Unbalanced Bias Current Problem

[MikeSafreed]

This is embarrassing to post - I've been building and repairing amps for a VERY long time, but one repair has me stumped, and I need the collective knowledge of this group to please help me out. (I had brain surgery a few years ago, and as a result, I've lost a few brain cells that contained some important facts - I think this is one of them!)

I had a 1971 Fender Virbrolux Reverb in my shop one year ago. The customer wanted it retubed, recapped, and blackfaced. Simple enough. I did the work, the amp sounded great and the customer sent me a lot of referral work. About 3 weeks ago, he called to say the amp started sounded "crappy" - from his description, it sounded like the power tubes were wearing out. I got the amp, took a listen and tried biasing in a new set of power tubes. No change. The sound of the amp is very distorted - notes beating against each other at higher volumes, but you can tell at lower volumes that the amp is not right. At HIGH volume is sounds like blown speakers - I tried another cabinet, but speakers are not the problem.

It also sounded like maybe one of the new filter caps failed, so I went ahead and changed them all (5). No change in the symptoms.

The problem occurs in either the Vibrato or Normal channel, so I moved to the Phase Inverter. I tried a new tube (actually tried all new preamp tubes), but no change. Since I had blackfaced this amp, I looked for a failed component in the phase inverter, and went ahead and rebuilt it - new caps and resistors. No change.

I moved to the power tubes. I found a bad pin on each tube socket, so I replaced each socket, with new resistors on each socket. No change in the symptoms.

I made sure I checked the negative feedback circuit and everything checks out OK. I replaced the 100 ohm resistor and the 820 ohn resistor. No change in the symptoms.

I tried a new OT. No change in the symptoms. I tried a new choke. No change in the symtoms.

Tonight, I played through the amp while I measured the bias current draw. The problem is that one tube is pulling much more current than the other tube. The 6L6 closest to the rectifier is pulling 120 ma when you strike a hard open low E. The other 6L6 pulls only about 65 ma doing the same thing. I checked the idle bias voltages on the power tubes and I'm seeing about -46 volt on each tube. All the components in the bias circuit check OK. The bias ground from each tube is through a one ohm resistor, so I changed them and made sure the grounds were solid. No change in the symptoms.

All the voltages in the amp are OK. I also hit all the solder joints with the iron just to be sure...

So, I have an amp that sounds unbalanced, and it certainly is. Does anyone have a suggestion on what to check next, please?

[rdjones]

Are the idle currents balanced ?
Does the imbalance follow the tube when swapping output tube-to-tube or stay with the socket ?
Can you scope the grid drive at the 6L6's or if not measure the AC voltages ?
Same question for the PI grids.

Basically you want to know if the imbalance is a drive problem or a current supply problem.

rd

[martin manning]

You said you've rebuilt the PI, does that mean you replaced or checked the PI output coupling caps for leakage?

[Prairie Dawg]

You said you've rebuilt the PI, does that mean you replaced or checked the PI output coupling caps for leakage?

That would be the place I'd be looking, the caps that couple the PI output to the grids of the power tubes.

[MikeSafreed]

The AC voltages on the grids of the power tubes are very imbalanced. The tube with the higher bias amp draw also has an AC voltage on the grid of around 46 VAC. The other tube shows 36VAC. This is measured using the same test as I did to determine that under an AC load the left 6L6 is drawing considerably more bias current that does the right 6L6.

*** The higher bias amperage draw does not follow the tube - it stays with the socket closest to the rectifier. ***

The idle current bias is pretty balanced - one tube measures 29.2 MA and the other tube measures 31 MA, and those measurements follow each tube. About what I'd expect for tubes these days...

I did replace the coupling caps in the phase inverter. I did that as one of the first steps in troubleshooting and it had no impact on the problem. As I said in my original post, I rebuilt the phase intverter and the negative feedback circuits and that had no impact on the problem.

