Fender Bassman 100 red plating problems

[soundmasterg]

Hi guys,

I am wondering if any of you have any ideas on this one. I'm out of ideas at the moment as to what is causing this problem.

Amp is from a working band's bass player and is a Silverface Bassman 100. Customer hadn't noticed anything wrong with the amp other than the fuse holder was cracked. So I replaced that and fired it up, only to notice that power tube #3 was glowing red. So I shut it down and tested that tube on my mutual conductance tester and it was almost completely dead. All the tubes were Tung-Sol reissue 5881's and had all been replaced at the same time. Power tube #4 had no cathode current and I replaced a blown screen grid resistor. Ordered a new Tung-Sol reissue 5881 for power tube #3 and replaced it. Power tube #1, #2, and #4 all tested out fine, so after retensioning and cleaning the sockets, I fired everything up again and monitered the amp. Set the bias to around 64% of the max and everything sounded good and worked well. Let it cook for 20 minutes with no problems. Called up the bass player and told him to pick up his amp. So he calls me two days later after a show and amp made it through 5 songs and then the sound got funky and he looked and power tubes #3 and #4 were red plating. He shut it off and I have it again on my bench now. Power tubes #3 and #4 are dead and burned, and power tubes #1 and #2 are in perfect health. With no tubes each socket has bias voltage. The screen voltage makes it across the screen resistors. The plate voltage is on all sockets. The coupling caps from the phase inverter to the power tubes don't leak on either side. The amp had been completely rebuilt several years ago and has mostly new carbon comp resistors, Sprague Atom electrolytics, and Mallory 150 coupling caps. Everything in the bias supply is new too. Previously before the second failure all tubes showed proper current draw and were biased and matched well. No evidence of carbon traces on the sockets. It is only power tube sockets #3 and #4 that have the problem, and they both are on the same side of the output transformer. It measures fine as far as I can tell though. The amp will run happily with no problems for hours with no power tubes in there. Somehow when the tubes are added, the two on that side go out of control. It seems like they are losing bias somehow but I haven't checked that since I got the amp back....sort of reluctant to blow any more tubes until I have an idea of what could be going on.....

I will add that the bass player who owns the amp used the right impedance. He is a professional working musician and has played this particular amp with the same rig for the last 20 years with no problems. I have a feeling the tubes are somehow losing bias, but why is the question. These particular tubes have been in use for 6 months for no problems in this amp...well at least until the recent problems. Whatever is going on, it seems to be sockets 3 and 4 as 1 and 2 have no issues and those two tubes are still working just fine. The sockets are fine...no evidence of arcing....pins are not shorted to the chassis....the amp has about -48v bias at the sockets with no tubes in them the way it is adjusted at the moment. The bias control has been changed by past technicians to a coarse bias adjust and it has enough range and works fine. All of the parts are relatively new in the amp as it was overhauled within the last 5 years. Someone on another forum mentioned the Tube Store recommendation about the Tung-Sol 5881 only being a 400v tube...anyone know if this is the case truly or is someone just going off the old datasheets? New Sensor doesn't really share their specs readily. I may have to call them and ask...but most tubes made these days can handle 400v+ as long as you don't exceed the max dissipation. I doubt New Sensor would be making a tube with such a voltage restriction as that would limit their sales. I would guess the 400v max is someone going off the old datasheets, but I suppose a call to new Sensor wouldn't hurt. Incidentally I did set the bias on this based on the 23 watts max and not the 30 watts max of a 6L6GC. Plste voltage on this amp is about 448v. I can try some other power tubes in it, but the fact that the problem seems localized to one side of the output transformer and not the two tubes on the other, and the fact that whatever is going on killed a brand new tube quickly also makes me think it is something else.

Normally I fix stuff like this myself without any issues, but this one has been perplexing for some reason....I'm stumped as to what to look at next. Anyone have any ideas?

Thanks,

Greg

[soundguruman]

The bias voltage has failed.
The filter capacitor for the bias supply may be shorted. OR the bias adjustment pot may be dirty or broken, OR a resistor is open (not usually)
If you measure the voltage on pin 5 of the output tube, it should go up or down when you turn the adjuster pot.
So usually, the capacitor is shorted, and it happens when the part gets too old.
If just two of the tubes turn red, you could actually have a bad 220K resistor, or a broken wire...

[soundmasterg]

Thanks for the reply.

The bias voltage is just fine...as I noted above, right now it is -48v at pin 5 on the socket. Previously I adjusted it to this level because that is where it needed to be for the tubes that were in there to bias up correctly, but it adjusts fine to any other level.

