Mesa Dual Rectifier Maverick -- DC on input

[xtian]

OK, getting into quantum entanglement, here. I pulled V1 AND V2, and I'm still seeing DC at the input. BUT! I turned the amp OFF, let the caps drain, and I'm still seeing DC, that sweeps slowly (60 seconds or more) between ~40mV and ~-40mV. So I now understand that I'm seeing stray 60Hz AC being picked up by the open wires, and both my meters are interpreting this as DC.

Notice that the Maverick has 2.5M resistance between the input tip and ground. This is probably why the stray voltage isn't shunted to ground. Most amps have 1M ground reference.

Not sure this helps with anything. I checked all the grounds (input jack, filter caps, pots), and they all read less than 1 ohm.

[Chuck H]

I guess I have to ask... When this started were you trying to solve a performance problem, scratchy pots, odd distortion artifact, etc., or was this just a puzzle you wanted to solve?

Why did you look for DC on the input in the first place?

[xtian]

Amp arrived with bad noise, so started troubleshooting, and noticed scratchy guitar vol pot, scratchy Volume pots on amp. You know how if it's a dirty pot, the noise is in one place on the carbon track, but if it's DC it's everywhere in its rotation? Anyway, that's what lead me to check for DC with multimeter.

I have swapped multiple tube sets. Tacked in extra 100uF at reservoir, and again at PI node with no change in hum. There was a dodgy area on the PCB that made the hum come and go when chopsticked--I ended up lifting the PCB (removing all the pots and jacks from the front panel is a long process) and retouched the solder connections that looked iffy. No help.

I'll call Mesa on Monday, in the hopes Rich says, yeah, those damn optocouplers are known to cause this issue, or something helpful.

[xtian]

The switching system sure seems a likely suspect for injecting noise into the circuit. And, in fact, I see a browned resistor in the 19vDC supply for the switching system. It measures within spec, however, 3.39K. The diodes also measure 0.5v forward drop. However, the test point is supposed (according to the schematic) to read 19v when in Lead mode, and it reads only 15v. So, maybe a failing filter cap in that LV circuit? But AC ripple is pretty small, only 0.1vAC. What do you think?

[Chuck H]

Glad to hear you're on a first name basis with a guy there. Mesa was always very helpful whenever I called them and I've been hearing that it's not like that anymore. So this is good news.

[g1]

The switching system sure seems a likely suspect for injecting noise into the circuit. And, in fact, I see a browned resistor in the 19vDC supply for the switching system. It measures within spec, however, 3.39K. The diodes also measure 0.5v forward drop. However, the test point is supposed (according to the schematic) to read 19v when in Lead mode, and it reads only 15v. So, maybe a failing filter cap in that LV circuit? But AC ripple is pretty small, only 0.1vAC. What do you think?


[ATTACH=CONFIG]56346[/ATTACH]

Also looks like the wrapper on that cap has shrunk (from heat probably)? I don't see it partially wrapping over the top like the other cap in the pic.

[Helmholtz]

I have tried a variety of tubes in V1...and the evidence is getting WEIRD.

I inserted a 12" instrument cable in the input jack and attached my multimeter to tip and sleeve. I turned the Output Level (master master volume) to zero and powered up. In Rhythm mode I see -12mv (thanks, @Helmholtz, for pointing out the importance of pos/neg). And in Lead mode, I see -170mV.

I tried several different tubes, and got similar readings.

Now the weird starts. I put the amp in standby while swapping V1...and I noticed I still had -ve DC on the input jack! And then I pulled V1 out, and STILL had -ve DC on the input! With no V1, in run mode, I see -15mV in Rhythm and -76mV in Lead, and in standby, I see -10 and -28mV.

Looking at the schematic, assuming no V1 is installed, wouldn't you have to suspect LDR1? I don't see any other path for stray potential.

What about the positive DCV? Can you still reproduce it?

[xtian]

Yes, I can reproduce the DC thing--I just replaced all the tubes and fired up. And instead of inserting an instrument cable, I just used a chopstick to open the input jack's switch (to help eliminate the stray DC I measured earlier). I'm seeing -80mVDC between tip and ground at the input jack.

[Helmholtz]

Yes, I can reproduce the DC thing--I just replaced all the tubes and fired up. And instead of inserting an instrument cable, I just used a chopstick to open the input jack's switch (to help eliminate the stray DC I measured earlier). I'm seeing -80mVDC between tip and ground at the input jack.

Sorry, this is negative DCV again. What about the original positive DCV?

A little negative DCV is normal. The tube datasheet recommends a 10M grid leak resistor for grid leak biasing. One can expect that the normal (negative) grid leak current produces around 1V bias across this 10M resistor. This means a typical grid leak current of -0.1ľA. So with a 1M grid leak it is normal to see something like -0.1V at the grid with open input. But if a guitar with a DCR of maybe 10k is connected to the input, the voltage at the input should drop to -1mV.