Suggestions for Boogie Trem-O-Verb Problem?

[Steve L]

I have a Mesa/Boogie Dual-Rectifier Trem-O-Verb head on the bench at the moment that is proving to be quite a challenge.

Here is the issue: The bottom half of the output is not present when the amp is turned on. After 3 minutes or so, the output will then be fine until another couple of minutes and then the bottom portion of the output will disappear again. Might come back, might not.

Here are the details: The signal is clean up to the grids of the output tubes. If I put a meter or scope lead on pin 2 of the phase inverter, the output instantly returns to normal, take it away and the bottom half is missing again.

Tried a new set of output tubes, even tried another set. New phase inverter, tried several of those, as well as all of the other preamp tubes. Voltages around the circuit look good, except that the plate voltage of the phase inverter looks high (pin 6). When the amp is not working properly, the plate voltage will be about 415V, when the amp does work properly, the voltage drops to about 350V. Output tube voltages, including bias voltage all look good. Power amp feedback loop appears to be OK.

I have changed every part in the phase inverter ciruit. (1M, .1, .047, 120pf, 754pf, 82K, 90K, etc., socket tests good. Freeze spray does nothing to locate anything.

I have talked to Boogie; they were no help.

Any ideas?

Thanks,

Steve

[Chuck H]

perhaps it's just a timing issue with the testing, but the voltage on the PI is probably rising because one half of the power tubes are dropping out and reducing overall current draw. With an intermittent problem like this it's hard to find out if it's the tubes, sockets, solder joints or even the OT. You have to be testing the right thing when the problem is happening.

I suspect it's either a cold solder joint or the OT. If the amp has internal fuses for the B+ rail I would check those for consistent continuity. I've read that those amps have trouble with bad traces on the board too. But I'd bet that the basic problem is that voltage is dropping off of half the power amp intermittently for some reason or the cathodes of half the power amp is going open intermittently.

Chuck

[chipaudette]

Looking at the Trem-O-Verb schematic that I could find, the plate voltage on the phase inverter should only be ~280V. The supply voltage should be ~415V, but the plate voltage should only be 280V.

If you're really measureing 415 working or 380 non-working, something sounds wrong in either case. Are you sure that you're really measuring what you think that you're measuring?

If you really are measuring 415-380 on the plates, what is the voltage of the supply on the other side of the plate resistor (ie, supply "C")?

Chip

[Steve L]

Hi guys. Thanks for the replies. First, yes those are the real voltages. I agree they do look wrong though.
I also suspected the output transformer, however having the signal come back when an impedance is put on pin 2 of the phase inverter seems to indicate otherwise, especially with the signal being intermittant. However you never know. I have seen transformers do some strange things before.

I really suspect a cold solder joint, however I can't seem to find one.
I will keep looking.

Thanks again,

Steve

[chipaudette]

To repeat my question below...what is the supply voltage to the PI?

If it is ~415V, then you know that the tube is not conducting. This means that either: (1) you have lost continuity on the plate side or on the cathode side of the one or both sides of the PI tube, or (2) the bias on the PI grid is really screwed up and it has shut down the normal conduction through the tube.

If the supply is higher than 415V, then you know that current is flowing through the tube. That's good. This means that you at least have continuity. Other than that conclusion, it'll take some more thought...

Chip

[Steve L]

Chip,

Thanks for the replies. The voltage at the power supply, point "C", is 417V. The voltage on pin 6 of the phase inverter is 415V, so, barely conductiong.
The voltage at pin 1 of the phase inverter is 265V, conducting fairly well.

I believe that the bias of the tube is the problem, especially since placing a meter lead on pin 2 makes everything better.
When I measure the voltage on pin 2 it is roughly 20V. I suspect that it is different with the meter probe off it, however I cannot tell what it is.
Everything checks fine without power. My next step is to place a few short clip leads between some places where I think that there could be an intermittent open and see what happens.
By the way, when the signal returns to normal on its own, it comes back slowly.
Any other suggestions?

Thanks again,

Steve

[chipaudette]

Wow, this is some good info. I definitely like your theory that the bias is messaed up. Even at 20V (which, as you said, is probably corrupted by the measurement process itself), the grid voltage is too high. I'm no expert, but isn't the point of cathode biased tubes is that the grid is near 0V ?.

So, if it is supposed to be 0V, then your grid doesn't have a path to ground (or an intermittent one). I would suggest that turn off the amp and pull the tube. Then, in the amp, measure the resistance from each grid pin socket to ground. Of course, this resistance should be kinda high (1-3M ?) but it shouldn't be infinite. Check it out.

Good luck!

Chip