plexi-6sl7

[andrew]

I would like to try a 6sl7 in the second preamp tube position on a plexi build. Should I change the 100k resistor value from pin 6 to pin 1? Thanks.

[David Root]

6SL7 has 2/3 the plate resistance of a 12AX7, so you should change both 100K resistors to lower values. I would try 68K. I've done 6SL7 in hi-gain circuits before, this will work better. You will not get the same gain out of V2B, about 38 vs 53 for the 12AX7, but it's still plenty.

Also maybe change the cathode resistor to 1K.

You should get nearer 1.5mA with the 6SL7 compared with about 1mA with the 12AX7.

AND most importantly use 6V DC on the heaters, high gain octal pre-s like the 6SL7 absolutely need DC or they will hum like crazy in a hi-gain application like this Marshall circuit. I know that may be a pain but you would only need to do it on V1 and V2.

[andrew]

Thanks David, if I want to use a 6sl7 in V1 what changes would you suggest for it. I am using a split cathode with 2.7/.68 and 820r/22uf. I will look for a 6v wall wart around here. Thanks again.

[David Root]

Again, 68K or 56K plate, probably 1K/1K2 cathode, 2K7 is too high for a 6SL7 I think, but you never know, that can be a matter of taste. I would think you would get some odd harmonics with 2K7. I should add that I did my analyses using 280V B+. If you move that up to 300V it shouldn't make too much difference overall.

As a matter of principle I tend to favor cutting the bass in V1A, and if necessary putting some back in V1B, which you seem to be OK with. Caps are also a matter of taste, I might go a bit higher on V1A and a bit lower on V1B, but that depends on taste, single coil vs HB, etc.

[andrew]

David, I can't buy an octal for the PI right now but until I get one I have a 12bz7 and a 6cg7 to play with. The 6cg7 supposed to be the same as a 6sn7 but I doubt it sounds the same.Do you have any experience with these types of tubes? Thank you.

[David Root]

6CG7 would make a reasonable PI I think. HiFi guys like it as a PI and a line driver, so I'm sure it's pretty clean. It will run higher current than a 12AX7, more like a 12AT7. 12BZ7 I think is an obsolete predecessor of the 12AX7, don't know anything else about it.

FWIW Andy Fuchs uses a 6FQ7, which is essentially a 6CG7 without the internal shield, as a driver in his Clean Machine, which is based on a Dumble SSS (sort of). Fuchs Amps are at the top of the heap in the large boutique amp class IMHO.

[andrew]

Hi David, should I lower the plate resistors for a 6cg7 also? Thanks.

[David Root]

Yes, depending on how tightly balanced you want the two sides of the PI to be. There are some wrinkles to consider though.

For example, with 400V B+, the 6CG7 on a long tailed pair with 56K plate in the input side and 59K in the other side, and a 33K total tail resistance (this includes the bias resistor, the tail resistor and the feedback resistor to ground) will give you pretty much perfect balance if the two triodes are reasonably well balanced. You should draw about 2mA.

The main drawback is the gain is very low, just under 9, where a 12AX7 would give over 20 with an 82K/100K plate setup. However the 6CG7 PI will not distort the signal as it can pass a huge voltage swing +/-114V. It is the 9-pin 6SN7 which is a superb HiFi PI. So as long as you can generate plenty of signal voltage into the PI you should be fine.

When you do get an octal for the PI, the 6SL7 is also very good and plenty clean for a guitar amp, with a much higher gain, probably double what the 6CG7 would do.

A great tool for this kind of analysis is a piece of software called TubeCAD. It is $40 and I use mine all the time, I have had it for about seven years now and it's the best $40 I ever spent in amp design tools. It is what I used to come up with these numbers in 5 minutes, compared with an hour doing it with plate curves and a pencil and ruler.

[solderstain]

6CG7 would make a reasonable PI I think... more like a 12AT7.

I have a Silvertone 1484 in the shop again to sort out, and it uses a 6CG7 for the PI...

And it's missing...

And... I don't have one in my tube drawer.

Just for the sake of trouble-shooting the amp for signal continuity, is a 12AT7 reasonably close enough to use in a pinch?

[David Root]

Yeah, sure no problem I think. I looked at both in Tubecad w/B+=400V, 47K/51K plate resistors, balance about the same I think, better than 10%, get about 260V plate, input and output impedance about the same, at 3mA current.

"good enough for rock n' roll"!

Bias points are quite different, 12AT7 is much hotter so it won't pass as big a signal without breaking up. That may or may not be a problem, they both pass lots of signal cleanly.

[solderstain]

Yeah, sure no problem I think. I looked at both in Tubecad w/B+=400V, 47K/51K plate resistors, balance about the same I think, better than 10%, get about 260V plate, input and output impedance about the same, at 3mA current.

"good enough for rock n' roll"!

Bias points are quite different, 12AT7 is much hotter so it won't pass as big a signal without breaking up. That may or may not be a problem, they both pass lots of signal cleanly.

Good. Thanks, David. I just want to see if the thing is actually passing a signal before I unscrew the two chassis and start probing around inside.