mvDC using the diode/cap. It's a bit odd, if I set my meter at 6v it reads 500mv... if I set it at 600mv it reads 60mvDC. Seems to be a weird meter symptom/issue. The hum is gone at power on and grows over the 40 minutes.

Doesn't seem like enough DC voltage to be symptomatic of an oscillation does it? Even the hum itself I would think would read higher on the meter as it's a good amount of hum. (technical speak)

[Enzo]

No, I was expecting like several volts.


You have a signal generator? Set it up to something like 50kHz or 100kHz. Something WAY above audio freq. Set the output level to something strong. What would be a couple volts at audio. Now measure that with your diode thingie. What sort of reading does it give? Just an experiment.

By the way, that diode circuit is what we call a "detector" and is the basis for an AM radio.

[lowell]

Don't have the signal generator either, but can probably make this happen in the coming days.

I am replacing the bulging main B+ caps. 330uf 315v. They are 20mm diameter and 10mm lead spacing. Cannot find these anywhere. Does anyone know where to find em?

[Enzo]

315v? Oh please. Look for something standard like 330uf at 350v instead.

COmpanies like Marshall order parts in huge amounts. They can order them custom for their application. Off the shelf items may not match up. But it is just a power supply filter.

[g1]

I guess it's the spacing giving you the trouble?
Are they snap in type or radial leads?

[lowell]

Haha I know Enzo... I don't care about the 315v... it's more the SIZE. 330uf 350v with the 10mm lead spacing and 20mm diameter is what's difficult. Marshall did not allow for larger radial caps. I order some larger ones on Mouser... and will make em work using silicone glue. Just figured I'd see if anyone had a direct hit on them.

[lowell]

got those caps in and managed to get the amp to behave for 4hrs. I replaced the 2 main filter caps. After 4 hrs the bias was at 85mv each side. Still not sure I fixed it, as it still rose over time to 85mv, but after 4 hrs it's not running away. And I just suggested to the owner that he turn it off or put in standby between playing it.

Enzo, I did your test. It seems that anything 50khz and above doesn't register on my meter. However at 50khz into the diode/cap filter, I injected 1vAC and had about 9vDC. Does NOT make sense haha but that's what my meter told me. I'm guessing the meter is no good at higher vAC, but does just fine with vDC.