AA964 Build Issue

[rawnster]

Need some extra troubleshooting brain power here. Forum member Atmospheric recently posted about his new aa964 build. It had been running nicely for a while, but then developed an intermittent sputtering/motor-boating symptom. Thinking it was the recto tube, he swapped it out, and the problem went away for a while but then came back after a few hours. After telling me about the problem I went to his house to see it first-hand. It started motor-boating again, for which I quickly looked in the back and saw one of the power tubes red-plating (glowing like a mutha). I threw it into standby for a bit, turned it back on and the issue disappeared again. My first thought that there’s probably a cold solder joint on that power tube socket, or some other discrepancy on that socket. So I refloated all of the connections on that socket. I fired it up and it sounded great with no issues. That took place this past weekend. Today I get the following email from Atmospheric:

That noise/motor-boating problem came back then went away. That one socket didn't seem to look any different than the other tube, so maybe that's NOT the problem. So I got those NOS tubes. I thought I would try an RCA duet (thinking that maybe these Phillips tubes ARE bad). I plugged them in, hooked up my DMM to the bias points, flipped it off Standby and heard a pop, then no sound. Bias test points now read +20 V (not mV, V). I'm bummed. I hope that I didn't fry a tranny (probably not). Maybe this noise problem has been in the bias circuit all along.

So, I’m not sure what’s going on here. Any help would be much appreciated. Thanks!

Here's the only pic and schematic I have of the build. The tube that was glowing came from the middle octal socket in the pic.

[martin manning]

Rawnster, is that a bias voltage of +20V from ground? I'd check the whole bias circuit AC and DC portions, with tubes out... maybe the bias rectifier has failed?

[jjman]

"AA964" era can be a Princeton or a Vibrolux Reverb. The schematic shown appears to be a combination of parts of each. 3 preamp stages appears to be a Fender reverb channel, sans reverb. The PI appears to be Princeton type.

Bias test points are not shown in the scheme so it's not clear if they are current sensing and if they are 1ohm.

[Atmospheric]

@martin: With my DMM at the bias test points, power on and Standby off, voltage starts around zero and gradually gets more negative. With Standby on, it reads ~20V DC. At no time is there any appreciable AC present (seems like bias diode is rectifying).

@jjman: This build is a Princeton derivative.

The 6V6 cathodes are tied together at one end of a 1 ohm resistor, the other end of the 1 ohm resistor is grounded. Bias voltage is read across the resistor. We've been biasing in the range of 32-42mV.

Here's an interesting data point, while I was measuring just now, I accidentally flipped my DMM past VAC to mA mode and I got hum out of the amp (makes sense... cathodes are grounding through the meter in mA mode). So it seems as if that 1 ohm resistor has opened up to ground and (whew) the OT is not fried.

But why would that 1 ohm resistor fail?

[KT66]

Looks like there is just one bias test point - both cathodes tied to one 1R resistor. I would change it out to 2, it's only a couple of parts and might make diagnosing issues like these easier.

20V at the bias test point? That measurement must be wrong because that suggests that the output tubes are drawing 20 amps or 400 watts of power across a 1W resistor.

Recently I turned on an amp and when I took it off of standby POP went the brand new 5U4. It does sound like something in the bias circuit ( or bias measurement circuit ) is not right - maybe that created excessive current draw and it toasted the recto.

[Randy Magee]

I worked on a '68 Princeton Reverb recently that had a failed diode in the bias circuit. The amp was non-working when it came in, had all the original electrolytics and even the original tubes. I serviced the amp and powered it up and both of the new 6V6 tubes I'd put in there were getting real hot very quickly. I started checking voltage and got +5V on the bias circuit. I shut it down and replaced the diode (everything else had been replaced in the bias circuit) and the voltage in the bias circuit then measured -34V and everything calmed down. That was a first for me... I'd never worked on an amp that had a failed diode in the bias circuit before...

[rawnster]

"AA964" era can be a Princeton or a Vibrolux Reverb. The schematic shown appears to be a combination of parts of each. 3 preamp stages appears to be a Fender reverb channel, sans reverb. The PI appears to be Princeton type.

Bias test points are not shown in the scheme so it's not clear if they are current sensing and if they are 1ohm.

Hey JJ, I uploaded an updated schematic to show the bias tap connections.

[rawnster]

Thanks for the replies. I'll see if Atmospheric will want to add the second bias tap and thus split out the power tubes to have their own bias test point with discrete 1ohm resistors. Might as well change out the diode while I'm at it.

[jjman]

If the current sensing resistor has gone out of spec, the voltage reading on it would also go wacky. 2 resistors allow for checking the tubes for matching. I think I use 3 watters but 1 or 2 watter are probably ok.

[Atmospheric]

We replaced the bias tap resistor and the diode and 100uf cap on the bias board.

Thanks one and all for the help. Much appreciated.

[rawnster]

We replaced the bias tap resistor and the diode and 100uf cap on the bias board.

Thanks one and all for the help. Much appreciated.

Yes, this is back up and running and sounding stellar as expected. This is a great little circuit. Eats pedals too! Barber Burn Unit sounds great through it. Nice. I'm definitely going to build one for myself. I've got too many brit-sounding amps around the house.

[marcoloco961]

Need some extra troubleshooting brain power here. Forum member Atmospheric recently posted about his new aa964 build. It had been running nicely for a while, but then developed an intermittent sputtering/motor-boating symptom. Thinking it was the recto tube, he swapped it out, and the problem went away for a while but then came back after a few hours. After telling me about the problem I went to his house to see it first-hand. It started motor-boating again, for which I quickly looked in the back and saw one of the power tubes red-plating (glowing like a mutha). I threw it into standby for a bit, turned it back on and the issue disappeared again. My first thought that there’s probably a cold solder joint on that power tube socket, or some other discrepancy on that socket. So I refloated all of the connections on that socket. I fired it up and it sounded great with no issues. That took place this past weekend. Today I get the following email from Atmospheric:

That noise/motor-boating problem came back then went away. That one socket didn't seem to look any different than the other tube, so maybe that's NOT the problem. So I got those NOS tubes. I thought I would try an RCA duet (thinking that maybe these Phillips tubes ARE bad). I plugged them in, hooked up my DMM to the bias points, flipped it off Standby and heard a pop, then no sound. Bias test points now read +20 V (not mV, V). I'm bummed. I hope that I didn't fry a tranny (probably not). Maybe this noise problem has been in the bias circuit all along.

So, I’m not sure what’s going on here. Any help would be much appreciated. Thanks!

Here's the only pic and schematic I have of the build. The tube that was glowing came from the middle octal socket in the pic.

Quick question about this schematic. There is a different RC circuit in between the second and third gain stages. It is a 3.3M resistor in parallel with a 10pf cap right before what would be the grid leak in a normal situation. Is that seen as a voltage divider AND a high pass filter?

The 3.3M would normally kill all highs if seen as a grid stopper, but it looks more like a voltage divider attenuating over 90% of the signal. If so, does the cap still act as a high pass??