The AC voltage on the PI measures around 4 VAC on pin 2 and around 1.4 VAC on pin 7, using the same input signal that causes the bias on the power tubes to go way high on one power tube (120 MA) and about 55 MA lower (65 MA) on the other power tube. Not right to my way of thinking.

I would have said that the problem is some type of grounding issue on the power tubes had I not already replaced both power tube sockets, and installed new 1 ohm resistors on pin 8 to ground on each tube.

I double-checked all the cap and resistor values and everything is really close to the AA964 schematic. Those 220K resistors inline with the PI couling caps measure at 220.5K each. Those coupling caps are the 2nd set in there, and they are both from a new batch of Orange Drops, and they meaure right on at .1uf. They show no leakage when in the circuit.

Here's an update on the DC voltages on the PI. Pin 2, 43.92 VDC, Pin 7, 46.6 VDC, Pins 3-8, 73.1 VDC, Pin 1, 205 VDC, Pin 6, 201 VDC. The DC voltages off of each 220K resistor/.1uf coupling cap feeding the 1.5 K resistor on each power tube measures -46.55 VDC.

I'm still thinking it's a ground issue, but where?

[Firestorm]

The AC voltage on the PI measures around 4 VAC on pin 2 and around 1.4 VAC on pin 7, using the same input signal that causes the bias on the power tubes to go way high on one power tube (120 MA) and about 55 MA lower (65 MA) on the other power tube.

I'm not sure what the AC relationship of those two pins is supposed to be ... signal enters the first triode of the PI through Pin 2, but enters the second triode through the cathode. Try to get an AC measurement (with the test signal) at both PI plates or at Pin 5 of both output tubes tubes.

Unfortunately, the problem is beginning to sound parasitic in nature. I don't know how extensively you rewired the amp when you blackfaced it, but the spaghetti wire Fenders from the '70s are prone to all sorts of phantom coupling. Might even be a conductive board.

Just for fun, you might try putting a .002 cap from grid to ground on the problem tube (just like it had when it left the factory) to see if it settles the tube down.

[MikeSafreed]

I, too, am beginning to suspect some type of parasitic issue. I've changed too many parts and not affected the symptoms in the least. Since it's an imbalance issue, you'd think that something would be wrong with one side of the PI or the power section, but changing out everything in both sections has no impact. Redoing the grounds on the power section had no impact. The OT could have been the culprit, but wasn't. The only weird part is that the amp played fine for a year and THEN started sounding crappy. I would have thought a parasitic would have shown up much sooner, but I guess weirder things have happened. When I blackfaced the amp, I didn't touch anything except the PI components and I think a couple of cap values in the preamp. No wiring changes except for the bias circuit, to and from the bias pot. The amp came from the factory with shielded grid wires on the inputs. I did remove those suppression caps from the power tubes. I'll put a cap on the tube that's causing the issue and see if it helps.

The suggestion about the board is something I am going to pursue. I've had conductive boards before - even had one that I could measure DC on bare spots on the board. Usually, those voltages and currents are so low that it's nearly impossible to measure them.

This is a lose-lose situation for me. I have so much time tied up is trying to solve the problem that I'll never recover my costs, and I'm beginning to think the only way out is to strip the amp and rebuild it - even more time...

Thanks for the suggestions... I think there's more golf in my future!

[martin manning]

I would expect to see an imbalance if the PI is being driven very hard for some reason. Are you looking at this with a 'scope, and is it something that will show up with a sine or square wave applied at the input? If so does it show up if you isolate the power stage (disconnect the PI input coupling cap from the mix resistors upstream) and apply the sine wave at the PI input?

[MikeSafreed]

Martin,

My scope is down - it's old and too expensive to repair. So, unfortunately, I'm limited in the types of signal tracing I can perform.

Since the issue occurs using either input stage (Normal or Vibrato), I thought I would try to make sure that something wasn't happening in the reverb section that was overloading the PI even when the reverb section was not supposed to be active, so I disconnected the signal wire at the "Y" junction 220K resistor input to the PI, but it had no effect. If something is overdriving the PI, it would almost have to be something happening in the vibrato section ground at the bypass cap/100 ohm resistor junction. But, I regrounded that and it had no impact.

