'66 Reverberocket 7591 redplating

[firemedic]

I restored my '66 Ampeg Reverberocket a couple years ago with a recap, retube, new Jensen speaker, a new OT (it seemed like a Deluxe Reverb OT matched impedance pretty well w/ cathode biased 7591s) and new cathode resistor/cap.
At band practice I played it for @ an hour when I noticed the life started to drain out of it. Looked at the back and sure enough the plates were glowing red, one worse than the other but both redplating.
 
My knee-jerk thought is that the the PI coupling caps must be leaking a lot of DC; I have not replaced those. But wouldn't it have to be a lot of DC w/ a cathode bias output arrangement? Maybe the cathode bypass cap is leaking. Maybe the cathode resistor has gone bad. 
 
I haven't had a chance to open it up but I will post when I figure this out.

[eleventeen]

Seems like you are on the right track with the suspicion of the coupling caps. Who can know what "a lot" is? 25-30 volts leaking through (especially) the plate-coupling cap would quite disrupt that circuit. Also makes sense that one 7591 is redder than the other.


I tend to doubt it's the output stage cathode resistor but it's definitely not impossible. If that resistor opened up, there would not be any current path across the output tubes, so no redplate that way. If the cathode resistor rose up in value quite a bit (I guess it happens sometimes) then for some period of time, cathode voltage would rise quite a bit and could pop the bypass cap around that resistor since it's only 25 WVDC. Now, maybe you have a partially-shorted cap bypassing the cathode R. Cathode R goes way down in resistance, effective bias is reduced, and then you have too much current thru the 7591's. Redplate.


Should be fairly easy to troubleshoot & fix. Given the highish cost of 7591's, I would not skimp on parts and I'd get a much higher voltage rated cap for the 7591 bypass.

 

[firemedic]

Well I finally got the time to open her up again. First I tried to use my bias voltage & current adapters to figure out my current draw & it took me an embarrassingly long time to remember they won't work with the 7591s.   
 
Output section nominal voltages were pretty good. The 14.1v on the cathode seemed low but apparently it's correct, per the schematic. Plate voltage was 390v though, & the schem showed 360.
No DC volts across the PI coupling caps.

Something was up with my cathode resistor/bypass cap, even though they metered ok. For whatever reason I had some generic 63v 25uF cap in there so that got replaced with a 100v 100uF (the only decent one I had around).
The resistor was a 270R 10W in parallel with a 330R 10w to make a 150R resistor. I junked that mess & replaced with a single 180R 10W (schem calls for 140R 10W). Hopefully a little cooler bias will sound all right.

Plate voltage is now 362v and cathode is 15v. So whatever the issue was, exactly, I think it's fixed. I will post later with the Tone Report.

[eleventeen]

Your bias checker thingy won't work with 7591 because those tubes have a different basing diagram than 6L6/6V6/KT88.


http://www.drtube.com/datasheets/7591s-jj2005.pdf


If you're not redplating, that's certainly the good news.


My suspicion is that your cathode bypass cap got leaky = partial short, and because it is in parallel with it, effectively lowered the value of your cathode resistor. More current could thus pass thru it/them and this also lowered the cathode voltage. At the same time, this made the grid(s) more positive, meaning, the effective bias was not moving in your favor and the tubes went overcurrent. If your meter has an ESR setting, it would be interesting to see if the 63u/25v cap you changed out has a cruddy reading, if you haven't tossed it out. Otherwise, an ohmmeter check if you know how to interpret same.

[firemedic]

I had a phase where I bought a bunch of questionable caps just to have them, just in case. For whatever reason I put this one in while refurbishing the amp. It's not the first time I've been burned by cheep caps.

I think your theory is correct about the partial (possibly intermittent) short.

[firemedic]

Okay, this amp is a monster again! It seemed a bit lifeless at that last practice and I think that crappy bypass cap was gimping my tone.
I also feel it's a bit safer with a slightly cooler bias since I can't directly test current, and it still sounds very warm & ballsy. Hooray Ampeg! Hooray me!

[Willabe]

Nice work.   

[eleventeen]

Congrats on not having to replace your 7591s!

[firemedic]

Yeah, whew. They're the new EH 7591s I think manufactured in 2001. They sound way better than the original tubes which were slap worn out by the tiny original OT.