The electrolytic cap in the bias supply is maybe a year old and again, the supply is fine and works fine so I don't think that is it.

Somehow the tubes on that side begin to red plate after awhile, though they were fine when the amp was here previously for more than 20 minutes of use. The tubes on the other side are always fine....so the bias supply is fine....but perhaps that 220k resistor is bad. I tested it, and it seems fine, but I can always just replace it to be sure it isn't the culprit. It does make sense why the 3 and 4 side would lose bias if that resistor was somehow going open. The wires all look and test fine.....

Greg

[Steve Conner]

Tubes will sometimes develop excessive grid current and go into thermal runaway of their own accord. Usually at the end of life, but poor quality ones can do it from new. When this happens in a 4 tube amp, the grid current from the bad tube will pull down the bias voltage for its partner on that side and burn it out too.

(Tech sidebar: The 220k resistor used for bias feed is outwith the manufacturer's spec for the 6L6GC. The original GE datasheet says 100k max for a single tube: two tubes should have 50k. So with a 220k bias feed resistor, even good tubes are technically within their rights to go into runaway and melt down. Insert rant about tubes not being what they used to be, etc.)

So, my first suspicion would be a "bad" tube out of the box. My second would be the 220k resistor drifted high or intermittent, my third a leaky coupling capacitor, and my fourth would be parasitic oscillations caused by poor layout of the screen or grid resistors.

[soundguruman]

Some of these amps have a "balance pot" that is not a bias control. The balance is off center?
(I'm not sure which one you have)
The tubes "are bad?"
The output transformer has shorted windings on one side?
Half the phase inverter is going into oscillation? (unusual, but not impossible)
The load is screwy? Bad speaker? Or are you using an attenuator box?
I suppose it's possible that half the output tranny has shorted windings on one side.

[soundmasterg]

Tubes will sometimes develop excessive grid current and go into thermal runaway of their own accord. Usually at the end of life, but poor quality ones can do it from new. When this happens in a 4 tube amp, the grid current from the bad tube will pull down the bias voltage for its partner on that side and burn it out too.

(Tech sidebar: The 220k resistor used for bias feed is outwith the manufacturer's spec for the 6L6GC. The original GE datasheet says 100k max for a single tube: two tubes should have 50k. So with a 220k bias feed resistor, even good tubes are technically within their rights to go into runaway and melt down. Insert rant about tubes not being what they used to be, etc.)

So, my first suspicion would be a "bad" tube out of the box. My second would be the 220k resistor drifted high or intermittent, my third a leaky coupling capacitor, and my fourth would be parasitic oscillations caused by poor layout of the screen or grid resistors.

Thanks for the reply Steve!

Given that power tube #4 had burned out a screen grid resistor previously, I suppose it could have been bad and pulled down the bias voltage for the new tube I put in the 3rd position and killed it also. I have some old Sovtek 5881's around that are still ok that I used for testing, so I guess I can stick those in and see what happens to the amp when I run a signal through it...make sure there aren't any oscillations going on.

Soundguruman, if you take a look in the original post, I did say that the bias balance had previously been converted over to a coarse bias adjust. The output transformer seems ok, and the amp did work fine for awhile before those two tubes failed. The load is ok....the customer was using a speaker in a known good cabinet, and I was testing with a known good resistor dummy load.

Thanks guys, I will try my extra tubes and run a signal through it and let it cook for a bit and see if I can see anything strage happening on the scope.

Greg

[Enzo]

And I might add, you may have bias voltage reaching pin 5 of the socket, but that doesn't mean pin 5 of the socket is making good contact with pin 5 of the tube.

[soundmasterg]

And I might add, you may have bias voltage reaching pin 5 of the socket, but that doesn't mean pin 5 of the socket is making good contact with pin 5 of the tube.

Yes I realize that Enzo, thanks. The last time the amp was here I retensioned and cleaned all of the sockets. I plan to check with the tubes partially inserted to make sure the tube is seeing the bias voltage when in socket before I do much else. At this point I am guessing that tube #4 was also bad and took out tube #3 even though it tested ok on the tester.....because it did blow the screen resistor previously. I won't get a chance to mess with it until wednesday AM, so I'll update after then.

Thanks!

Greg

[R.G.]

