Problem with DR reverb driver

[luthierwnc]

Hi All;

This is my first post here but I recognize some of you from the Ampgarage and maybe even one from the old alt.guitar.amps newsgroup.

At any rate, I pulled a project out of the rack this weekend and got stumped. It is a homebrewed Deluxe Reverb. It only has one channel and no trem. Other than that, it is very close to the schematic in resistor and cap values. My problem is that the reverb driver tube is running away and I can't tell what is doing it. Simply plugging in the tube drops the B+ 11 volts. It is a known-good 12AT7. I swapped the inverter tube and got the same result. I am measuring 91 VDC on the cathode even with the input to the grid unsoldered. On the schematic is around 9v. For laughs I threw a 5751 in the hole and got 33 VDC and the same drop in plate voltage. A 1k sine wave goes into the tube very clean but comes out pretty nasty. Doesn't look like a parasitic.

I pulled the supply from the screen node. This amp is slightly overcapped at 40/20/20/20 but I wouldn't think enough to cause any problems. The reverb is overwhelming even at 2 on the dial. I have a voltage divider on the input grid at 220k/750k to take a little of the hair off it. 3-spring Accutronics 8 ohm tank. The tube has the stock 2k2/25 cathode set-up and it measures fine. Tube socket pins have no continuity.

Has anyone seen this before? I do have a spare reverb transformer but thought I'd try the gray matter first.

Thanks, Skip

[Chuck H]

the reverb driver tube is running away and I can't tell what is doing it. Simply plugging in the tube drops the B+ 11 volts.

Drops ALL the B+ 11 volts? Or just the B+ on the driver plate socket pin? The B+ SHOULD drop at the socket pin with a tube intalled because the tube draws current.

even with the input to the grid unsoldered.

If the input grid has no reference to ground the tube will always runaway. It needs the ground reference for bias.

A 1k sine wave goes into the tube very clean but comes out pretty nasty.

How do you know it's going into the tube clean??? The grid of a tube has a high impedance. If your checking the signal at the grid pin you are seeing the whats at the door. Not whats gone through it. Because you've had this problem with multiple tubes I would suspect the circuitry or the tube socket. If the tube socket is dirty, and not making good contact with the grid, this could cause runaway. Also, if the ground reference for the input grid is wrong you could have runaway or incorrect bias.

The reverb is overwhelming even at 2 on the dial. I have a voltage divider on the input grid at 220k/750k to take a little of the hair off it.

That shouldn't be needed. Are you sure the rest of the amps signal is getting through the amp OK? If the circuitry is correct there is no reason for the reverb to be too loud or disproportionate with the mix.

Are you doing your output testing for the driver at the recovery triode? That would include the recovery grid, cathode and mixing pot in the equation.


Chuck

[pdf64]

Like Chuck says, the grid needs a dc path to ground for bias reference, otherwise it's just going to float.
Check the ground path for the cathode resistor and cap, 91V there is nuts. They would surely need replacing if they really have had that voltage across them.
I can't see how the reverb transformer could cause this. Peter.

[luthierwnc]

Hi Chuck;

It is a problem with the whole rail. Everything I gave you was with no signal (except to check the wave form in the preamp). I did it with the grid intact, disconnected and grounded with the same result. Last night I pulled the tube and did some tests with signal. The B+ dropped from 429 to 384 by turning up the volume to 2. Draw per tube was biased at 25 ma and it shot to 70. The reverb tube must have been picking up something with no signal that the rest of the amp gets as it is turned up. The waveform is still perfect through the first tube. Looks the same with or without the reverb cables attached.

I'm going to shield a couple of signal wires to start and then look for parasitics. I'll report back when I get something useful.

Thanks, sh

[Chuck H]

Well I can't say for sure without poking at it. But it sounds like an ultrasonic parasitic oscillation. When you turn up the amp a feedback you can't hear is making the amp work overtime and draw current. This could be why the PT is getting hot and why the voltage on the rail is dropping. If your problem is that extreme I don't think a couple of shielded cables will fix it well enough. This is not a really high gain amp. You may want to re-examine the grounding scheme and layout.

[Enzo]

Take a clip lead and ground the grid of that tube and see what happens.

[luthierwnc]

Thanks guys,

I think it is just a weak transformer. The no load B+ is 436. With both power tubes biased at 25 ma each and no preamp tubes, it drops to 413 VDC. With the preamp tubes it is down to 411ish. A dimed Strat power chord at about 7on the amp volume will take the B+ to 394 before it recovers. And I replaced the 5AR4 with a solid state rectifier. When I was running a 1k sine wave in (8 ohm dummy load) with not a lot of gain it took the B+ consistently down to around 384.

To get to the heart of it, I put a working amp side-by-side on the bench, yanked the power tubes and inverter and jumpered the preamp signal over to the subject amp. The volts were the same. Not a parasitic.

It sounds like a low voltage amp should. No weird squeaks or hums. Can't say much for the bottom end and not much headroom -- but not bad. It just isn't a DR. Figuring a 7% boost from the 115 wall voltage of the day with my 122 I should be getting around 425 with a tube rectifier. Using the diode string bit more.

You never want it to be the most expensive part so I wanked around enough to get to the bottom of it. There is a 50 watt OT in there so If I replace the power trannie I'll probably just get a matching size and use a pair of 5881's.

Hey, another project.

Thanks again, Skip