SF Vibrolux redplating, losing voltage

[AxeAmpTN]

A month ago, a customer brought me a '74-ish Vibrolux with the classic "found it in an attic" story.  His complaint was that it made no sound.  I opened it up to find that something had previously blown up inside (see photos), but had apparently been replaced.  Most of the components look to be fairly old, so whatever happened was awhile back.  Long story short, I found a broken PI tube and replaced it.  Customer wanted to try it out, so I let him.  It worked and sounded good...too good.  I noticed a lot of heat coming from the back and noticed the 6L6's starting to redplate.  Infrared thermometer read nearly 500 degrees on the plates!  I warned the customer that this was bad and it needed to be addressed, but he insisted on taking the amp anyway so he could use it for a recording.  Fine.

He brought it back last week to see if I could adjust the bias back down real quick.  I looked everything over and ran a few tests to see what was going on. 

First of all, the plates measured at 269VDC. By my estimates, the current was probably running near 80mA!  (I didn't bother with actual current readings because clearly there was too much.)  But I was definitely puzzled at the low voltage.  I explained to the customer that not only was this well outside the range of normal, but with only a bias-balance in the circuit, this was going to take some investigating.

I checked the rectifier output.  (Brand new JJ 5U4.) On standby, it measured 410-ish VDC, but dropped to 269 with standby off.  I started checking connections to find what could be shorted and draining off voltage.  The filter caps are clearly old and need replacing, but didn't raise other red flags. 

With the power tubes removed (brand new Ruby 6L6's), the rectifier output and pin 3 of the tube sockets measured correctly (445 VDC).

I checked the bias board--which appears extra crispy--and sure enough, it's not passing voltage to the tubes.  The PT bias tap measured 49.9VAC, but the DC after the diode only measured millivolts.  So clearly, I'm looking to replace the bias circuit.

So, I'm posting here to glean from y'all's experience and expertise.  This is easily the worst amp I've seen (included a post flood Champ that had gotten all moldy).  What am I missing?  Any extra precautions you recommend (besides the usual)?  Anything else?

[sluckey]

Fix the bias! Since you only measure a few millivolts after the diode (I'm assuming you mean immediately after the diode) my first suspect would be a shorted filter cap. Second suspect would be an open diode. Both are easy to check. The bias caps (there are two) need to be replaced even if they are not the problem.

DON'T PUT THE 6L6S BACK IN UNTIL YOU CAN MEASURE APPROX. -50VDC ON PIN 5 OF EACH POWER TUBE SOCKET.

[AxeAmpTN]

Fix the bias! Since you only measure a few millivolts after the diode (I'm assuming you mean immediately after the diode) my first suspect would be a shorted filter cap. Second suspect would be an open diode. Both are easy to check. The bias caps (there are two) need to be replaced even if they are not the problem.

DON'T PUT THE 6L6S BACK IN UNTIL YOU CAN MEASURE APPROX. -50VDC ON PIN 5 OF EACH POWER TUBE SOCKET.

Thanks, sluckey.  I'm checking my stock of parts right now to replace this whole board.  I'm recommending to the customer to convert to an actual bias adjustment pot too. 

Quick question:  I don't see a designation for the diode.  I have a few 1N4007's in stock...would that perform sufficiently?  Nevermind...found it! 

[sluckey]

1N4007 is fine. But check the one that's in there. Also check that resistor (maybe 2.7K) between the red/blue lead and the diode. And check for shorted bias caps too. Only talking two minutes to do this.

[sluckey]

Nevermind...found it!

Found what? The diode part number? The problem? Let us know.

[AxeAmpTN]

Nevermind...found it!

Found what? The diode part number? The problem? Let us know.

I found Doug's page on bias circuits. https://el34world.com/charts/Biascircuits.htm

I'll definitely still check in individual components (diode & caps) when I have it in front of me tomorrow.

[AxeAmpTN]

Okay. Got to check the bias board:

Resistor between PT and diode measured just under 1K. (Schematic says 1K.)
Diode checks bad; open both ways.
Both caps measured bad...capacitance measured in nF instead of uF.

My recommendations to the customer will be a recap and replacing the circuit with an actual bias pot in order to make real adjustments (since the bias circuit needs to be replaced anyway).

[PRR]

> the plates measured at 269VDC. By my estimates, the current was probably running near 80mA!

If the bias should be negative "dozens of volts", and you have "mV", then the poor 6L6es will try to pass several *hundred* mA. 6L6 with 270V on G2 and "zero" on G1 is off the chart at about 170mA, so 340mA for a pair.

This is FAR more than any normal operation. Highest number in 6L6GC data is 210mA/pair.

And *directly* explains the massive sag. When my house water pump starts-up my lights dim. (long wire from street) If I dropped a rock in the pump so the motor strained flat-out, my lights would DIM.

[mikeymike]

This looks so familiar......... 

What does it smell like? I wonder if it's the same smell my 65 has. It's a very headache-incuding type of smell that gets stronger when the amp warms up.

[AxeAmpTN]

This looks so familiar......... 

What does it smell like? I wonder if it's the same smell my 65 has. It's a very headache-incuding type of smell that gets stronger when the amp warms up.

It's that distinct, slightly fishy, burnt oil/burnt electronics smell.

I have 2 work benches at 90^o to one another--one for amps and the other for guitar work.  The Vibrolux is sitting open on the amp bench for now.  As I'm doing guitar work, once in awhile I catch a whiff of the Vibrolux and have to stop and doublecheck that nothing is on the verge of blowing up (not that it's even plugged in).

Incidentally, I've dubbed this amp the "Scary Vibrolux."

[mikeymike]

Scarylux 

My 65 pro reverb had a magic smoke event near the pilot lamp of some sort, I posted about it on here recently. It looks almost exactly the same as what happened to the vibrolux in your photos. There's this golden tint on everything inside my amp that was exposed. It took some scrubbing to get it off the brass grounding plate. I fear my pots have this crap embedded on the all carbon strips (except for the exact spot where the control was set when it went kaboom). Definitely give yours a check for that if you can.

[AxeAmpTN]

Scarylux 

My 65 pro reverb had a magic smoke event near the pilot lamp of some sort, I posted about it on here recently. It looks almost exactly the same as what happened to the vibrolux in your photos. There's this golden tint on everything inside my amp that was exposed. It took some scrubbing to get it off the brass grounding plate. I fear my pots have this crap embedded on the all carbon strips (except for the exact spot where the control was set when it went kaboom). Definitely give yours a check for that if you can.

Scarylux indeed, mikeymike.  If yours recently did this, definitely clean out those pots!

As for the Scary Vibrolux, whenever that bias board popped (presumably before the current owner got it) it was probably not the only thing that happened.  I don't know if you can see in the photos, but there are definitely burn marks near the rectifier and power tube sockets, as well as near the transformer. However, the PT works (correct AC voltages) and the tube sockets have no marks from arcing, so I am guessing they were replaced at some point after the initial blowup event.  Adding to this, all the date codes are from different years ('70 thru '74 as far as I can tell), which leads me to believe that there have been replacements made. 

The pots all clearly have burns on the back of them (especially near the bias board), but don't even crackle!  Someone must have used a ton of Deoxit on them at some point.  Even with the power tubes in runaway mode (with 0VDC on the grid), the controls all work as expected.