Non-CF Clipper?

[sepulchre]

I've been threatening to try a CF clipper (ala tarded or Jose) by have not gotten there yet. But now I've built this little amp base loosely on a Gibson GA 8-T. Actually, it's based on a Univox U45b which in turn based on the Gibson but that's getting into hair splitting territory.

Anyway, I'd like for it to have a gain boost switch. I tried couple of things in vain and so now I'm wondering if a CF clipping circuit could be used in an amp without the CF. My thought is to just put it right before the PI. Any thoughts?

Here's the schematic:

[sepulchre]

Okay, I just read Miles saying "Clippers sound like crap in an amp with few stages." Well this amp has few stages so maybe I should try something else.

Anybody got any ideas on how pump some gain into this thing? I've already got a switch on the faceplate so I need to make it do something. If worse comes to worse I suppose it could be a bright switch but I'd rather have some gain.

[matt h]

There's not really a lot you can do with just one preamp gain stage. (the second stage after the tone/volume is really your PI's driver. incidentally it's also the most influental distorting stage in the amp).

So assuming your quest for "gain" is really a quest for distortion, you could spend time playing with the value of the Rk on the driver stage. Reducing the value until your anode sits at about 1/2 B+ at idle will get you more swing, so you can hit your power tubes harder.


Since this appears to be an awkward 6bm8-a-like amp with medium/(low)high gain triodes for your PI and outputs, the only other thing you can really do is dink with V1 (preamp and oscillator).

Three options:
1) Replace the trem oscillator with solid state (currently being discussed right here in the tech forum) and have an extra series gain stage in front of the amp.
or
2) replace V1 with another pentode/triode type, but rather than a power-pentode, use something like a 6u8. The triode *should* still have enough voltage gain for a trem oscillator, but the pentode should give you a substantial boost to drive the driver harder (the only real distorting stage in the amp). You could use the switch you have on the front panel as a pentode/triode switch.
3) run an IRF820 or similar as a bootstrapped cathode follower after the input stage. You'll not only get more voltage gain from the input triode, but you'll reduce insertion loss for the volume and tone controls. The switch can switch in/out the bootstrap.

[sepulchre]

Yes, by gain I mean distortion, as in "high gain" amp, though I'm obviously not shooting for "high" gain here.

It's definitely a 6BM8 amp, with all the awkwardness that brings.

Hmm, a solid state trem; that sounds interesting. I could definitely use the extra stage. But I would have to completely rebuild the board. Maybe I'll keep that in mind for the next stab at this. As it is I have postponed building this for months in lieu of other "projects" (read chores). So I might just do the bright switch so I can be done with it.

I do appreciate the reply; good ideas for such a limited playground. Thanks!

Btw, I do have a TS808, so . . .

[matt h]

You wouldn't really have to rebuild the whole board. I mean, think about it, whether it's going to a tube socket or to the FET, it still only connects at three points! As for what you do with the newly freed triode? Well, I imagine you could hang a small terminal strip off one of the tube socket's bolts. Again, you're just dealing with, at most, an Rk/Ck pair, a plate resistor and its coupling cap (which ends up flying to your switch anyway).

Yup, high gain amps, with a lot of sequential stages proviing cascaded voltage gain leading to a bunch of compression. It's just important to remember that while gain may cause distortion, distortion ain't gain. Case in point, you put in diode clippers, and you *lose* a metric shit-ton of gain. Why does that matter? You wouldn't be driving anything downstream of them nearly as hard, which could be a real problem for certain topologies.

[martin manning]

What have you tried so far? A couple of simple things that come to mind are increasing the Ra and Rk on the input stage (roughly in proportion such as 220k/3k3, 330k/4k7) to increase its gain and distortion, and increasing the Rk on the second stage to bias it more towards cut-off to get some asymmetric clipping.

[sepulchre]

I tried increasing Ra on the first stage by a factor of 2. I rigged it so that there were two 200K resistors in parallel, one on a switch. That way I could switch between 100K and 200K. There was a difference but not as much as I expected. I didn't alter either of the cathode resistors. After looking at the load lines I see how that could add to the effect.

However, this may all be a bit on the moot side. The tremolo does not work properly. When activated it works for about a second then stops. I'm not really wild about tremolo anyway and am considering doing away with it altogether. That way I could use that triode as a gain stage. Then the "Speed" control could become "Drive" or something similar. I'd have that second stage going through a voltage divider so that engaging it would not create a huge difference in volume. That would have to be tweaked of course. Perhaps the switch could adjust that divider rather than engaging or bypassing a second stage.

Without the tremolo there would be plenty of space on the board for whatever I wanted to do, like having a relay on the switch so that a foot switch could be used. I'll get a new faceplate made (the local trophy shop does a great job for cheap).

But first I will raise Rk of the PI driver stage and pump the first stage too and see what I get. Thanks guys!

Couple of pics: (the cab is not quite as orange as it looks here)

[Structo]

Looking real good!

How did you do the logo on the amp?

[sepulchre]

Thanks!
It's burned on. And I didn't even use my soldering iron!

I don't know why the cab came out looking so orange. The stain is "Gunstock" and that's what it looks like. My crappy camera I guess. I should try my wife's, it's Much better (of course).

The back:

[Reeltarded]

Yes! It does look great!

[M Fowler]

I like that orange stain? on the cab.

[sepulchre]

Thanks!

The pictures make it look orange, I don't know why. The stain is "Gunstock" and that's what it looks like.

[Cameron]

Okay, I just read Miles saying "Clippers sound like crap in an amp with few stages." Well this amp has few stages so maybe I should try something else.

Anybody got any ideas on how pump some gain into this thing? I've already got a switch on the faceplate so I need to make it do something. If worse comes to worse I suppose it could be a bright switch but I'd rather have some gain.

You could try a PPIMV and use the the clippers across the + and the - of the PI. With some experimentation.. this can get some cool sounds... sometimes...... if you can't crank up the amp. Put in..a cut like control ....but instead of the cap..try different clippers. Making a small low wattage attenuator... I think.. could be another option...that may sound better.

[tubeswell]

Bias-vary trem needs to be stronger when it is feeding a cathode biased stage, because the 'auto-bias' characteristic of cathode biasing 'fights' against bias-vary trem wiggle. Should you wish to try improving the tremolo, some options to consider are;

1) With the strength of the trem in your circuit, you are limited by the lowish (~250V) supply voltage. You can make the trem slightly stronger by decreasing the AC load on the LFO gain stage. This can be done by increasing R21 to (say) 1M, and/or making R23 330k or 470k and making R20 4k7.

2) You could copy the pawnshop excelsior trem which uses an impedance buffer (IRF820 source follower) and separate (470k) connecting resistors and separate (470k) grid leak resistors for the output tubes.

3) You could convert the amp to fixed bias (with a brown-princeton-type fixed bias supply) and copy the 6G2 bias vary trem.

[martin manning]

BTW, I wouldn't worry too much about keeping the reservoir down at 10uF. The 6X4 data sheet lists that value as typical, not maximum, and in any case as long as the series resistance at or above minimum (see http://www.valvewizard.co.uk/fullwave.html) you should be ok.