B+ unstable (=slowly decreasing) in old amp

[slidincharlie (Carlo P)]

A friend borught me his Gibson BR-6 (the two-jack version from the late 40's) with a field-coil speaker. He bought it recently; probably this amp had been sitting in a closet for a long time. The guy told me that, while playing it cranked for the first time after purchase, the volume dropped down at about half its former perceived level, and did not went fully up again.
Now I'm trying to find out the cause of this volume drop, and also wnat to overhaul the amp.
First thing I did was reading voltages in the PS and power section with its stock tubes (which are very old, most likely as old as the amp), and I noted a progressive decrease of all voltages, in the following way:

- at start:
plates of 5Y3: 370 VAC;
B+ on the OT center tap: 250 VDC;
plates of 6V6's: 246 and 242 VDC.

after 15 minutes:
plates of 5Y3: 365 VAC;
B+ on the OT center tap: 223 VDC;
plates of 6V6's: 221 and 211 VDC.

The above readings continued to decrease for further 15-20 min., then I switched the amp off.

The same behaviour, with slightly different voltage values, was displayed with new NOS rectifier and power tubes.

The amp is apparently bone stock.
A schematic is here: http://hem.passagen.se/ekabjan/Amps/...tics/br-6f.jpg .

What can cause this progressive decrease of voltages inside an amp? Maybe a drifted resistor, or one that drifts under heat?

TIA,
Carlo

[Amp Kat]

Not trying to be a smartass here but after several years wouldn't you maybe suspect the power tubes may be shot ? They don't last foreever and the first sign of burn out is loss of treble and power drop-out. I'd re-cap the whole amp as far as Electrolytics and retube and rebias it and you should have a nice sounding and long lasting beaut. The preamp tubes are most likely all good.I'm sure some others have some expertise advice as well.

[Marc]

carlo,
since you don't say if the electrolytic caps have been changed I'll assume they have not. So, the caps are probably at least a little leaky and the resistors may have drifted and/or they may drift more once they heat up I'd look to those possibilities, If you have a known good pair of 6V6's you could try that too. You may also want to check the coupling caps for leakage-a 50 year old cap is automatically suspect in my book.
In short give it a complete overhaul-unless you want it to be stock (and unplayable).
Hope for something simple, it could be just one or two resistors that are out of whack.....
Good luck!
Marc

A friend borught me his Gibson BR-6 (the two-jack version from the late 40's) with a field-coil speaker. He bought it recently; probably this amp had been sitting in a closet for a long time. The guy told me that, while playing it cranked for the first time after purchase, the volume dropped down at about half its former perceived level, and did not went fully up again.
Now I'm trying to find out the cause of this volume drop, and also wnat to overhaul the amp.
First thing I did was reading voltages in the PS and power section with its stock tubes (which are very old, most likely as old as the amp), and I noted a progressive decrease of all voltages, in the following way:

- at start:
plates of 5Y3: 370 VAC;
B+ on the OT center tap: 250 VDC;
plates of 6V6's: 246 and 242 VDC.

after 15 minutes:
plates of 5Y3: 365 VAC;
B+ on the OT center tap: 223 VDC;
plates of 6V6's: 221 and 211 VDC.

The above readings continued to decrease for further 15-20 min., then I switched the amp off.

The same behaviour, with slightly different voltage values, was displayed with new NOS rectifier and power tubes.

The amp is apparently bone stock.
A schematic is here: http://hem.passagen.se/ekabjan/Amps/...tics/br-6f.jpg .

What can cause this progressive decrease of voltages inside an amp? Maybe a drifted resistor, or one that drifts under heat?

TIA,
Carlo

[slidincharlie (Carlo P)]

As I said in the starting post, I already tried a new pair of 6V6 and a new 5Y3 (all Philips) to no avail: I still can detect a progressive decrease of B+ until it tends to stabilize at around 20-30 volts lower about 30 minutes after startup.
I have also replaced all the filter caps.
I still have to check the resistors in the power supply, the cathode resistors and the coupling caps. Hope to find simply a drifted resistor or a leaky cap (but boys, ain't that chassis messy! there is almost no room for working in there! In comparison Harmony and Silvertone amps are much much easier).

