By popular demand, and because....

[Malcolm Irving]

Without doing the math... You could actually omit the voltage divider. Just use a 58k load against the tubes internal resistance. Because of the lower impedance you'd need to bump the coupling cap to .017uf to maintain the same knee frequency. But this should still reduce the time constant even further. Then...

You could reduce the grid stop value. Which would add a little gain so you could then reduce the load resistance even further, again necessitating a change of the coupling cap value. And so on...

I'd bet that a simple .02uf coupling cap, a 47k load and a 10k grid stop would give better results with a small increase in gain.

Not sure if you mean replacing the 100k plate load with 47k or keeping the 100k plate load and reducing the 'grid leak' of the following stage to 47k - probably the first option?
The plate load forms the DC load line and the parallel combination of the plate load and the grid leak form the AC load line.
A heavy AC load (i.e. using a low value grid leak resistor) is found in 'Trainwreck' designs. I have tried it on a couple of amps - it seems to increase harmonic distortion even when the amp is played 'clean' (if that makes any sense at all). Because the AC load line becomes more vertical you get less output voltage swing - possibly avoiding the need for an attenuation stage. Can be a good thing.

[pdf64]

Not sure if you mean replacing the 100k plate load with 47k or keeping the 100k plate load and reducing the 'grid leak' of the following stage to 47k - probably the first option?
The plate load forms the DC load line and the parallel combination of the plate load and the grid leak form the AC load line.
A heavy AC load (i.e. using a low value grid leak resistor) is found in 'Trainwreck' designs. I have tried it on a couple of amps - it seems to increase harmonic distortion even when the amp is played 'clean' (if that makes any sense at all). Because the AC load line becomes more vertical you get less output voltage swing - possibly avoiding the need for an attenuation stage. Can be a good thing.

Also most every Fender with an opto trem has a 50k load on the last (fx channel's) pre-amp stage, due to the trem intensity pot.

[daz]

By the way Chuck, i forgot to mention that after i posted about the gain being a tad too low i changed the 47k to 56k and that brought it back. Very little component difference but it was only very little difference in gain i needed to make up because i like my gain pot to be in a certain spot because if i have to turn it up i lose a but too much of the treble bleed frequencies from the .001 on the pot. If not for that i would have been able to make up for the lost gain by turning the gain up. I use the gain pot to tweak the tone more than the gain because of the treble cap on it. Within the range i use to tweak the tone the gain doesn't change enough to matter.

[Chuck H]

Malcolm,

The arrangement doesn't always mean less gain (daz has already corrected for that). There actually still is a voltage divider, but the tubes internal resistance acts as the series resistor. This allows for the lowest possible load resistor for the desired gain level at the lowest possible impedance for the standard type triode stage.

I'm talking about replacing the 220k grid leak with 47k, replacing the 470k grid stop with 10k, omitting the 100k series resistor and bumping the .0047u cap to .022u

Like this:

[daz]

I'm talking about replacing the 220k grid leak with 47k, replacing the 470k grid stop with 10k, omitting the 100k series resistor and bumping the .0047u cap to .022u

Like this:

Thats exactly what i did, tho as i said a 56k in place of the 47k to bring the gain to where it was. After a nite's sleep and fresh ears i can say for sure It's really sounding great Chuck, thanks much. I'm playing it and looking at the rest of the circuit now to see i if theres anything I should do to accommodate the way is sounds and responds now because as you know when you make such a radical change other things may now be better with different values or added parts. I have 2 or 3 things i want to try before it's cement, mainly things i'd like to make switchable. But any way it goes as long as it feels and sounds as good as it does now i'm good to go. The balance and feel are just about perfect now. One things for sure...being able to make such a positive change via theory without being able to hear the result or even hear what the amp in question sounded like before the change.....well, the you defnately know your stuff !

[pdf64]

Despite the Trainwreck thing, to my ears, a (next stage eg grid leak / vol pot) load of much less than 100k on a regular CC 12AX7 stage tends to make the tone thin and constricted.
I like to replace the trem pot with a 100k type, it just seems to sound better, even when the gain difference is compensated for.
I've not tinkered much with Trainwrecks, so not tried increasing that 3rd stage grid leak yet.
Maybe someone could try an AB test on one, keeping that corner frequency and gain compensated both ways?

[Chuck H]

Despite the Trainwreck thing, to my ears, a (next stage eg grid leak / vol pot) load of much less than 100k on a regular CC 12AX7 stage tends to make the tone thin and constricted.
I like to replace the trem pot with a 100k type, it just seems to sound better, even when the gain difference is compensated for.
I've not tinkered much with Trainwrecks, so not tried increasing that 3rd stage grid leak yet.
Maybe someone could try an AB test on one, keeping that corner frequency and gain compensated both ways?

One thing about the Trainwreck (and daz's and one of my designs) is that they have a fairly saggy HV rail. So you get your squishy back. Regarding the TW and my design, they both beat the piss out of the power tube grids, so that's pretty squishy too. Without the lower impedance driving the triode stage that gets clipped the tone get's TOO mushy and compressed and lacks articulation. So, I suppose, the lower impedance effort becomes a useful part of a recipe rather than something to be generically classified as "constricted".

[Chuck H]

One things for sure...being able to make such a positive change via theory without being able to hear the result or even hear what the amp in question sounded like before the change.....well, the you defnately know your stuff !

