Ampeg V4 filament problem

[melvin]

So, I modified a V4 to handle KT88's. Right now the original filament winding handles the four KT88's and an auxiliary filament transformer handles the preamp tubes. The ground reference for the KT88 filament is on one of the sockets consisting of two 100 ohm resistors one from pin 2, one from 7, both to 8 (cathode) which is at ground potential. The KT88's are biased at about 30 mA each. The amp is fine until it is seriously cranked for a bit at which point it fries the two 100 ohm resistors.

Some other things to consider:

This is the second time this has happened and this is a whole new set of KT88's.

I have a fuse (6A slo-blow) installed on one leg of this filament winding that did not fail.

The path to ground on the cathodes consists of a single wire from pin 8 of one of the sockets to the ground path on the circuit board.

Anybody know what might be going on?

[Billy R aka DynaFreak]

bad grnd on cathodes? check the circuit board grounding. try grounding the 100 ohm resistors directly to chassis ground and see what happens. what watt rating are they?

[Rhodesplyr]

I agree with DynaFreak's recommendations, but this is a weird/interesting problem. The thing I'm wondering is that, if the KT88s are drawing cathode current through the 100 Ohm resistors connected as you've described, it sounds like there would have to be another ground reference somewhere for the filaments. A V4 schematic shows a hum balance pot. Is it still there, or did you remove it from the circuit? If the hum pot is still there, where is it grounded? The V4 has a slightly complicated grounding system, especially with the standby switch in the ground path and a diode between the rectifier bridge ground and the main audio grounds. How are the standby switch contacts?

Also, one important difference between 6550s/KT88s and 7027As is the metal shield around the KT88 tube base connected to Pin 1. Are you using Pin 1 as any kind of tie point and/or is there any contact between the metal band and the chassis? I have occasionally heard of problems with amps that use Pin 1 as a tie point when swapping in KT88s.

For a single 1/2W 100 Ohm resistor, it would take over 71mA to exceed its power rating, but that's still well within the 6A fuse limit. Two KT88s at full power could certainly swing 140mA.

[Tom Phillips]

Joe,

All is fine until you crank the amp. Therefore, it appears that the wiring is correct. I’m going to assume that the cathode pin where you have the 100 Ω resistors grounded is really at system ground potential. If it wasn’t the output tunes wouldn’t work correctly. You can verify that it is a good ground with your Ohm meter.

Under normal operation each of the 100 Ω resistors have ~3.15 volts across them and are therefore dissipating only ~ 0.1 W. The resistors are being fried when something happens that puts much higher than normal voltage across those resistors. Following are what I think are the possible causes and what I would do if the amp was on my bench.

1) A wiring short between the heater supply and the plate or screen HV supply. I would inspect all the wiring paying close attention to the areas around pins 2 & 3 of the power tube sockets. Pull the power tubes and inspect both sides of the tube sockets. With the amp unplugged and all the caps fully drained of charge I would connect my DVM in continuity test mode between the screen supply and the heater supply. (Pins 2 & 4 of the power tube socket is a good place to connect) Then I’d wiggle the wiring and verify that there is never any continuity. I’d then repeat the test connecting between the plate supply and the heater supply (Pins 2 & 3). If there is ever the slightest peep out of the continuity tester I’d isolate to the cause and fix it of course. Note: This works best when using a test meter that has a quick response time when in continuity test mode.

2) A sort inside one or more of the power tubes between the plate or screen to the heater. This is the most common failure mode related to the problem you have described.
I know you said that this has happened with two sets of tubes but it’s still a likely possibility. Your new tubes are not “Known good parts” as many people assume. They are just “new”. A known good set would be tubes swapped from a similar amp that were working flawlessly in that other amp.
To check the tubes I would install one power tube at a time. Power up the amp with a speaker or dummy load. Then tap and viriate the tube to try to force the possible intermittent short to happen. Repeat with each of the four tubes swapping them in one at a time. If you find a bad tube keep testing the others until you have checked all four.

