Troubleshooting Selmer T'n'B Mk III

[EFK]

Yes, PI is V4. If I pull V2 - the last gain stage for both channels, looks like one half of 12ax7 only is used - the amp goes silent but DOES NOT function. Sorry if I was not clear on that - that would in fact be weird wouldn't it? Voltages on V1 through V3 are where they should be, @ 130 V on plates and 1 V on cathodes. I'll double-check today.

6.1 does seem a hair low for the heaters but that really shouldn't be a problem, should it?

[MWJB]

Hang on, you have:

V1 - input & tone stack recovery for one channel, then ...

V2 - gain stage for BOTH channels, then...

V3 input & gain stage for the other channel, then...

V4 PI tube.

...?

Sounds like a recipe for osscilations, the 2 channel input & tone stack recovery tube should be V1 & V2 surely? Can you rewire them like this?

Low heaters may/may not be a problem, less than 6.3VAC can be sign of excessive load on the PT. You are using the 250VAC PT primary (if in EU)?

[EFK]

Yes, stock design on this amp (preamp has never been touched):

V1 (Bass channel): Input signal pin 2 ---> Tone stack ---> Volume ---> signal into pin 7 ---> 470K mix resistor....

V3 (Treble channel): Input signal pin 2 ---> Tone stack ---> Volume ---> signal into pin 7 ---> 470K mix resistor....

From the 470K mix resistors, the signal from either channel then goes through a .022 cap into one half of V2 which is the last gain stage for either or both channels ---> then on through another .022 cap to the PI.

Early on in this thread, in a discussion about the echo loop, it was suggested I could remove the 1M resistor which FOLLOWED the two 470K mix resistors prior to the signal from either channel entering V2. In other words, the signal from either V1 or V3 ---> 470K resistor ---> 1M resistor ---> V2. It was NOT a ground reference resistor or anything like that - see schematic, the signal off the 470K mixers then went *through* the 1M. I am still unclear as to what purpose this 1M resistor would have served, but just for fun I'm going to put it back and see what happens. Wouldn't this attenuate any signal going into V2?

I'm using the PT on the 120 VAC setting as I am in the US. It has a 115 setting too but I haven't tried it as the wall voltage here is generally @ 119-121. I suppose if I used the 115 setting it would jack up the heater voltage a bit.

[MWJB]

If you try 115VAC make sure heaters don't go over 7VAC, B+ will also rise so you'll need to rebias...I don't think at this point that low heaters is specifically the cause of your issues ...so perhaps save this for a rainy day?

Perhaps the 1M resistor was there specifically to attenuate signal...& keep oscillations at bay?

If your amp is laid out physically like the amp in Steve O'Cs photos, you might end up just "fighting the tide" as it were...Selmer T'n'B based on the brown tolex bassmans seem to have a much better layout, the SV in Steve's photos is not laid out well at all...I'm amazed that ANY of them worked at all!

[EFK]

You know every time I prepare to try something, I think, "This is going to be it!" And yet, it's always the same. The 1M resistor seems to have no effect.

Now here is where it gets even weirder if that is possible.

It seems like more of the hum is coming through the treble channel as opposed to the bass channel. W/ both volumes down, there is humming. Now if I turn the bass channel volume up to about noon, the hum almost completely vanishes. There is still crackling if I touch the knob, but the hum is practically gone. If I continue to turn the bass vol. knob up to full, some of the hum comes back but then it's on full so I would expect some noise. Now, bass volume back to 0. Turning up the treble channel volume, the hum gets louder and louder. Aha! So I return both volumes to 0 and think, I'll pull V3 (treble channel triode) and I'll bet the hum goes away. No banana. Doesn't seem to make any difference other than that the treble channel is then cut out.

As a side note: can resistors be tested in-circuit? When amp is powered down and drained, the resistors all measure the values they should. Is there a way to measure resistance under power, to see if maybe something is drifting or not working properly?