[Amp Kat]

Ok Carlo, I saw down there where you did say you replaced the tubes with NOS so my bad but they may also be flaky but most unlikely. Your almost to the point where your going to need a scope to find out where the signal is dropping. It's normal for the B+ to drop after some time when the tube settles down and heats up good especially in new tubes and 20 to 30 volts is about right. I'm not sure you have a B+ problem more than a cold solder joint intermittent problem. You may want to touch all the solder joints on the 6V6 tube sockets and maybe the 6S?7 sockets as well. If you can read 200 ohms from the 6V6 cathode to ground your connection is good. Another thing to try is measure the DC voltage across the Cathode 200 ohm resistor and it should be around 20 to 30 volts. If it's not you have a problem there with either a connection or resistor open or shorted. If your plate voltage drops you should see it across the Cathode resitor and divide it by the 200 ohms to see how much current is actually dropping off. If I didn't have a scope I'd take a ts jack and connect a lead from ground to the sleeve and a wire right after the last coupling cap to the tip and send that to another amp or power amp and see if the volume still drops then the problem is back in the preamp. If it doesn't drop it's in the power section and at least you've cut your work in half. I'm thinking solder joint but can't say for sure without it in front of me.

[Satamax]

Hi everybody!

Carlo, do you have a digital camera doing IR or an IR thermometer? Check for something heating up, iirc the B+ goes through the field coil, check if it isn't this part shorting to something, and heating up. Touch the parts you can touch when the amp is slightly warm, and then recheck when the amp is fully warmed up and droping the voltage.

Bye.

Max.

[slidincharlie (Carlo P)]

A little update and a few odd things

Re: heat.
When the amp has been on for about 30 minutes and the B+ as well as the plate voltage on the 6V6's has dropped by about 30 volts, the power tranny is definitely warm (maybe around 100°F) and the 220k cathode resistor on the 6V6 is hot.

One odd thing ("ghost" resistor):
there is a big 20k resistor (the old-style brown ceramic hollow tube type: maybe a 10W?) between pin 4 and ground on V3 (the first power tube). This resistor is NOT on the schematic, and it also gets hot when the amp has been on for 20-30 minutes. What is the function of this big resistor?

Another odd thing:
there is no bypass cap on the 6V6 cathode resistor. Should I add one?

Last odd thing (low B+):
isn't the B+ (and plate voltage on the 6V6's) dramatically low? I've never seen an amp (even a low wattage old amp) with such low values. Can this be considered normal, at least for this kind of amp?

P.S.: I haven't checked for leaky caps. I did check all the resistors and a few have drifted a bit. I think I'll replace all "critical" resistors anyway.

[Amp Kat]

Re: heat.
When the amp has been on for about 30 minutes and the B+ as well as the plate voltage on the 6V6's has dropped by about 30 volts, the power tranny is definitely warm (maybe around 100°F) and the 220k cathode resistor on the 6V6 is hot.

One odd thing ("ghost" resistor):
there is a big 20k resistor (the old-style brown ceramic hollow tube type: maybe a 10W?) between pin 4 and ground on V3 (the first power tube). This resistor is NOT on the schematic, and it also gets hot when the amp has been on for 20-30 minutes. What is the function of this big resistor?

The transformer will get pretty warm dependending on it's design and that Cathode resistor will get very warn too. The resistor from pin 4 of the tube is the screen resistor and should be going to pin 6 which isn't used and then tied to the plate with a wire, not ground ! 20k is much to large for a Screen resistor. Are you sure it's not 2k ? They are normally 470 ohms to 1K 5 watters but 10 watts isn't a bad thing. If it's open regardless of the value the tube will not current at all and you would get no sound. If it was going to ground it would create a short and the B+ would most likely be zero and the fuse would blow so somethings not right there.

