Possible issue with a guitar amp

[quad]

Hello folks,

I am a frequent guest at DIYstompboxes, reading up on all kinds of fun projects and problems, but not a frequent contributor when it comes to discussion. I suppose this is not the optimal place to talk about such an issue on a stompbox forum, since the trouble I'm having is with a guitar amp. On the other hand, this community is great all around and hopefully someone might have a hint with what went wrong.
    So, here's the situation. I finished building a JTM45 guitar amp. I sourced the schematics and layout from TubeDepot and parts from all over. Mostly the US and Germany. A few changes that I made to the original design of the amp include the Larry's grounding scheme and PPIMV (post phase inverter master volume). Now, everything seems to be working alright. The voltages check out (including the bias voltage for my KT66's), the amp is dead silent and the sound is sweet! Yesterday I pushed the amp to its limits. Both volumes at 10 and master volume towards 10 as well. I had protective ear plugs, so at least no harm was done to me. 
    After playing for just a minute, I noticed my KT66's were red plating! The blue glow of the valves mesmerized me for a few seconds but then I came back to my senses and shut the amp down immediately. I pulled out the output valves and started testing voltages. Bias, plate, cathodes... Everything checked out. Placed the KT66's back in again, turned on the amp and it blew my B+ fuse and HV center tap fuse (note, that no fuses were blown with output valves taken out). Replaced the fuses, did more testing for cold/bad solder joints, mistakes, shorts etc.. Found nothing. Voltages are spot on again. The bias section appears to be operating just fine. This time after the fuses were replaced and with KT66's back in position, the amp fired up without any issues. But I made no changes! Thanks to those fuses, nothing previously got cooked. I felt a little scared about my new, precious KT66's. Not to mention any of the transformers! Re-biased the output valves to read ~ 38mA (using 1ohm resistor mV=mA). Although my bias voltage jiggles back and forth by a margin of ~ 2volts. Even after keeping the amp on for a while to let it stabilize. Not sure if that's normal operation.
    So, that's the story. As of yet I have no idea what happened and whether it would repeat at the same conditions, which is something I'm not going to attempt for a while. Perhaps the bias voltage went out of whack for some reason, so the grid wasn't biased properly and the KT66's started conducting excessive amounts of charge? All kinds of questions in my head. Puzzling...

To summarize: The output valves (KT66's) started red-plating after pushing the volume of the amp to the max.

Couple of pictures of the amp:
http://oi60.tinypic.com/20s9mqu.jpg
http://oi60.tinypic.com/e0k0wm.jpg

[alfafalfa]

After playing for just a minute, I noticed my KT66's were red plating!

It could have been a bad bias connection, the tubes went without bias voltage .
I would check the negative biasvoltage connection, you might have a bad solder point there.

And at the same time I want to complement you on the terrific job you did. It looks fantastic, everything very neat and tidy . Great job .


Alf   

[KazooMan]

Beautiful build!

No solution for you, just two suggestions.

First:  Whenever you have had an issue with your amp or make any mods fire it up using a lightbulb limiter.  That will protect things and indicate if you have any major problem causing excessive current draw.

Second:  Post your question on the ampgarage.com forum.  The regulars over there are really good at helping with problems.

[vigilante397]

I'm afraid I'm not any help, I tried my hand at amp building and discovered I'm terrible at it. But I just wanted to tell you the amp looks fantastic and I hope you get it working!

[alfafalfa]

Ampage was the old name of the forum . It's now:

http://music-electronics-forum.com/

I had a look at the schematic, it's fixed bias I noticed , you could solder a 1M resistor across the 50 K bias pot ( from input to wiper ) , so if the wiper doesn't make good contact  you still have a very negative voltage and this prevents redplating.

[alfafalfa]

Sorry , didn't read very well. I thought it said "ampage" but it read "ampgarage".  

 

[R.G.]

It is very likely that you have an intermittent connection of your bias to the tubes. Any loss of bias will cause redplating and fuse opening.

Remove the output tubes, clip your meter's ground lead to the chassis, and stick the hot probe lead into the grid pin hole on an output tube socket. When you turn on the power, you should see the negative bias voltage on your meter. This should be in the range of -35 to -60V.