I'm going to disconnect the vibrato circuit and make sure there's nothing going on there anyway.

[MikeSafreed]

I put the suppression caps back on - no change in the symptoms. I disconnected the vibrato from the circuit - no change in the symptoms.

On Sunday, I am going to remove the PI tube socket, replace it and all the wiring attached to it, and see what happens. If this causes no change, I am going to call my customer and give him 2 options - he can take it back as is and make a nice yard planter out of it, or he can let me rebuild it completely as an AA964 using new pots, new tube sockets, and a new circuit board and new ground buss. I'll reuse his xfrmrs, but that's all. I'm at the point where I'm ready to do it for free...

[Firestorm]

Wow. You could drive the amp over to Dayton some night and leave the car doors unlocked for a few hours...

Before you turn it into a planter, try disconnecting the negative feedback at the speaker jack and see if there's a change. With the feedback still disconnected, try swapping the grid leads and see if the problem moves. If you have a signal generator, you could also try pulling all the preamp tubes and feeding the PI from the generator, basically isolating the PA from everything else.

After that, the drive to Dayton...

[Phil_S]

Mike: With all respect for what you are doing -- I sense you want to do right by the customer, maybe you can find enough cash to get a scope that works. A scope may have saved you a bundle. If it were me, I'd want to keep this from happening again.

I don't know how good this place is, but they always seem to have an OK low end new scope on their price list for around $350. Here's one that I found today. http://www.tequipment.net/RigolDS1052E.html I recognize, everyone's idea of "a lot of money" is different. In case you were thinking a scope is north of $1000, I figured it couldn't hurt to show you the $350 one. If you search the website, you might find one for even less.
http://www.tequipment.net/dualsearch.as ... illoscopes

Good luck finishing this one.

[MikeSafreed]

I will try disconnecting the NFB at the speaker jack on Monday afternoon. Good suggestion...

Regarding the scope, thanks for the tip - I will check it out tonight. The scope I had was really high-end, but old. It died a horrible death - while rearranging the shop, moving some benches, my helper moved the scope from it's shelf above the bench to a concrete floor. The total "move" operation took only milliseconds to complete, and the "move" was completed without the use of any hands. It proved scientifically that concrete is denser than the materials that house a scope (the guts of the scope have no ability to fend for themselves when subjected to a multiple-G-force impact). Stupid scope. Stupid helper (ex helper)...

Anyway, here's something interesting about the problem. It appears to be frequency and/or gain sensitive. By changing the frequency of the signal, the syptoms increase or decrease. Lower frequencies make it worse - higher frequencies reduce the distortion and reduce the imbalance. Also, the strength of the signal impacts the symptoms. For example, turn up the bass on the amp, increase the volume, and the symptoms are horrible. Reduce the bass without reducing the volume and the symptoms improve. Leave the bass up and reduce the volume and the symptoms improve.

[alvarezh]

Mike, I have been following your posts very carefully because a have a similar (not exactly the same) problem with a Blockhead and I want to see if your solution can help me. But, I am writing not about my problem but about the fact that:

Bass = 60 or 120 Hz….. does that ring a bell? Could the problem be in your P.T.?

If I am correct, you have stated you have changed the O.T. but not the P.T.

After all the logical steps have been taken, like you have, maybe the solution will show itself by looking into the illogical ones!

All the best.

[Phil_S]

... The scope I had was really high-end, but old. It died a horrible death - while rearranging the shop, moving some benches, my helper moved the scope from it's shelf above the bench to a concrete floor. The total "move" operation took only milliseconds to complete, and the "move" was completed without the use of any hands...

Ugh! Too bad this happened. It may be hard to move from a Mercedes level scope to a Chevy level scope, but, given the waste of time factor, it will get you to the store a lot faster than walking. You can always sell it later -- that $350 scope will fetch you $175 easily; a low dollar/low risk proposition -- when you are ready to invest in a better one.