Ampeg Reverberocket R-12-RB

[xtian]

I have a modded and much abused 1960s Ampeg Reverberocket R-12-RB on the bench. I have replaced all e-caps, and am now having an early-morning, am-I-really-awake? moment.

I am measuring 365v on the plates of the 7591 plates (two tubes in push-pull), 13.7v on the cathodes (single 155R cathode resistor).

I am inserting a 1kHz 1v p-p sine at the input, and it looks beautiful on the output--symmetrical and strong.

Too strong, in fact. I have triple...quadruple checked my readings. I am seeing 37v p-p before clipping, into an 8R dummy load. Now, remember I've been taking this reading on my amps over the past few weeks, and have never seen something like this: (37v x 0.707)^2 / 8R = 85 watts.

This can't be right. This amp has a Hammond 1620 OT (rated at 20 watts).

Am I supposed to divide the output p-p by 2? Then (37 x 0.707 x 0.5)^2 / 8 = 21 watts.

[xtian]

Weber's bias calc is telling me 15 watts idle diss per tube. Damn. Sounds hot. No signs of red plating.

[xtian]

Changed original cathode 150R to 220R. Now seeing a much more reasonable 33mA per tube, 11watts per tube.

[Stevem]

Your meter must be messed up on the ac scale, how does it read the on the primary side of the PT?
Did you blow a internal fuse in the meter?
By the way, be sure to replace that cathode resistor, as by now they are well used and anytime I have referberb one of those and not changed it out, it goes open in short order!
Step up to a 15 watter and the tubes idle much more stable when the amp gets cranked out.

[sluckey]

Am I supposed to divide the output p-p by 2? Then (37 x 0.707 x 0.5)^2 / 8 = 21 watts.

Yes. First convert p-p to peak. Then convert peak to RMS. Then plug into the power equation. Or, all in one step as you have done.

[xtian]

@Stevem: I did replace the cathode resistor, as mentioned. I'm using a scope, not a meter, for output power.

Thanks!

[xtian]

Am I supposed to divide the output p-p by 2? Then (37 x 0.707 x 0.5)^2 / 8 = 21 watts.

Yes. First convert p-p to peak. Then convert peak to RMS. Then plug into the power equation. Or, all in one step as you have done.

OK. Now, with the new cat R, I'm getting 27p-p, which is 11.4 watts clean, and about 22 watts total plate diss at idle.

[xtian]

Owner and I both hated the voice of this amp. It was extremely mid focused. Mid humped. Nasty. So I replaced the one-knob tone control with a dual concentric pot and a AA1164 tone stack. Now it sounds MUCH better, but this stack is much more lossy (10:1) and reveals that I'm losing a whole lot of amplification somewhere.

I traced the signal with my scope, and this is shown on the scan as red numbers in circles. One volt p-p at input. 30v after first stage, drops to 3v after tone stack (this matches the behavior of my working Princeton build).

This 3v signal hits V2, the lone 6SN7. Note the 22K resistor tying the cathodes of 1/2 V2 and 1/2 V3. This is where the "dry" signal is supposed to pass thru, while the "wet" signal goes to the reverb pan (currently unplugged) and I observed 20v headed headed out to the last stage before the reverb pan.

This is where my signal gets too small. I read 1.5v at the cathode of V2a, then less than a volt after the 22K resistor. I have only 1.6v or so going to the PI.

I have tried two 6SN7 tubes. No change. Having trouble thinking of the next step.

[xtian]

I've been thunking. Thunk.

I think the amp is working as it should. I introduced a new tone stack, much more lossy. there is no third gain stage, as in blackface topology. There is the first stage > tone stack > makeup stage > then V3a which looks like a unity gain mixer, then the PI.

Please look at V3a (has 47K Ra/5.6K Rk). Unity gain mixer?

[xtian]

Aw, headslap. Earlier, when repairing the output jack, I swapped it's leads. Repaired by Bubba(TM).

All better.

[Stevem]

Use another 6sl7 and rebias that stage.