screen and plate voltage

[Jazz P Bass]

You could try the tests from the DR RI schematic.

There is a known input signal level & all other voltages & signal levels come from that.

[Old Tele man]

Remember, guitar amplifiers are designed to & tested assuming (a) 1,000 Hz sinewave input into (b) a resistive load at (c) 5% (or less) THD distortion.

[Randall]

I wondered about that RMS input voltage. But that puts me back where I started. I realized from the beginning the amp wasn't clipping, so I turned up the input signal with one channel volume wide open until I saw clipping at the output load and measured power there but was told that was wrong. Otherwise how do I know how much input signal it takes to clip this amp? I can now tell you it is approx 176 mV RMS, which leaves me the same as before, 14 watts at the reverb channel, 19 watts at the normal. This time with tone controls at 7.

One more thing to clarify, when you say clipping, you do mean measuring at output, correct?

[Chuck H]

Yes, clipping is measured at the output.

I'm baffled, honestly. I can't even imagine that a DR with the tones fived and 200mV at the input (even peak 2) won't be clipping with the volume past six or seven. Something is wrong with something. Either the test, the amp or the measurements. My next move in your position would be to reevaluate everything again and again until I have eliminated the impossible so that I was only left with what IS. Because something is wrong.

[dstrat]

Have you tried to feed each channel separate because I'm with Chuck , there's something very wrong there.
I keep thinking feeding both channels your loading your generator too much throwing off your input.

the DR RI shows 37mv to drive to full power which I am inclined to believe.

Edit: Unless you got 680K resistors on your input.

[J M Fahey]

I wondered about that RMS input voltage. But that puts me back where I started. I realized from the beginning the amp wasn't clipping, so I turned up the input signal with one channel volume wide open until I saw clipping at the output load and measured power there but was told that was wrong.

Not "wrong" but incomplete.
You feed signal through one channel, reach clipping and measure power out.
Then you feed the other channel and *again* drive it into clipping, remeasure.

Otherwise how do I know how much input signal it takes to clip this amp?

Please CHOOSE ONE.
Either you measure power output or measure sensitivity, those are different experiments.

I can now tell you it is approx 176 mV RMS, which leaves me the same as before, 14 watts at the reverb channel, 19 watts at the normal. This time with tone controls at 7.

Incomplete. See above.

One more thing to clarify, when you say clipping, you do mean measuring at output, correct?

Yes yes yes yes yes.

EDIT:
Yes, clipping is measured at the output.

I'm baffled, honestly. I can't even imagine that a DR with the tones fived and 200mV at the input (even peak 2) won't be clipping with the volume past six or seven. Something is wrong with something. Either the test, the amp or the measurements.

That.

there's something very wrong there.
.......
the DR RI shows 37mv to drive to full power which I am inclined to believe.

That.

[Randall]

This thread is going to explode my head.

May we take a step back and break this down a bit? I have a HP military signal generator, dual trace scope, 100w 8 ohm dummy load and an average RMS Fluke meter. My current configuration is I have normal channel only connected to the generator feeding a 1K signal @ 175 mV measured at input jack by one side of the scope before the 68K grid resistor. Tone controls on 7. At output I measure (given for bad eyes and margin of error) 35 vp-p just before clipping at about 9 on the volume knob. (35v/2) x .707 =12.37 v RMS. (12.37 x 12.37)/8 = 19.13 watts. My average RMS Fluke meter also connected to output reads 12.8 vac, maybe not accurate, but close enough to confirm my scope.

So let's say I'm happy with the power output at said 19.13 watts at this point, this confirms the power amp section, correct? But now looking at the fact that it takes 175mV at input to drive the power amp into clipping, would this suggest there is sensitivity issue, or loss of gain in this channel? If so, how to go about measuring this?

Same test on reverb channel, reverb and trem at 0. At 9 on the volume is where it begins to clip, but not in the same way. This time the top of the wave clips, but the bottom just sort of rounds out and never really 'clips', even at 10. Plus, I measure only 30 vp-p this time, which is by the same math 14 v RMS. So with this channel I have both a sensitivity and an output problem? Juan said "power amp clips at the same point no matter what", so this eludes me because this does not appear to be the case here.

[dstrat]

Hi Randall, My comment about the 680K resistor was hoping you did NOT have that value. If you do that would (Might) explain it.

Verify you have 68K input resistors, I say that because that mix up happens a lot.

[Randall]

Yes, they are 68K not 680K.

[eschertron]

Same test on reverb channel, reverb and trem at 0. At 9 on the volume is where it begins to clip, but not in the same way. This time the top of the wave clips, but the bottom just sort of rounds out and never really 'clips', even at 10. Plus, I measure only 30 vp-p this time, which is by the same math 14 v RMS. So with this channel I have both a sensitivity and an output problem? Juan said "power amp clips at the same point no matter what", so this eludes me because this does not appear to be the case here.

The shape of the waveform when it limits is a clue that the signal is being distorted before the power section. You are not getting a good sine wave to the output tubes. From what you've said, there seems to be a loss of gain on both channels, but this one even more so. Follow your test voltage and note RMS voltage before and after each gain stage. Find where the signal drops.

[pdf64]

I wonder if something about that 3rd stage plate circuit is wonky, hamstringing its output Vpac, eg wrong value plate resistor, coupling cap bad/wrong, trem intensity pot bad somehow?

[g1]

I wonder if something about that 3rd stage plate circuit is wonky

Same here.
Before doing AC checks, it's important to verify all DC tube voltages correct with no signal. The 65 DR reissue schematic should show ballpark DC voltages.

edit: I see you mentioned voltages correct in post #1, but I was unclear if that was for power tubes, power supply, or everything.

[Randall]

I can tell you that the signals at both channel preamp tubes second plate are identical.

[J M Fahey]

Same test on reverb channel, reverb and trem at 0. At 9 on the volume is where it begins to clip, but not in the same way. This time the top of the wave clips, but the bottom just sort of rounds out and never really 'clips', even at 10. Plus, I measure only 30 vp-p this time, which is by the same math 14 v RMS. So with this channel I have both a sensitivity and an output problem? Juan said "power amp clips at the same point no matter what", so this eludes me because this does not appear to be the case here.

Then you have a preamp problem.
Something is clipping inside it (or its output is attenuated) so it can not drive the power amp to clipping.
You must find what and where.

1 tip: repeat the experiment through the Reverb channel (the other one has proven it works fine so letīs set it aside forn the moment).
Rise volume until on speaker out "the top of the wave clips, but the bottom just sort of rounds out and never really 'clips', even at 10." and then, touching nothing, move scope clip from speaker out to PI input ... I bet you will find one signal peak is clipping too.

Please test just this and nothing else and post results.

I guess you have a preamp problem which makes it clip early on its own and does not let it fully drive the power amp.
Thanks.

[Randall]

I'm working from the original AB763 schenatic and can confirm that all voltages are within 20%. The reverb channel second plate is higher than the others almost to tolerance limit, and the PI cathode and resistor network are the only voltages lower than printed, and only by a few volts, but all else is good.