Selmer T'N'B evolution - question

[EFK]

Take a look at the two attached schematics for a Selmer Treble N Bass 50. The earlier version, MKII, only has (3) 12AX7 preamp tubes, whereas the later MKIII model added an effects loop and thus a 4th 12AX7 to drive it. So, the earlier version only has two gain stages, both sides of a triode per channel, while the later version adds a third gain stage that is shared by both channels (one half of the extra 12AX7 only) following each channel's respective signal travel through both halves of its particular 12AX7.

A few questions: (1) Does it matter *which side* of this extra triode is used? For example, if the signal coming off V1 or V3 (each channel, respectively) is on the side of the triode using pins 6/7, should the next stage enter in on 1/2 or 6/7? Or does it matter? Thinking of oscillation potential here.

(2) If you look at the MKIII schematic, when they added the extra stage, they added an extra resistor AFTER the two 470K mixer resistors off each channel. The signals from the respective channels travel through each 470K mixer, then prior to entering the extra gain stage of the MKIII they have to travel through a shared 1M resistor. WHY? Was this added to knock the signal back a bit to avoid excessive gain? I'm very interested in *why* this may have been added, and can it affect the sound other than to reduce the strength of the signal prior to entering a final gain stage? Or can it be serving some other purpose? They also added a .022 cap just prior to the last gain stage, which seems to defeat the purpose of using larger .047 caps as couplers? Isn't the .022 just going to further cut low frequencies? Maybe a slap-dash adaptation to orient it more towards a guitar amp as opposed to a bass amp?

I realize these are burning questions which are of intense global interest. I've been dwelling on them, that's for sure.

[Enzo]

1. No.

The signal at the plates of the second triode in each channel is WAAAY too hot for a line level effects loop. You wouldn't want 40 volts of signal going into your effects unit, would you? So those large value resistors combine with the 47k resistors to form a voltage divider and knock the signal level down to a fraction of what it was.

SO then a nice polite level of signal can flow through the FX loop.

Since the circuit is otherwise about the same, the signal returning at the loop is now waaaay to low for the PI stage. So the extra triode is gain to restore the signal level to what it was before the loop stages.

The cap? Well, you want it there to block any DC coming in the loop return from getting on the grid of the tube. As to the low frequencies, remember, it is all about the circuit, not about the parts. You can;t say what a cap all by itself will do to the frequency response, it is only when it gets in a circuit with other impedances that you can determine this.

And all thing equal, yes, a .047 cap would pass more lows than a .022. But what was the low freq cutoff of the larger cap in its stage anyway? COnsider when messing with cathode bypass caps that raising the cap value might push the bottom end down in freq, but if the change is from 5Hz to 2 Hz, what differtence does it make to the signal from a guitar? In other words, we need to determine WHAT you would be missing before we start missing it.

[EFK]

Thank you Enzo for the very clear responses. I guess my next comment would then be: if the effect loop were to be removed from a MKIII, then there is really no need for the extra half of the triode is there? If you were to pile that extra stage on top of the signal from either channel without 'making it polite' as you put it, in other words fed the signal off the 470K mix resistors right into the extra triode, is it likely to become a squealing pig of over the top, gain-y noise?

[Steve Conner]

Sort of. Before that effects loop stage, you have a potential divider. It's the 1M+470k, and the 47k on the tip of the echo out jack. If you didn't remove the 47k, then the signal is still getting knocked down before it gets amplified back up, so the gain won't be any different.

If you did remove it, then the signal path has about a factor of 20 more gain than it should, and that probably explains your noise problems.

If you don't want an effect loop, you can leave that triode stage out, and rewire the amp as per the Mk2, and it should be less noisy than a properly working Mk3.

[EFK]

Steve you're on to me! I'll bet it was just a little hum, and adding all that extra gain on there amplified it something like 300%!! It needed a rewire anyway and I'm taking my time - also using the MKII schematic. I see there is a subtle difference in that the MKII uses a slightly larger resistor to ground in the tone stack (which according to Duncan's calculator will make the mid dip a bit shallower) and also the treble channel has slightly different cathode resistor/cap values. The schematic is not clear, but I have a good picture of stock MKII board and one of those bypass caps is a .1!!! That's going to be bright.

Anything interesting I could do with an unused 12AX7?

[EFK]

BTW - question on diodes: the BY105 and BY128 are apparently archaic because I can't find any information on them. Somebody replaced the two BY128 rectifiers with hexfreds. Don't know what they are, but I have some 16A/1200V which look identical. Are these suitable? Is there is a problem with freds in this situation?

Re: the BY105, which deals with the bias circuit: would a plain old 1N4007 work here?

[Enzo]

1N4007 should work fine all way around.

[EFK]

Thanks Enzo. Anyone have any ideas why the 600 ohm resistor is added to ground at the output jacks? Is this some kind of protective measure? The one that is there is big, comparable to 5 watt size. It also looks a hundred years old. Wondering why they added it - protect against no speaker?

[Enzo]

That would be my guess.