Another odd thing:
there is no bypass cap on the 6V6 cathode resistor. Should I add one?

That's up to you as it's more of a frequency thing. Some like it ,some don't !

Last odd thing (low B+):
isn't the B+ (and plate voltage on the 6V6's) dramatically low?

Yeh that is pretty low but it could be a Class A push pull amp where the plates would run at about 250 volts.

I've never seen an amp (even a low wattage old amp) with such low values. Can this be considered normal, at least for this kind of amp?


It is a tad low but could also be in the deisn if it's a class A amp. Amps designed for class AB would have a higher plate voltage.

P.S.: I haven't checked for leaky caps. I did check all the resistors and a few have drifted a bit. I think I'll replace all "critical" resistors anyway.

Let us know on that screen resistor and exactly where it's going and it's ohmage value. Check with your meter diectly across it for it's true value.

[slidincharlie (Carlo P)]

Odd resistor!

The resistor from pin 4 of the tube is the screen resistor and should be going to pin 6 which isn't used and then tied to the plate with a wire, not ground ! 20k is much to large for a Screen resistor. Are you sure it's not 2k ? They are normally 470 ohms to 1K 5 watters but 10 watts isn't a bad thing.

Amp Kat,
I know screen resistors, and this does not seem a screen resistor to me!
I confirm: it is a 20K and goes from pin 4 to pin 1, which in this case is a tie point for ground. I checked for continuity, and I am sure that pin 1 in this 6V6 goes to ground.
Please have a look at the picture below.
The resistor is absent from the schematic (for which I provided an URL at the start of this thread). But I've read that old Gibson schematics often did not match their amps...

One further detail:
pin 4 in both 6V6's receives B+ from the OT center tap (that is, a red wire that carries 250VDC). I suspect this is a "non-standard" configuration if compared to, say, Fender amps, am I right?

[Amp Kat]

Somethings not right there Carlo. In the schematic the Screen is tied to B+ Which could really use more filtering than that little 10uf cap and a choke as well but if it were me I would take that resistor off of pin 4 to ground and wire it up like the schematic. You also said the Cathode resistor was 220 ohms and the schemo shows 200 but 220 is cool I just wanted to point out more differences in amp and schemo.

[Enzo]

CArlo, you have given us a bunch of voltages, but you left out the most important one. WHat is the voltage at the cathode of the rectifier? The rectifier and first filter ther. Does that voltage drop too or does it stay up? The field coil of the speaker acts as a choke here, and I would not be surprised if the field coil is warming up and its resistance climbing a bit. That would mean the voltage across the coil raises. So in that case, the voltage at the first rectifier would remain up while the voltage at the other end drops. Does it?

RUnning the B+ directly to the screens and also to the OT center tap was pretty common in small amps back then. The furst filter was a 20, the field coil choke and the second filter 10 smoothed things out pretty well, besides, the push pull cancels most of the ripple anyway. The third filter for the preamp tube makes that pretty clean.

That 20k to ground from the second filter is at best a bleeder, and at worst, someone's idea of a way to reduce the B+ voltage. That's what I see anyway.

A completely different idea now. leaky coupling caps. What voltage is at the grids of the power tubes when cold and when warm? If the caps are leaky and after a moment leak more, then the power tubes will have a small positive voltage on them, so they conduct harder, which drags the B+ down. Either measure at the grids looking for DC there, or just disconnect the caps to the power tube grids and see if the voltage stabilizes.

[Amp Kat]

You know I forgot about that field coil Enzo as it's been a while on that set-up but those guys back then did some freaky things that you always seem to run into that throws a wrench into the mix. Seems like every amp has it's own little twist but I can't see those 20k's helping out in any way.

[slidincharlie (Carlo P)]

Enzo,
I guess you nailed the main problems here.