Perhaps I'm approaching this incorrectly; perhaps the solution is to get drunk enough that the amp sounds good.

[MWJB]

Some resistors can be read in circuit...others not, because of parallel paths to ground skewing the reading. If you suspect a power supply resistor, power up & check voltages, and check mA drawn through the resistor (stick it under load & play an I-pod/cd player through it if you like), 12AX7 draw about 1mA per triode in a lot of circuits, anything from 0.7mA to 1.5 might be reasonable. Really, if you suspect a power supply resistor just chage it for a 3W metal oxide & be done with it.

Did you ever measure the dc voltages at all power tube (pins 3, 4, 5) & preamp (pins 1, 6, 3, 8) sockets/

Have you ever heard the amp function as you are expecting it to?

Will it not be possible to swap V2 & V3 wiring?

[MWJB]

Just a thought, have you tried bypassing the echo loop triode? Connect the junction of the 2x470K mixer resistors (output from V1 & V3) directly to the 0.022uf cap at the PI input, lift this cap away from the echo loop plate to which it connects.

[EFK]

I've got the echo loop completely out of the circuit. That was one of my first thoughts. As it stands now, I put back the 1M resistor following the 470K mixers, then have that going through a .022 cap directly to LTPI input. I don't know if the 1M/.022 really need to be there, but I thought better safe than sorry.

V1: Pin 1, 120. Pin 6, 123. Pin 3, 1.27. Pin 8, 1.33.
V2: Pin 6, 139. Pin 8, 1.24.
V3: Pin 1, 132. Pin 6, 130. Pin 3, 1.20. Pin 8, 1.22.
V4: Pin 1, 200. Pin 6, 211. Pin 3/8 tied: 22.3.

Today I also tried completely bypassing V2 (last gain stage) - I disconnected a few things and ran a jumper wire so that the signal came right off the 470K mixers, bypassed V2 and went straight to the PI. No difference!!!!!

The only way to shut this amp up is to ground out the PI.

If I set my meter to the <200m VDC setting, I can read anywhere from 3 to 6 mV on the pot lugs *both channels* (measurement changes as I move the bass and treble pots). Is this acceptable? Or, is this the source of the scratchiness?

Hey by the way, not to sound stupid, but looking back to basics: is there any problem (for testing purposes) with using a speaker cable that plugs into the amp and hooks to the speaker with tight gator clips?

I have never heard this amp without all this noise. Only had it a few weeks.

[Steve Conner]

If you haven't any previous experience working with old tube amps, you might find the level of hum and hiss intolerable, when in fact it's perfectly normal.

MWJB mentioned that the grounding in these amps is known to be bad. Why don't you try my earlier suggestion of removing the connection to chassis at the HT fuse holder, and relocating it to the negative terminal of the main 50uF can. Or omitting it altogether, there may already be a connection to chassis through the brass plate on the front. But be careful leaving out ground connections, that things don't become live that shouldn't. You don't want -400V on all your heaters.

If that doesn't work, just take it to a gig, you won't hear the hum over your drummer. It's maybe asking a bit much to expect a vintage 50 watt amp to be docile enough for home use.

[EFK]

Steve it's likely the truth. All of the amps I've used have been new, almost silent. I've reduced the hum in this quite a bit through shielding grid wires and grounding as per all of the suggestions. The HT fuse holder had two ground connections, one a later add-on for no apparent reason, and I removed that one. I've moved a lot of grounds to the filter cap neg. and moved that chassis ground from the brass plate to a lug next to the PT. All of this has helped. At the moment, there really isn't much else to do except removing the brass plate and running an added ground bus for the preamp and tying all of it back to a lug near the inputs. But then the pots would have to be isolated from the chassis and all of this seems like a ridiculous amount of work.

My real concerns right now are (1) the low 6-7 VDC on the PI grids and (2) what is with the squealing and crackle/scratchy sounds on brand new, thoroughly-grounded pots? I feel like they are picking up DC somehow, but for the life of me I cannot figure out how.

