Reverb issues on princeton build

[rutledj]

Posted this over on amp garage but got no replies. Hoping you gurus can help. I did a "retro" build of a princeton reverb. Everything works perfectly unless I try turning the reverb up. When I do, the volume decreases quite a bit (more than half). Reverb does work, just lose a lot of volume.I have checked and rechecked wiring and components. Replace tubes a couple of times. Does it whether tank is plugged in or not. Tried replacing tank and calbes. I've attached pics.

This one is a real puzzler. Can anyone tell me what kind of signal level I should see coming back from the reverb tank compared to the tank input? Also, shouldn't the signal out of the reverb recovery tube be in phase with the signal out of the 2nd preamp tube (input of 3.3 meg resistor)? Mine appears to be out of phase.If I disconnect the tank and jumper across the reverb jacks, there is no loss of signal. Really confused on this one.

[rutledj]

Another pic

[rutledj]

Another

[rutledj]

another

[dogburn]

What is the code/model for your reverb pan? Does it have the right input & output impedances? 

[stratomaster]

Pull up the 65 Reissue Princeton Reverb schematic. It has a test scenario and test voltages. The ones in the oval are signal (AC).  That will be a good guide despite the circuit differences.

Also, it looks like you're missing the dry input to the 3.3M/10pF network--unless it's an under board connection, but that one looks like it's going to the 470k resistor off the reverb wiper--which would be incorrect and would explain why the reverb is acting as a master volume control.

[rutledj]

This schematic seems to indicate the reverb wiper is connected to the 470k resistor. There is an under-board wire for the mixer input. I'll check out the 65 reissue for ac levels. Thanks

[rutledj]

I've tried both a MOD 4AB3C1B and accutronics of same model. I took some voltage measurements using the 65 schematic as a guide. I have about 100vac on the plate of the reverb driver but only about 52mv on grid of reverb recovery tube. The schematic shows it should be around 1.7v. 

DC voltages look correct on all tubes

[SEL49]

about 52mv on grid of reverb recovery tube. The schematic shows it should be around 1.7v. 

That 1.7V is a schematic error. There will only be a few mV signal coming out of the reverb tank. Your 52mV reading looks fine. Look at some other reissue schematics to verify.

Add a 220K resistor between V3 pin 7 and ground as seen in the attached pic. This makes the reverb circuit look just like all the AB763 circuits. Does this help?

[rutledj]

The 220k did help some. The drop off in volume isn't nearly as much but it does introduce some distortion at low volumes that I wouldn't expect.

[stratomaster]

Something else is happening. You shouldn't need that added resistor.

What happens if you disconnect the wiper of the reverb control? Does the amp behave normally? If so, reconnect the wiper and disconnect the lug connected to the reverb recovery stage. Turn the reverb knob.  Does the volume fluctuate significantly with just the reverb control connected? If so, then something is up with the pot itself.  If not, then there's an oscillation or something in the reverb circuit that is interfering.

[rutledj]

Ok. Just tried that. The reverb pot has no effect on anything with each terminal wire disconnected one at a time. Tried moving wires around to no effect. Very odd. Wonder if the reverb driver transformer could be defective.

[SEL49]

What happens if you disconnect the wiper of the reverb control? Does the amp behave normally?

Don't do that. The grid (pin 7) gets its ground reference through the reverb pot. The tube will misbehave without that ground reference. I believe that Fender included that 220K later on to ensure that the grid has a dedicated ground reference.

Try this... Measure resistance between V3 pin 7 and chassis. With the reverb pot set to zero, you should measure 470K. With the pot set on 10 you should measure 570K. Should also vary smoothly from 470K to 570K as you turn the pot.

The problem almost has to be in the blue highlight on this pic. Maybe a missing underboard jumper, or a jumper or wire connected to the wrong eyelet. Take a close look and use your meter to verify jumpers are connected correctly. Nothing to the left of V3 pin 2 can cause the volume drop symptom you posted.

One more thing to try... Swap the RT red and blue primary leads ---OR--- the green and black secondary leads, but not both.

[rutledj]

I am getting the same readings you suggested. 470k with reverb off, 569k with it full on. Tried switching the transformer green and black wires and it made no difference. Rechecked wiring but just don't see anything out of place.

[stratomaster]

Ok. Just tried that. The reverb pot has no effect on anything with each terminal wire disconnected one at a time. Tried moving wires around to no effect. Very odd. Wonder if the reverb driver transformer could be defective.

Cool. Now we're getting somewhere.  Is you reverb footswitch wired as shown in the schematic?  If so, does switching the reverb out (grounding the recovery stage grid) eliminate the behavior of the reverb pot affecting volume or does it only silence the reverb sound?

[rutledj]

It only silences the reverb.

[stratomaster]

It only silences the reverb.

To clarify with the reverb return disabled via the footswitch the Reverb control still changes the dry signal volume?

[rutledj]

Yes. Just checked it again. The reverb pot still loads down the signal with the reverb disabled via the footswitch.

[stratomaster]

Do you have a large (0.022μF+ @ 600v) cap handy? If so, put it across the 100k plate resistor of the reverb recovery triode.  This functionally grounds the plate from an AC perspective.  Then retest.  If the control loads down the signal still, then you've ruled out everything except the 0.003μF cap and reverb pot.

[AlNewman]

I did a "retro" build of a princeton reverb.

What does that mean?  Do you have an as built schematic?