A completely different idea now. leaky coupling caps. What voltage is at the grids of the power tubes when cold and when warm? If the caps are leaky and after a moment leak more, then the power tubes will have a small positive voltage on them, so they conduct harder, which drags the B+ down. Either measure at the grids looking for DC there, or just disconnect the caps to the power tube grids and see if the voltage stabilizes.

I checked the coupling caps for leakage, and they ARE incredibly leaky: I read 36, 86 and 102 VDC respectively on their distal end. After 30 minutes they were leaking even more (45, 100 and 116 VDC respectively).
Please confirm if I used the correct test technique: I disconnected the grid end of the three caps and kept them disconnected during the 30-min test.
On the other hand, if I leave them connected, I see 4 and 9 VDC on the grids of each 6V6, respectively (which probably accounts for the different plate voltage on them, like if they were mismatched - see the start of this thread).

WHat is the voltage at the cathode of the rectifier? The rectifier and first filter ther. Does that voltage drop too or does it stay up? The field coil of the speaker acts as a choke here, and I would not be surprised if the field coil is warming up and its resistance climbing a bit. That would mean the voltage across the coil raises. So in that case, the voltage at the first rectifier would remain up while the voltage at the other end drops. Does it?

The rectifier gives 387 VDC, and this is what I see on the first filter cap. With the coupling caps connected I see what you have foreseen: the voltage at the 5Y3 and at the first filter stays up (well, it decreases by a few volts actually) while the B+ drops dramatically.
BUT... with the grid end of all filter caps disconnected, the B+ stays nearly constant (it drops only by about 10 V after 30 minutes actually), and it's the same on each 6V6 (it's a matched pair).
Should I expect that everything will be okay with fresh coupling caps? (the caps currently in the amp must be original, that is almost 60 yrs old).

That 20k to ground from the second filter is at best a bleeder, and at worst, someone's idea of a way to reduce the B+ voltage.

May I safely disconnect this resistor and see what happens?
If tone is better with that resistor in place, may I leave it?
If on the other hand tone is better without that resistor, may I toss it?
If it was meant to keep the plate voltage down, I think I might leave it after all (after checking bias AND tone).
If it was meant as a bleeder (and probably it is, since the residual HV after switching the amp off drops almost immediately to a few mV), what should I do with it?

[Enzo]

If those interstage coupling caps are leaky, replace them before wondering about the amp any further.

If you see any positive DC voltage on the grids of the power tubes, other than a shorted power tube, there is only one place that DC can come from - through a leaky coupling cap. Disconnect the coupling caps, but leaving in place the grid to ground resistors, and if the voltages now stabilize, you just verified it.

I bet new couplers will take care of things. Chances are most of them are leaky.

As to the power resistor, it isn't on the schematic, so I doubt it is needed. Disconnect it and see what happens. If it was intended to reduce the plate voltages, we have to ask, do the voltages actually need reducing? A lot of old amps were designed with 110VAC mains in mind, so now with today's 120VAC mains, all the voltages come out 10% higher than the schematics say. This may or MAY NOT be a problem. If the B+ is only 250, I doubt another 20 volts matters much.

If it is a bleeder, it is nice to have, but simply exercising caution in your work can take its place.

Tone? WHo knows? FInd out for yourself.

[slidincharlie (Carlo P)]

Fixed! The problem is gone

Replacing all three coupling caps (badly leaky) with new caps did the trick. Now the voltages are constant and the amp has an overall higher volume than before (this might be due also to the new filter caps I installed).
I used three Orange Drops that I had in my parts box, even though Mallory 150's would have been probably better.

I left the big 20k grid-to-ground resistor on the first 6V6 in place. It served to lower the B+ by about 10V as well as as a bleeder.
I tried to disconnect it and listened to the amp: it seemed to have a tad less breakup and maybe a very little more high end. I preferred to leave it in place (even though that resistor wasn't in the schematics - but inconsistencies between schem and amp were very common in old Gibsons).

Thank you all.

Carlo