Now use a chopstick or other non-conductive probe (because this isn't worth dying for!) to wiggle things around and find out what causes the voltage reading on the meter to drop to nil.

[quad]

Thank you all for the kind words and suggestions! There's a good possibility that I left a weak connection after doing the PPIMV mod.

It is very likely that you have an intermittent connection of your bias to the tubes. Any loss of bias will cause redplating and fuse opening.

Remove the output tubes, clip your meter's ground lead to the chassis, and stick the hot probe lead into the grid pin hole on an output tube socket. When you turn on the power, you should see the negative bias voltage on your meter. This should be in the range of -35 to -60V.

Now use a chopstick or other non-conductive probe (because this isn't worth dying for!) to wiggle things around and find out what causes the voltage reading on the meter to drop to nil.

I've tried measuring the bias voltage after taking out the valves - checked out. But trying to wiggle those cables around a little bit is a good idea! Oh, and I'll definitely use a non conductive material to poke around. Wooden stick will do 

[KazooMan]

Yep, the amp garage and music electronics forums are distinct.  The former is really focussed on amps.  It also seems to have a lot more traffic and a more involved membership.

Question about the intermittent bias suggestions.  Wouldn't you hear that in the output?  From what I read I gathered that the amp sounded just fine when he noticed the red plating. 

[quad]

Question about the intermittent bias suggestions.  Wouldn't you hear that in the output?  From what I read I gathered that the amp sounded just fine when he noticed the red plating. 

Indeed. The amp still sounds fine right now, as it did before I turned bright/normal and master volumes up to 10 and played guitar. Only at that point I noticed my output valves were starting to red plate and glow blue. It's weird. You'd expect the valves to red plate at any volume if there was a poor connection to set the bias. Well, I have to check those connection first to rule out the possible intermittent issue.

[R.G.]

A really sneaky problem that happens only rarely is that the filter cap for the bias supply gets leaky when it gets hot.  That just reduces the bias voltage a lot without necessarily making it go away, and the problem is not there when it cools off.

[PRR]

> Re-biased the output valves to read ~ 38mA (using 1ohm resistor mV=mA). Although my bias voltage jiggles back and forth by a margin of ~ 2volts. Even after keeping the amp on for a while to let it stabilize. Not sure if that's normal operation.

Two Volts?? 2V in 1 Ohm is 2 AMPS. The tubes can only pass a half-amp peak!!

If you mean 2 milliVolts, 36mV-40mV, that's not extreme, but is suggestive.

Jiggle is not normal. It is motorboating, but not enough to flap the speaker.

What is the jiggle rate? 5 per second? Once a minute? (Have you watched your wall-voltage to be sure it isn't utility-company jiggle?)

Are you sure you got the "negative" feedback right? Swap the OT plate leads. If the amp gets a little less-loud, that is correct. If it gets more-loud and howls, you were right the first time.

Power-cap and bias-cap troubles can cause many funny things. (Crowhurst writes about a hi-fi which was fine except on sustained trumpet.... fixed with new power cap.)

[quad]

Sorry, I meant 2mV. It's ~38mV +- 2mV. Not sure what the rate is. I'll check out the wall voltage. Got to do more measurements. My OT leads are connected: red lead to V4 and black lead to V5.

The OT datasheet:
http://www.classictone.net/40-18039.pdf

Amplifier schematic:
http://oi58.tinypic.com/14tpx1j.jpg

[quad]

Hi,

I performed a couple of measurements. First off, the mains voltage varies by 2 - 5 volts, which seems reasonable. I couldn't find a loose solder joint. I moved the wires while performing the measurements and the readings stood still. What I did find, however, is the suspicious bias diode (visible in the photograph, a good portion covered off by the black wire). It was the one component I wasn't sure about. The metal can, general purpose rectifier type of diode, which seemed okay when I measured it. I couldn't find a proper datasheet for it, just that it was a rectifier diode. I replaced it with a 1n4007. The bias current is more stable now at ~40mA. And no catastrophic loss of bias voltage to cause red plating either. Seems okay now.