3) Could be an intermittent short between the HV & the heater winding inside the power transformer. This seems the least likely of the problems mentioned. Especially since it only happens when the amp is cranked. The pro test is to put the PT on a high pot tester. However, that is not an option in most shops and requires that the PT be disconnected from the circuit. I just mention this for completeness. Hopefully the information in #1 or #2 will help.

Regards,
Tom

[melvin]

Thanks for the responses.

Billy R- The rating of the resistors is 1W. They disintegrated almost completely and vaporized part of the tab at the pin one of them was connected to.

Rhodesplyr- To answer your questions, the hum balance pot was ineffective on the power tube filaments so I put it on the pre-amp tube filament circuit. It is, however, grounded at the same place as the two 100 ohm resistors. Also, pin 1 on each socket is not connected to anything.

Tom Phillips- I will try #1 and #2 as you described. Hopefully, one of these will bear fruit.

One question though, is it possible that there is something I did not consider when converting to KT88's?

[Billy R aka DynaFreak]

Hmmm. You say the hum balance pot was ineffective? Did you test it? Those things are notorius for going bad. And if recall correctly those V4s ground everything in one place at the input jack. Have you checked the ground there, not just with a meter but by taking it apart and visual inspection? I would still be very tempted to temporarily take out the hum balance pot, and re-do the 100 ohm resistors and ground to the chassis, the idea being to separate the power amp circuit completely from the filament heating circuit.

[melvin]

The hum balance pot is new and effective in the preamp circuit.

Separating the grounds does sound like a good idea.

And I will check the ground at the input jack.

[Rhodesplyr]

Hearing that the resistors were vaporized, I think I'm now with Tom that it's a higher voltage short, not just cathode current somehow getting pulled through the 100 Ohm resistors. As Tom points out, you've got the KT88 plate on pin 3 and the heater on pin 2. The V4 is already a high plate voltage amp, and, cranked, there are going to be some VERY large voltage swings at tube socket pin 3. Once you've got a carbon track on the tube socket from an arc, it can easily recur.

I would consider replacing the output tube sockets if you don't find another obvious cause. I've had two amps that suffered from intermittent crackling and popping noises that I could never track down, but they vanished when I installed new sockets.

[melvin]

Well, as it turns out, one of the KT88's is bad. There's a short between pin 2 and 3 on the tube (not the socket). Thanks for the help. I guess I'm just being paranoid, because there's no way for the amp to cause something like this except vibration, right?

[Tom Phillips]

... I guess I'm just being paranoid, because there's no way for the amp to cause something like this except vibration, right?

I'm not quite sure what you are getting at. If the amp is well constructed it should stand up to normal transportation and in use vibration. Vibration can be hard on tubes however, and shorten their useful life. It can also help weed out substandard tubes.

Glad you found the problem,
Tom

[fixitim]

are the rest of pin 8's connected together?
they need to be

[Rhodesplyr]

What tube factory are these KT88s from? I don't think I've ever run across a direct plate-heater short from something happening inside a beam tetrode, though it's possible. (I've seen more shorted screens.) It's also possible that there's some sort of conductive debris inside the base that shorted the adjacent pins. It's hard for me to think of a way the amp itself could cause that kind of catastrophic failure.

Also, you said that this happened with two sets of KT88s. Were they all from the same batch/source? There is a manufacturer that's been known to send out some bad batches of these.

[melvin]

Tom - That's all I meant. The amp weeded out the substandard tube. I'm just still green enough to first assume it's my fault. Thanks again for the help.

Rhodesplyr- The KT88's are from 2 different manufacturers. EH and then JJ's.

As far as the bad batch thing goes, I'm betting they don't put their tubes through quality-control testing that involves the conditions in a V4 at full volume. And while this tube did not fail until then, from reading the published tolerances, I wouldn't think it'd be a problem. I guess one question is do they test them at maximum ratings or at typical?

Of course, there is the jarring shipment.

I will replace the tube and test the thing at full volume, believe me. I will let you know what happens. (I just hate when things bounce back.)