Fender Princeton Reverb II

[gbono]

I have a PR II which came to me with the complaint that it was making funny noises, cutting out and possibly smoking. All tubes check out on a tube tester, no visual of smoked/damaged components, no funny smells, plate voltages without 6V6GT plugged = +420V, power supply voltages also look normal and the amp works fine on the bench. Checked the ESR on all the capacitors and nothing remarkable. Tube bias are 20mA/26mA.

Chassis goes back into cabinet and I play through it for about 15 mins and when you play hard transient notes/chords the output gets distorted and drops out and then comes back. Very hard to repeat though. Reverb control doesn't seem to make the problem stop or start. The amp gets the rubber mallet test and there is not reaction - even when the problem is present.

I'm going to do:

Test the speaker in the cabinet with another amp - I did a quick impedance check on the driver and it looked normal
Swap out the 12AX7s one at a time with know good tubes.
Start chasing a signal (gated sinewave burst) through the preamp to power amp to see what stage is problematic.
?????

I don't work on much tube gear so bear with me.

[olddawg]

Not to be condescending, but have you tried a different speaker cab? An intermittently rubbing voice coil can cause what you are describing. Also a trick I used to use on intermitant pcbs was to vibrate them with an electronic engraver with a piece of hard rubber on the tip. Not so sure how that would affect sensitive tube elements though. It's a good way to isolate internally intermitant components. Freeze works too, but don't hit a tube with it. Might also want to scope the signal path at various points to find out where the problem starts. Do you have a load bank where you can cook the amp near saturation for a while and monitor a calibrated source tone at the output? It would be a lot easier to troubleshoot in a continuous fault mode.

[gbono]

Well after testing the driver with another amp I didn't find a problem until I noticed that the rivet that held the tag panel for the tinsel leads had sheared off. New rivet and one happy customer

[gbono]

Well I got back on this amplifier since it still has a "crackle" issue at higher volume levels. I'm going to replace the driver to eliminate that as a possible root cause. If that doesn't solve the problem, then it is back to looking at HV arcing on the tube sockets (this design has point to point wiring - no PCB). Has anyone ever looked for trace or component arcing with the amp in a darkened room - like they way you could find a bad ignition wire on a car?

[Enzo]

That is exactly how I recommend looking for arcs in a live amp.

[gbono]

So if I have the amp on the bench and play into a dummy load can you also "hear" the arc or should I use a loudspeaker as a load (amplify the arc)?

[Enzo]

If I can see it, it is a problem, whether it makes a noise or not.

[gbono]

I had an arcing power HV supply problem in a previous job/life that you could hear but not see. I used a plastic stethoscope to hunt down the location of the fault.

[Enzo]

That answers your own question then. Certainly I would never see an arc snapping inside a transformer. We have many senses, and we use them all. Most experienced techs, for example,recognize the different smell of burning resistor compared to burning transistor compared to a transformer, compared to a wire, and of course NOTHING smells like a burnt out selenium rectifier... NOTHING.

[gbono]

The issue now is that the output sounds ratty and distorted when the master volume is on "9" and the volume is at "4". With a dual trace scope on pin 1 of V3 and the second channel on the output (8 ohm driver connected) you can see the output clipping while the voltage out of V3 is not clipped. I was monitoring the DC voltage at 80uF filter cap and it goes from 404V (idle) to 385V under clipping. Is this normal operation for an amplifier like this (I don't do many tube amps) it appears that there is a good deal of "sag" in the supply and I wonder if the filter capacitors could need to be replaced?

[Jazz P Bass]

What is the max unclipped (clean) AC voltage at the output.
Princeton amps where never known for unbridled clean output.
I would assume this one is no different.
The power supply is not regulated, so there will be some sag.
19 volts does not seem all that extreme.
If the filter caps are good, as far as removing the ripple, I would let them be.

[gbono]

I get 12.5 VRMS at 1000 Hz and 8 ohms which is pretty close to the rated 20W. This amp seems to start out sounding clean and then gets dirtier over time. Believe me, I don't want to recap this amp - how do you change the caps on the point to point "fish paper" eyelet board? The pots are all grounded to a brass channel that's soldered to the chassis. Do you have to remove the tube sockets to get the wire board free of the chassis? I don't want to just solder on top of an eyelet.

[gbono]

After looking at this design and doing some web crawling it appears that the phase inverter in this design departs from the "old" Princeton Reverb cathodyne type PI. The PI in this design (PR II) is a Schmitt inverter. How do you check the "balance" in this type of design?

[gbono]

It looks like the PR II and the older Deluxe Reverb are very similar designs. Can anyone explain the difference in the feedback resistor choices in the PR II? I was able to control the output stage clipping (but not eliminate it) by using a 12AU7 in the PI. I have another novice question and that is why does the pentode bias have the same voltage on the plate and screen grid? The data sheet for the 6V6GT shows the plate and screen grid voltage of 285V.