High gainer, clipping, post PI MV, tone shaping, etc!

[Tooboob]

Right. The gain knob on modern lead channel preamps reduces volume and decreases distortion. A Post-PI MV reduces volume and increase PI distortion to compensate for loss of power tube distortion. Both are useful. Input signal level from the preamp to the PI alters the distortion and changes the required Post-PI MV setting, just like changing a preamp gain knob can force some adjustment of channel or pre-PI volume controls to get the desired tone. But lots of folks think a gain knob is a good thing. It's worth the bother.

Multiple volume controls in the signal chain of a distorting circuit allow flexibility at the expense of interaction. If you have a preamp gain control, a channel volume, and a pre-PI MV, and you add a post-PI MV, you gain control over channel relative volume, preamp distortion, output level and power stage distortion, but you have to master dealing with their interaction, especially since the goal of the post-PI MV is to allow various output levels while maintaining system distortion. Lots of great amps have a single volume knob for all these functions, and it forces the amp to be played at a fixed level for a given level of distortion. I suppose it's OK if you're seeking to play very loud, or the amp is lower power and it's just a source for a mike to a PA, but few guitarists spend all their time in these situations.

Looked at this way, to maintain full control, you either wind up adjusting lots of interactive knobs, or you need to add a computer to take separate preamp distortion, channel volume, output distortion and output level controls and make the necessary interactive adjustments (and wouldn't that suck!). Add an effects loop in the middle, and things get even harder.

Really, defining the number and function of the knobs should be a first step in amp design, and I get stuck there. Just volume, bass and treble has been proven to be useable, and these days, the knobs and switches are frequently the limiting factor for front-panel size, requiring months of experimentation and discoveries. There's definitely a simple-versus-flexible trade off, and few guitarists can or want to pull a Santana, Gibbons, or King, and only make one noise.

[daz]

If the cross line treble cut won't work, can anyone help me find something that will work?

I'm the least knowledgeable person here, and i haven't read this thread aside from a few posts so i may be answering out of context. But it seems you have NFB and want to cut treble in the power section, yes? Try putting a cap (start with a .01uf and work smaller if thats too much) across the NFB resistor. Instead passing highs as in any other area, because it's negative feedback it attenuates highs. If the resistor isn't big enough it won't make a lot of difference but you could add a larger resistor or use a larger cap. Use your imagination....larger resistor, same resistor PLUS a larger one with the cap across it, different cap values, etc etc. But it can cut highs in a very smooth way. You just have to play around with it. I tried a .01uf across a 100k NFB resistor and it cut a lot, but i have a variable NFB pot too so i was able to use less NFB and tune it just right. You could use a pot in series with the cap too. Many ways to implement it, but you get the idea. Hope i read enough of the thread to know what i'm talking about here, if not disregard...

[Gingertube]

Daz's (least knowledgeable or not) method is the go for cutting highs in the power amp, that is provide more high frequency feedback.

No tone control system which is inside a feedback loop (which a post PI cross line cut system would be ) is worth a pinch of the proberbial. As Alan pointed out you may well run into stability problems but the bigger problem is that the feedback loop works hard to eliminate what ever tonal changes that your tone control makes. Its an exersize in futility.

Cheers,
Ian

[Chuck H]

To expand on daz's excellent suggestion I think you could implement a very effective control using a dual ganged pot. One gang used as a variable resistance for a cap across the feedback resistor and the other used as a variable resistance for a cross line cap. Figuring that subtracting the highs from the loop should allow a cross line cancellation simultaneously and this arrangement should give greater range to the control.

[Dave H]

Daz's (least knowledgeable or not) method is the go for cutting highs in the power amp, that is provide more high frequency feedback.

No tone control system which is inside a feedback loop (which a post PI cross line cut system would be ) is worth a pinch of the proberbial. As Alan pointed out you may well run into stability problems but the bigger problem is that the feedback loop works hard to eliminate what ever tonal changes that your tone control makes. Its an exersize in futility.

A capacitor across the feedback resistor (more high frequency feedback) sounds like a recipe for oscillation to me whereas most LTP PIs already have a cross line capacitor for stability.

It’s not ideal having a tone control/MV inside the feedback loop but it’s not entirely futile. A tube amp isn’t an opamp. It doesn’t have much negative feedback so it is only able to reduce the effect of the tone control/MV not eliminate in completely. When I simulated Daz’s amp it had a loop gain of about 6dB so I’m guessing that would only halve the effect of any controls inside the loop.

[Tooboob]

No tone control system which is inside a feedback loop (which a post PI cross line cut system would be ) is worth a pinch of the proberbial.
Ian

If the tone control is after the speaker in the feedback line (still inside the loop), it can work, though it will change loop phase at the affected frequencies, and that needs to be addressed. Then the speaker doesn't see what the PI's feedback input sees, and the PI makes the tone change instead of cancelling it. I think it's a nice place for a presence control. You get more distortion, but I subjectively think it's the good kind.

You know this, or you wouldn't be going along with daz. I'm just clarifying.

I do disagree with daz. I am least knowledgeable, but mostly only here. Scary folk here.

-Herb

[jimboyogi]

I sure that I am the least knowledgeable here. Even more reason for me to be grateful for all of the assistance you have all been giving me.

I've been very busy lately, so I haven't had much time to experiment further, or to post any results.

But I have had enough time to try a couple of things.

First up, as I previously posted, I tried adding the 47K resistors between PI outputs and a 'cut control'. Dave H thanks for this tip, and I also did as you suggested and wired the 250K log pot the other way, so that the pot was a 'treble' control rather than a 'cut' one. Huge improvement in adjustment range this way.
But as many of you have been saying, there are effects of having the control inside the NFB loop.
As I am only using 6dB of NFB, the control was still effective even at full postPIMV setting, where the full NFB was in play. But as the MV is turned down, and the NFB decreases, the control filter becomes more effective, and there is less treble at the amp output. This is exactly the opposite effect than I was after, and because of this I have removed this control, at least for now.

Daz, Gingertube, Chuck H, thanks for your advice about 'resonance' type controls. I did try a few different values of cap across the feedback resistor, and I found all of these values to cause the amp to oscillate. I then tried a cap in series with the feedback resistor, and this definitely works in this amp. I am still fine tuning this cap value, but this is a handy trick, up to 6dB of bass boost when MV all the way up, decreasing as the MV is turned down.

I have also done a little experimenting with various treble boost and bass cut filters in the early stages of the preamp. The rationale of the treble boost early, bass boost late high gain amp was the reason i began this thread. But I have so far been finding that a little bit of treble boost early goes a long way, and it's easy to overdo the treble. Perhaps that's the skill of the high gain amp designer, to utilise a range of different eq filters to avoid the boosted treble getting too harsh, while still sounding natural