Some practical applications of theory:

- Redplating as you describe is ultimately a DC-conditions issue.
- Given that it is DC, it helps to realize that you can *remove the signal coupling caps from the grids of the power tubes* and you have not changed the DC conditions at all. Note that you may need to remove the PI tube or wherever AC feedback from the output returns in order to prevent oscillation issues if the amp is prone to it.
- With the coupling caps to the grids gone, nothing in the rest of the amp *can* affect the output tubes except the power supply.
- The OT does not significantly cause or contribute to this issue. It's DC voltage drop at zero signal is not a significant part of the DC voltage drop on the plate. And power beam tubes and pentodes are essentially constant current sources on their plates. External impedances there may affect the DC voltage, but not the current they draw.

If I were in an intellectual quest to understand the cause of the disease, I'd keep debugging. If I were under a big time pressure, I'd remanufacture the bias supply, and replace the tube sockets and resistors/caps around the output tubes. If I were under a lowest-total-cost constraint where I had to minimize *both* parts and labor, it would be tricky to pick the best direction.

This is one way of stating the truism that knowledge always costs you something, even if the cost is expressed in terms of your time.

[soundmasterg]

Some practical applications of theory:

- Redplating as you describe is ultimately a DC-conditions issue.
- Given that it is DC, it helps to realize that you can *remove the signal coupling caps from the grids of the power tubes* and you have not changed the DC conditions at all. Note that you may need to remove the PI tube or wherever AC feedback from the output returns in order to prevent oscillation issues if the amp is prone to it.
- With the coupling caps to the grids gone, nothing in the rest of the amp *can* affect the output tubes except the power supply.
- The OT does not significantly cause or contribute to this issue. It's DC voltage drop at zero signal is not a significant part of the DC voltage drop on the plate. And power beam tubes and pentodes are essentially constant current sources on their plates. External impedances there may affect the DC voltage, but not the current they draw.

If I were in an intellectual quest to understand the cause of the disease, I'd keep debugging. If I were under a big time pressure, I'd remanufacture the bias supply, and replace the tube sockets and resistors/caps around the output tubes. If I were under a lowest-total-cost constraint where I had to minimize *both* parts and labor, it would be tricky to pick the best direction.

This is one way of stating the truism that knowledge always costs you something, even if the cost is expressed in terms of your time.

Thank you R.G! That was very helpful in focusing my thoughts towards what is truly happening. Unfortunately I'm in a lowest cost/total cost constraint, and a time constraint on this one. The customer will be back from out of town next monday and needs a working amp for some gigs. At least I know the problem was localized to the one side of the OT and not both sides....that simplifies having to replace parts or sockets. I plan to work on it tomorrow morning and will see what happens with my known good tubes after I make sure of a couple other things.

Greg

[clyde1]

DC on the grids? Possibly a leaking coupling cap like Steve mentioned?

[R.G.]

Lowest total cost and time? Ack.

Remember the old adage: fast, good and cheap; you can have any *two*.

[soundmasterg]

Lowest total cost and time? Ack.

Remember the old adage: fast, good and cheap; you can have any *two*.

Yeah I remember it, but the customers almost never do!

[soundmasterg]

DC on the grids? Possibly a leaking coupling cap like Steve mentioned?

Checked that already. Coupling cap is a Mallory 150 and is just fine. No leakage...

Thanks,

Greg

[Gingertube]

First the blown screen resistor is a pointer to that tube having suffered excess current.

Next
With a single bias adjustment and no bias balance you must use a matched quad
With a single bias adjustment Plus balance you can get away with usin a matched pair on each side.

Changing a single tube is always asking for trouble unless you have individual bias controls for each tube.

If you have a tube tester you can select a suitable replacement for a single tube, remember to always do the grid current test as well as tesing gm and bias voltage AND test at as close to actual operating voltage and current as you can set on the tube tester, in guitar amps this will often be quite different than the test conditions listed in the tube tester manual for a particuular tube.

Because nearly every fixed bias guitar amp around abuses the max Rg1 specification for the output tubes testing for grid current is a MUST. When one tube exhibits excess grid current then bias is reduced and in a quad output it's pair (the tube on the same side off the push pull) will then also suffer a melt down.

Of course the worst abuse of max Rg1 values I ever saw in a commercial design was not in a Guitar Amp at all, it was in a very expensive JADIS JA80 Class A HiFi Amp, A quad of KT88 for each channel with the 2 tubes on each push pull side sharing a single 510K resistor, guess what? Those amps blow up spectacularly on a regular basis. I have a friend who keeps bringing them (2 mono-blocks) back to me to repair and each time I have to test about a dozen KT88 to find 4 with low enough grid current. I see them regularly every 18 months to 2 years.

Cheers,
Ian