Blonde Bassman strage behavior in the PI

[Dave]

Hi everybody,


I'm working on a Blonde Bassman. Its an interesting amp. Serial number dates to early '62 and it seems to be a transition amp. The circuit adheres to 6G6-A but it has the 8 ohm OT and tube rectifier. All the transformers are dated '61.


I gave it a good going over and put in new electrolytic caps. Everything looks good, clean, and original.


So, because I am a worry wart, before I fired it up, I added series diodes between the HT secondaries and the rectifier tube, flyback diodes to the power tube plates, snipped and shrink wrapped the 6.3 CT and replaced with 100r resistors, and added a fuse between OT and B+. Obviously, I was worried about losing one of those transformers.


Well, the amp works and sounds good... But... If I hit a big bass note with the volume up past 5, the signal cuts out in a weird way.


The attack is strong. Note sounds for almost 1 second and then drops out for about half a second and then comes back. There is no distortion (audible) with the drop out. It almost sounds like somebody grabbed the volume control and did it manually.


It does it on both channels which leads me to believe that it has to be in or after the PI. I checked all tubes and everything I can think of.


I hooked up my meter to the plates of the PI there was about 300 volts there. With the meter running, I hit a big E and watched the meter. During the drop out, the meter dropped to 60 volts and then back up to 300.


What could I be missing?

- Update... Its seems to be only dropping the voltage on the plate fed by the 82K resistor. The resistor measures good although I'm going to change it anyway just to make sure. I thought maybe the coupling cap was leaking, swapped it and there was no change.

Dave

[tubenit]

Yrs and yrs ago,  I had a resistor that would measure "good" with my meter, but when I hit a bass note ........... the resistor had a broken wire inside the enamel where you could NOT see it and the wire would open and close with the vibration from the bass note causing it to cut in and out. 

It was the hardest amp problem to diagnose that I ever had because it was so intermittent.

I am not saying this is the issue with your amp.

With respect, Tubenit

[Dave]

Changed plate load resistor - no change.
Changed cathode resistor - no change.


Plate voltage after 82k - 330 volts - during drop... 50 volts
Plate voltage after 100k - 330 volts - during drop ... 270 volts
voltage before the resistors 420 volts - during drop ... 400 volts


Could this be a bad filter cap?


It seems like that when the tube is drawing current to keep up with the load on it, its dropping tons of voltage. I tried another tube and it did the same thing.


Dave

[Dave]

Changed the filter cap on that node. No change
Eliminated flyback diodes (just to see). No change
Eliminated B+ fuse (just to see). No change.


I'm lost.


Dave

[DummyLoad]

have you swapped out the coupling caps?


--pete

[sluckey]

Disconnect the NFB wire. Does that change the behavior?

[Dave]

I changed both coupling caps - No change.
Lifted NFB - No change.


Now I am really at a loss. I don't see anything else it could be.


Dave

[clyde]

http://el34world.com/Forum/index.php?topic=20922.msg221544#msg221544

[sluckey]

What resistance do you measure from pins 3/8 of the PI to chassis?

What are the voltages on pins 2, 3, 7, and 8? How are they affected by your problem?

[Dave]

Almost 13k from paralleled cathodes to chassis.


322 volts on pin 2 drops to about 50 volts while dropping out.
21 volts on pin 3 drops to about 12 volts while dropping out.
320 on pin 7 drops to about 270 volts while dropping out.
22 on pin 8 drops to about 17 volts while dropping out.


Clyde, did you ever find resolution?


Dave

[Dave]

I found some references online to very similar problems from other people. I finally found someone who said that the earliest circuits had no decoupling cap into the PI and this was responsible for causing blocking distortion. I checked my amp, and sure enough, it doesn't have one. This may be the solution. I will install the cap as per the 6G6b schematic and report back.


Dave

[sluckey]

320V on pins 2 and 7 is bad wrong! That's probably the problem. One or the other of those plate couplers from the two preamps is leaking badly. Voltage on these pins will be less than the voltage on pins 3/8 when working properly.