[rutledj]

ALNewman - I just mean I used the original layout and brass plate grounding scheme. Still followed the AA1164 schematic.


stratomaster, The cap across the plate resistor changed nothing. I went ahead and swapped out the .003 cap and 100kl pot. Neither made any difference.

So strange that the amp sounds so good (even with lower volume with reverb on) and has this issue.

[AlNewman]

ALNewman - I just mean I used the original layout and brass plate grounding scheme. Still followed the AA1164 schematic.


stratomaster, The cap across the plate resistor changed nothing. I went ahead and swapped out the .003 cap and 100kl pot. Neither made any difference.

So strange that the amp sounds so good (even with lower volume with reverb on) and has this issue.

Maybe it's just a layout issue.  Oscillations and such.

[stratomaster]

stratomaster, The cap across the plate resistor changed nothing. I went ahead and swapped out the .003 cap and 100kl pot. Neither made any difference.

Ok, now I'm very intrigued.  I think a voltage chart is the next step. I'm anticipating this may be a ground scheme or improper D node decoupling issue, but we'll need to rule out the obvious before we start considering the less likely.  The voltage chart would be the best step towards that.

[rutledj]

I probably need to find some software that would allow me to edit (type on) a schematic. Any suggestions for that?

[shooter]

if you didn't fall for the life of ease n convenience,  MS paint works well 

[rutledj]

I measured the ones all around the reverb and preamp.

[stratomaster]

Can you repeat these with the reverb knob up high enough to cause the problem you're experiencing?

I promise we're going somewhere with this.

[SEL49]

Have you fixed the pin 1 to pin 6 jumper as suggested by xtian on the TAG forum?

     https://ampgarage.com/forum/viewtopic.php?p=479284#p479284

[rutledj]

I'll try to get that later this evening.


SEL49, Not yet. Since it does it even when the reverb footswitch turning reverb off, I figured it was bypassing that issue, but I can still try it.

[stratomaster]

That jumper can help with issues, just not this one. There are thousands of reissues running just fine without that jumper going around the socket, though it's definitely preferred to do it as recommended.

[rutledj]

Ok. Rewired the driver tube to add that loop around 1 & 6. No change. Took voltages with lower volume issue. 


Also loading an image of my scope. Top waveform is on the input of the driver tube. Bottom waveform is on the .003uf cap. As you can see, they are out of phase. Shouldn't they be in phase at this point? It should get flipped going through the driver tube and then again through the recovery tube. Of course, that doesn't explain why they issue persists with the reverb grounded out.

[tubeswell]

Top waveform is on the input of the driver tube. Bottom waveform is on the .003uf cap. As you can see, they are out of phase.

If you mean there's a difference in phase between the grid of the 12AT7 reverb driver vs the plate/output of the reverb recovery stage - there's no requirement for the phases before and after the reverb pan to be in phase. All sorts of phase weirdness happens in those springs

[passaloutre]

I don’t think you can infer any phase relationships when one of the signals has experienced a time delay. I think it’s just a coincidence that they look 180 out of phase.

You could replace the reverb tank with a resistor (or a pot) if you really need to see the relative phases.

[dogburn]

Isn't the phase is inverted when it goes through the recovery triode? Which would mean it's doing what it's supposed to do.

[tubeswell]

Isn't the phase is inverted when it goes through the recovery triode? Which would mean it's doing what it's supposed to do.

Yes but it appears we're not talking about signal inversion between the grid and the plate on the same triode. rutledj seemed to be saying that the two sets of scoped points were the grid of the 12AT7 reverb driver, and the output from the plate of the reverb recovery stage (and I asked rutledj to clarify that this is what he was measuring)

[stratomaster]

The waveform at the output of the recovery triode should be smeared and chaotic.  If you somehow have dry signal going through the reverb stage unaffected, and your transformer is wired incorrectly,  then you'll get destructive interference that reduces volume with increased Reverb levels--just as you've observed.  I think you may have given us the biggest clue yet.  To confirm, look at the signal at the input side of the 3.3M resistor vs the output side as you rotate the Reverb control.

The problem with this theory is that the behavior is the same with the reverb grid grounded.

[rutledj]

I'll check that signal when I have a few minutes from work. Just to clarify, I was saying that from the grid of the reverb driver to the plate, the phase should be 180 degrees different from grid to plate. The reverb recovery tube would then introduce another 180 degrees phase shift, putting it back in phase with the dry signal passing through the 3.3m mixing resistor. Therefore, they wouldn't cancel each other out (to some degree).


But again, doesn't explain the behavior when the reverb signal is grounded out.

[rutledj]

Here are the pics. Top waveform is in front of 3.3m, bottom on opposite side.
First image is with reverb on, second with it off.

For some reason the pics are labeled wrong. The one with the reduced amplitude sine wave on bottom is with reverb on.

[stratomaster]

This is very useful as you now have a visual indicator of the interference.  You can ground grids, outputs of coupling caps, or bypass plates and see how much that affects the issue. I'd encourage you to repeat the grounding of the reverb return grid to confirm how it affects the issue, then ground the input to the pot (output of 0.003 cap) and see how it affects the issue.

[rutledj]

Finally found it. As much as I hate to admit it, it was a wiring error. I had the cathodes of the reverb recovery tube\preamp 3 were reversed. Going to the wrong set of cathode cap\resistors. I assume that was causing the out of phase issue. All is well now.


Thanks much for the help here.