Vibration Induced bias runaway?

[Dyerseve726]

One of my Princeton clones got brought back to me today after only a year or so. Guy says it's crackling. Fired it up and all seemed alright. Played on it for a little while, turned it up a bit, then it started. Checked the back, one of the 6V6's was red plating. Pulled the chassis and tested bias circuit. Good negative voltage on the control grid. All solder connections on the sockets look good.

Connected my bias sockets. Both tubes start out about the same, then one tube slowly starts creeping up and up. What's weird is that a light tap makes the current jump around like crazy and usually increases the rate of runaway. One of the times I tapped it, it almost immediately jumped up to 100ma and red plated. Luckily, with a tube swap, the problem followed the tube, not the socket, so looks like it's just an early tube failure. However, I'm a geek and curious why this might happen? Something shorting out in the tube itself? Is there anything else I should test just to be 100% sure it's just a tube issue?

You can see in the video how the tube on the right meter goes crazy with a light tap. It's also super microphonic, so the sound you hear is coming out of the speaker.

[pdf64]

Yes, it seems reasonable to regard  the root cause as being a bad valve, eg grid structure not adequately held in place, leading to transconductance being unstable.

[tubeswell]

If the tube has overheated previously, it can mean that the internal structural integrity of the electrode cage has been comprised, e.g. from plates melting and deforming or grid support rods melting and bending etc, with the distances between the grids and other electrodes decreasing, leading to irreversible changes in gm and tube conduction.

[Dyerseve726]

Thanks guys! I wish I had some spares lying around. Getting my hands on a fresh pair of 6V6's is not gonna be fun.

[Dyerseve726]

Well, something still isn't right. I got new tubes and biased them up. Everything seemed normal enough. Lightly tapping the tubes didn't cause any crackling or crazy bias jumps. Hooked it up and played for a while. Hit some low notes and got a crackle and saw a burst of light in the left tube. This is a combo so the speaker is right next to the tubes. Turned off the amp and swapped tube positions. The right side tube had a much tighter fit in both sockets, so now the tighter tube was in the left socket. I no longer got loud crackling or bursts of light, however, I was getting some unusual sizzle type distortion on high notes. It did not sound like normal tube overdrive. And it seemed to come and go. I'm thinking that left socket isn't making good contact on one of the pins. The tighter fitting tube made the problem less apparent, but not entirely resolved. That's my guess. Wondering if I could mess with those socket pins or if I should just order a new set of sockets. It's the only thing that makes sense. I chop sticked all the pins on the inside, tapping every connection and wire. The only thing that causes the problem seems to be vibrating the tubes. I really hope I haven't destroyed these tubes already :(

[PRR]

https://el34world.com/Forum/index.php?topic=28750.msg316448#msg316448

[Dyerseve726]

Got it back on the bench and it's still acting weird. Had my bias sockets on again. Had 19.5 and 15.5 mA (not a great match but ok, from my understanding) on the tubes and went to take a quick plate voltage reading (407VDC). These are 6V6GT, BTW. I normally have the test leads clipped in place before firing up, but I forgot so just held a lead on the ground bolt and touched the other on the plate pin. Not anything I haven't done before. Speaker made a light pop when the lead touched, so I removed the lead after I saw the reading. It then made an extended crackling (not very loud), and the 15.5 mA tube climbed to 17.5 mA and stayed there. Other tube stayed at 19.5 mA. I left the amp on for 15 minutes or so and it was quiet and the bias did not fluctuate on either tube after that. Wiggling them in the sockets, light taps, and taking additional readings off the plate pin had no effect. I'm feeling awfully frustrated and confused, so I just ordered some sockets on Amazon. I'm going to replace both sockets as well as the grid stop and screen resistors attached to them. If that doesn't take care of it, I'm gonna lose my mind, lol.

[shooter]

add some 1-ohm resistors, cathode to ground, so you can eliminate the probes being part of the problem.

[Dyerseve726]

add some 1-ohm resistors, cathode to ground, so you can eliminate the probes being part of the problem.

I'll probably do that when I install the new sockets, but I don't think the test sockets are the issue, because during the playing tests I obviously remove them, and still having problems.

[mresistor]

I find it easier and probably safer to use the spring-loaded test leads so you can clip the test lead on then turn amp on/off doing measurements.  Of course a couple (or three) meters is handy for this too.


these   https://www.ebay.com/itm/10-Pack-of-Red-Spring-Loaded-Mini-Hooks-for-Building-Test-Leads/123814655107

[Dyerseve726]

I find it easier and probably safer to use the spring-loaded test leads so you can clip the test lead on then turn amp on/off doing measurements.  Of course a couple (or three) meters is handy for this too.


these   https://www.ebay.com/itm/10-Pack-of-Red-Spring-Loaded-Mini-Hooks-for-Building-Test-Leads/123814655107

That's usually how I do it also and do have several meters that I use at once. I just forgot, and the plate pin is easily accessible, so no big deal unless I've had too many cups of coffee, haha. That faint crackling and current rise might have just happened at that moment by coincidence. It was actually V5 that changed bias while I was measuring plate voltage on V6. Nothing adds up, lol. It's presented me with all kinds of unusual/confusing behavior and all the components and wiring look and test how they should 

[dude]

Did you read the link PRR suggested..? Takes a few minutes to re-tension the socket pins.

[Dyerseve726]

Did you read the link PRR suggested..? Takes a few minutes to re-tension the socket pins.

They're not the circular clamp style. I don't know how to describe the way the contacts work. I'm not sure how to go about re-tensioning these contacts. Here's a picture. Just seems to be two opposing pins, but they're set super deep inside the socket. I couldn't find anything long, pointy, and strong enough to get down in there and bend them. Maybe I should switch to the other style. Seems like clamp style would make better contact.

[dude]

Those are one of the best sockets you can get. Try an ice pick, spray some D-5 too, could just be dirty sockets.

[sluckey]

Everyone needs a jeweler's screwdriver set...

     https://www.amazon.com/Precision-Screwdriver-XAQISHIRE-Magnetic-Electronics/dp/B096XB6DHP/

[Dyerseve726]

Those are one of the best sockets you can get. Try an ice pick, spray some D-5 too, could just be dirty sockets.

I'll try some contact cleaner before I install the new sockets. The amp is only like a year and a half old, so they're probably pretty clean still, but worth a shot. I guess it's just as likely they're dirty as loose in such a short time. I might change the sockets either way, just for peace of mind. The amp belongs to a gigging musician and tubes are so crazy expensive right now, so I'd rather be 100% sure the problem won't come right back.

[shooter]

a pair of long "needle-nose" pliers re-worked on a grinder comes in handy also, just make sure powers off before using