A coupling cap on the PI input like that on the 6G6B is not needed because the .1 and .05 couplers from the preamps do the same thing when not leaking. Putting a cap there will be a band aid fix but it simply masks the real problem.

If you compare the 6G6A and 6G6B you will see why you don't need the cap on the PI input for the 6G6A.

[Dave]

No, that was a typo... Sorry. Both plates had about 320 on them. The grids had 20ish volts.


The capacitor I mentioned above solved the problem. It is fixed now and working great.


Thanks for the help. Actually when I realized that Clyde was having a very similar issue, it prompted me to search on the issue itself thinking maybe it was common. Sure enough, it is a common complaint with both the 6G6 and 6g6A circuits. The 6G6 B circuit addressed the issue.


Dave

[clyde]

No Dave,
I wasn't able to resolve the problem, just mitigate it somewhat.  It was only the one channel so I'm not sure if our problems are the same.  As well, I can't see that adding a cap will solve the problem if the two coupling caps as mentioned by Mr. Luckey are working as they should.  I don't have the amp anymore but I'm sure I can get it back and see if that will fix my issue.  I hope yours is resolved. 

[Ed_Chambley]

Not sure if you are keeping this amp, but I have an early 62 6G6B.  If you placed cap after the mixing resistors, you will add additional series capacitance.  To my knowledge, the 6G6A amp had no problems per say.  It is my understanding the revision was to make the Bass channel pass lower frequencies and give the treble pot more control, not that anything was wrong with the design.

If you are adding a cap after the mixing resistors, will it not add additional series capacitance to both channels?

[HotBluePlates]

...  As well, I can't see that adding a cap will solve the problem if the two coupling caps as mentioned by Mr. Luckey are working as they should.  ...

This is likely the real answer.  The coupling caps prior to the 470kΩ mix resistors should probably be replaced, one at a time, to figure out which is leaking.

Unfortunately, measuring the phase inverter grids directly won't give a useful reading because the meter loads the bootstrapping and drops the (measured) voltage.

I dislike replacing a coupling cap in a collectible amp as much as anyone, but that appears to be what's called for here.

[Dave]

Thanks for the additional input guys. Ed, Yes, I am adding additional series capacitance to both channels into the PI, but... both channels were suffering from the same problem.


Many of the complaints I have read online have been that the problem only showed up on their bass channel, but mine was doing it on both.


Could very well be that one or both coupling caps on both channels are leaking, but at this point, the small additional capacitor has resolved the problem entirely and the amp sounds fantastic as it is.


I think for now, I am going to leave it like it is because it sounds so fine and, of course, to echo HBP, hate to change out those caps.


Dave

[DummyLoad]

your fix was is a "band aid": please view attached.


what you did do is add another RC time constant. if you're satisfied with it as-is, then no worries. if you want it stock, find the leaking cap and remove the added cap. 


--pete

[Dave]

Yes, I was just about to reply back saying that I had responded without rubbing the proper brain cells on the situation. All three of you are obviously correct. If those two coupling caps are blocking all the DC up to that point (which they should be), there is no need for another cap to do the blocking for them.


I got it. I'll suck it up and replace the coupling caps when I get a chance.


Dave

[DummyLoad]

If those two coupling caps are blocking all the DC up to that point (which they should be), there is no need for another cap to do the blocking for them.

there are THREE caps blocking DC - the third cap under the pot blocks DC from g1 of the PI to ground. that cap lets the pot "float" in the DC bias of the PI.

respectfully,

--pete

[plumcrazyfx]

I have one of these really early Bassmans- mine is early enough to still have a tweed Bassman power transformer. I think they use the same output transformer as the early blonde Twins.  I would love to find one that needed to be fixed so I could have Mercury make a run of them.

[clyde]

If those two coupling caps are blocking all the DC up to that point (which they should be), there is no need for another cap to do the blocking for them.

there are THREE caps blocking DC - the third cap under the pot blocks DC from g1 of the PI to ground. that cap lets the pot "float" in the DC bias of the PI.

respectfully,

--pete

I got the aforementioned amp back and replaced all three of the suspected culprit caps and no improvement.  I then did what Fender did-turned it into a 6G6B, end of problem!