Tweet
Howdy,
I have a long thread concerning some amp squeal going in the "theory & design" subforum, and although I hope to close it out soon, I encountered something today which I figured I would start a different thread on. (the squeal thread can be found here http://music-electronics-forum.com/t41633/
Having reassembled what is essentially a modified tweed bassman amp (I have some switchable plexi like gain mods), I was playing it today with the back panel off and the volume/gain maxed and after a few minutes noticed some smoke coming out of it. It looked to me like it was coming from somewhere near the output tubes, but it was not flowing constantly and was very light in color, so it was hard to tell. I shut the amp off, looked for any visible signs of damage but saw none. I fired it up again, played some more, and again started seeing intermittent smoke.
After the second time, I tried to feel around for hot parts. My 5W 470 Ohm screen resistors as well as the 1/2W 1.5k grid stoppers were not hot at all. The base of the 5881 output tubes themselves were extremely hot, but I would not be able to say if they were abnormally hot. I have a string of 5 10V 5W zener diodes between the center tap and ground in order to lower the B+, and those were hot enough that I could not hold my finger on them. However, I did not think the smoke came from where the zeners were located, so this may or may not be a related issue.
The 5881 output tubes are no longer closely matched, which "may" be a result of them having gotten damaged during my debugging phase of the thread referenced above. They idle about 5mA apart, and although this is probably not an extreme mismatch, my first instinct is to replace the tubes with a fresh pair an monitor the amp for more smoke. Before I do that and potentially destroy another pair of tubes, though, I'd love to get some thoughts on whether there are other things to check first.
I have the tubes biased at around 35mA and 30mA, respectively. With a plate voltage of ~430V and a rating of 23W, that essentially equates to somewhere in the 56-65% range of max plate dissipation. The tubes are reissue Tung-Sol 5881s and are rated only for 400V, but I have read that with a tube rectifier, they can handle more. I checked the bias again before playing the amp the second time, and I monitored it while playing - any time I stopped playing the bias was back to where I set it, so it did not seem to drift off.
I am a little bit concerned about the zeners getting very hot as I somehow shorted them all during my debugging phase and had to replace them (not sure what caused that to happen, but at around the same time my GZ34 rectifier tube started flashing, so I don't know if that caused the zeners to short or if the failed zeners damaged the rectifier, or if these are unrelated - I did also replace the rectifier tube). Anyway, my PT has a max current rating of 250mA, so even worst case the 10V zeners should in theory be dissipating 250mA * 10V = 2.5W which is half of their 5W rating. Is there any situation where more current than this could be flowing through the zeners and CT?
Anyway, does anyone have any experience with their amps smoking? Can a power tube itself ever smoke if it going bad , or is it usually a component such as a resistor, PT, choke, etc.? Any thoughts on what I should do first - replace the output tubes and hope that is the issue, or something else? I assume there is too much current flowing somewhere, but what, where and how do I test for that? My plate and bias voltages seem fine, but since the bias is checked when the amp is idle, I don't know that everything is fine when I am actually playing.
I have a long thread concerning some amp squeal going in the "theory & design" subforum, and although I hope to close it out soon, I encountered something today which I figured I would start a different thread on. (the squeal thread can be found here http://music-electronics-forum.com/t41633/
Having reassembled what is essentially a modified tweed bassman amp (I have some switchable plexi like gain mods), I was playing it today with the back panel off and the volume/gain maxed and after a few minutes noticed some smoke coming out of it. It looked to me like it was coming from somewhere near the output tubes, but it was not flowing constantly and was very light in color, so it was hard to tell. I shut the amp off, looked for any visible signs of damage but saw none. I fired it up again, played some more, and again started seeing intermittent smoke.
After the second time, I tried to feel around for hot parts. My 5W 470 Ohm screen resistors as well as the 1/2W 1.5k grid stoppers were not hot at all. The base of the 5881 output tubes themselves were extremely hot, but I would not be able to say if they were abnormally hot. I have a string of 5 10V 5W zener diodes between the center tap and ground in order to lower the B+, and those were hot enough that I could not hold my finger on them. However, I did not think the smoke came from where the zeners were located, so this may or may not be a related issue.
The 5881 output tubes are no longer closely matched, which "may" be a result of them having gotten damaged during my debugging phase of the thread referenced above. They idle about 5mA apart, and although this is probably not an extreme mismatch, my first instinct is to replace the tubes with a fresh pair an monitor the amp for more smoke. Before I do that and potentially destroy another pair of tubes, though, I'd love to get some thoughts on whether there are other things to check first.
I have the tubes biased at around 35mA and 30mA, respectively. With a plate voltage of ~430V and a rating of 23W, that essentially equates to somewhere in the 56-65% range of max plate dissipation. The tubes are reissue Tung-Sol 5881s and are rated only for 400V, but I have read that with a tube rectifier, they can handle more. I checked the bias again before playing the amp the second time, and I monitored it while playing - any time I stopped playing the bias was back to where I set it, so it did not seem to drift off.
I am a little bit concerned about the zeners getting very hot as I somehow shorted them all during my debugging phase and had to replace them (not sure what caused that to happen, but at around the same time my GZ34 rectifier tube started flashing, so I don't know if that caused the zeners to short or if the failed zeners damaged the rectifier, or if these are unrelated - I did also replace the rectifier tube). Anyway, my PT has a max current rating of 250mA, so even worst case the 10V zeners should in theory be dissipating 250mA * 10V = 2.5W which is half of their 5W rating. Is there any situation where more current than this could be flowing through the zeners and CT?
Anyway, does anyone have any experience with their amps smoking? Can a power tube itself ever smoke if it going bad , or is it usually a component such as a resistor, PT, choke, etc.? Any thoughts on what I should do first - replace the output tubes and hope that is the issue, or something else? I assume there is too much current flowing somewhere, but what, where and how do I test for that? My plate and bias voltages seem fine, but since the bias is checked when the amp is idle, I don't know that everything is fine when I am actually playing.
Comment