Last night, I had a thought: any chance there could be solder which has bled through the eyelets and is making improper connections *under* the board?

[Steve Conner]

Well, have you tried replacing those mustard capacitors yet? If brand-new pots are still scratching, that suggests leaky coupling capacitors. I have no experience with the mustard ones, my Selmer had green tubular caps in it.

Leaky caps might also explain the low PI voltages.

Can you measure the following:

PI supply node voltage
PI plate voltages
Values of the PI plate load resistors (they had drifted up considerably on my amp)
Value of the PI tail resistor

[EFK]

Steve - every cap in the amp has been replaced along with the pots.

I've got 303 VDC on the PI rail prior to it hitting the 82K and 100K resistors (both of which meter pretty darned close to said values). Plate voltages are 200 and 211. The schematic says the tail resistors off the cathodes are 820 ohms and 6.8K, and sure enough they both meter that. Cannot figure out why the grid voltage is so low!

Here is something interesting, maybe normal or not I don't know: if I stick a test probe at the junction of the line off the PT into the bias circuit, where it hits the 180K/2.2M resistors prior to the BY105 diode (see schematic), the voltage fluctuates very rapidly between -2.80 and -3.80 vdc. I'd say every half second, back and forth, back and forth. Don't know if this means anything.

Also, if I set the meter to <200 millivolts dc and ground one end, and then move the probe around the area of the choke and OT *in the air!!!!* I pickup very low millivolt readings, like 1 to 4 mV, just out of the air. Is this normal? The choke and OT are very close to each other, almost touching. They're also directly under the treble channel of the circuit and the treble pots, which seems to be the noisiest area of the amp.

The humming has gotten better with all the shielded wire I have run and it does quiet down more when I lay a sheet of aluminum foil over the chassis while it's out of the amp. But it's still there - what is still there is definitely a 120hz power ripple and it vanishes the instant I switch the amp over to standby.

And through it all, these damn pots are still scratchy and screechy!!!!!! Could the choke, OT or PT be somehow radiating a small DC charge into the pots, either through the air or through the chassis? I get absolutely 0 dc reading on chassis and all grounds.

[MWJB]

"Cannot figure out why the grid voltage is so low!"...it does seem a little low, but your PI plates don't run at particularly high voltage and that 6.8K tail resistor is relatively small...might be ball park voltages on the grids?

Do you have room to move the OT away from the preamp tubes & inputs?

[EFK]

Well I think I'm done chasing the hum. Between all the shielded wire I used and all of the grounding advice you guys have given me, it's not bad at all for an old amp. What's left of the hum probably has something to do with either a really old PT, old choke, or maybe cramped quarters wherein the PT, choke and OT are all practically on top of each other. Probably some lead dress issues to but at this point I'm living with it.

But the scratchy pots!!!! This is driving me nuts. Actually, they are not scratchy in the classic sense - when I grip the knob and turn it, there is no scratchiness from turning. The static and crackling is occurring when I touch or let go of the knob (volume only!). With the volume pots turned up higher, it actually squeals when I touch it. It's fine when I let go, no static or anything when not touching them. There is absolutely no dc or ac voltage on any of the grounds or chassis, but I think this is a grounding issue somehow and maybe I'm becoming a better ground when I touch them? I can get static and noise just tapping on the volume knob, almost exactly like tapping on a microphone or stethoscope. Ah well. I'm going to start playing it and restrain myself from poking around in there much more. hpefully I'll get a good sound and won't have to fiddle with the pots!

[MWJB]

You check for dc on caps by lifting the end of the cap opposite the B+ supply, connecting a voltmeter between the free end of the cap & ground & measuring the dc, replace if over 0.25vdc. As well as any coupling caps, check tone caps too.

Static - SF Fenders often have similar symptoms, it's usually down to poor lead dress. Does the amp make unwanted noises when you are playing through it, or just when you touch the vol knobs?