Daz... Thank you! But I'm a total techno tard. Any theory I have is a combination of sheer exposure (time and pressure) and rote learning rather than actual understanding. I use A LOT of on line calculators when designing to avoid formulas for which I simply don't speak enough mathematical language. BUT... My focus has been predominantly on circuit design anyway and I've had some experience there. I often feel the need to qualify my lack of tech AND experience equally and avoid any notional considerations so I don't come off like some ignorant half baked amp guru type asshole.

[daz]

Daz... Thank you! But I'm a total techno tard. Any theory I have is a combination of sheer exposure (time and pressure) and rote learning rather than actual understanding. I use A LOT of on line calculators when designing to avoid formulas for which I simply don't speak enough mathematical language. BUT... My focus has been predominantly on circuit design anyway and I've had some experience there. I often feel the need to qualify my lack of tech AND experience equally and avoid any notional considerations so I don't come off like some ignorant half baked amp guru type asshole.

Just telling it like it is. That wasn't the first time you gave be advice thats worked out real well but it's probably the best time. Amp is just sweeter'n hell now. One thing i always notice whenever I have an amp sounding so right is that then and only then do they respond to other things properly. For example, when it's right you can cut the value of a coupling cap in half and it does what you would expect and to a very noticeable degree. When they aren't quite right you can hardly tell the difference other than the larger cap might be a bit wooly in the lows. Treble perkers also work very different in a "right" circuit and not at all the same or as useful in one thats not quite right. the great thing about that is you can then REALLY start to tweak it to you taste because now you can determine what it needs and what will accomplish it. heres the current schematic.

[Malcolm Irving]

Just for interest, the time constants for the 0.022uF coupling cap in the new design are:

when the grid is conducting:
( 40k + 56k || (10k + 2.5k) ) x 0.022 e-6 = 1.1mS

and when the grid of V2A is not conducting:
( 40k + 56k ) x 0.022 e-6 = 2.1mS

(The 2.5k value is a guestimate to allow for the resistance from grid to ground of V2A during grid conduction)

These are similar figures to the earlier design. Also, I expect we want some dynamic bias shift during overload to get the desired guitar sound and playing feel.

[Chuck H]

Just for interest, the time constants for the 0.022uF coupling cap in the new design are:

when the grid is conducting:
( 40k + 56k || (10k + 2.5k) ) x 0.022 e-6 = 1.1mS

and when the grid of V2A is not conducting:
( 40k + 56k ) x 0.022 e-6 = 2.1mS

(The 2.5k value is a guestimate to allow for the resistance from grid to ground of V2A during grid conduction)

These are similar figures to the earlier design. Also, I expect we want some dynamic bias shift during overload to get the desired guitar sound and playing feel.

Similar, yes. It looks that way more so because the numbers are small. 1.36 - 1.1 = .26 So that's about a 24% reduction in discharge time when it counts. As to dynamic bias shift... Different strokes and all. I don't find it objectionable even when it's pronounced in someone else's tone and some players really exploit it as an extra dynamic effect. I guess I just prefer the sound and feel of less of it for some amps.

[Chuck H]

If the peaking is in the extreme HF you can try the common 47pf snubber across the PI outputs. Some classic amps use 100pf here. Those are the only values I've used. The exception being a "Voxy" sort of design I did that used the classic designs "high cut" circuit (.0047uf + variable pot) which is a whole other thing because those amps have no NFB and can sound harsh without it (presence on 10 all the time).

If the typical PI snubber doesn't offer satisfaction you COULD always use component values in between where the circuit was and where the circuit is. That actually works. So, a little less like it is and a little more like it was, for better of worse. I'll avoid the math and SPICE up some values if this interests you

[daz]

If the peaking is in the extreme HF you can try the common 47pf snubber across the PI outputs. Some classic amps use 100pf here. Those are the only values I've used. The exception being a "Voxy" sort of design I did that used the classic designs "high cut" circuit (.0047uf + variable pot) which is a whole other thing because those amps have no NFB and can sound harsh without it (presence on 10 all the time).

If the typical PI snubber doesn't offer satisfaction you COULD always use component values in between where the circuit was and where the circuit is. That actually works. So, a little less like it is and a little more like it was, for better of worse. I'll avoid the math and SPICE up some values if this interests you

I actually came to remove that peaker in front of the gain pot after i posted that schematic. I was getting too much high end with it and even posted about removing highs but deleted the post after i realized that was the cause. I dunno if u saw that post b4 i deleted it or was assuming there was too much top by the schematic after seeing that peaker there. I still trick myself often by playing with the volume too low which fools you into thinking the highs are fine when in reality turning it up reveals too much. The peaker there gave the feel even more give in the attack and i liked it but it was with the volume too low and once i turned it up the top was overbearing. I decided it's not worth having it switchable so i just removed it altogether. Now it actually seems like the top may be too little after playing it in a mix last nite at home. Turned up some mp3's and jammed along and found myself cranking the top up a lot more than when i play it solo to make it sound right in the mix. I even switched in the 100k NFB resistor to brighten it up instead of using the 56k i normally use. So i may put that peaker back in maybe with a smaller resistor or find another way to add highs and make it switchable so that when i do play out i don't have to worry about the tone getting buried.

If you guys are insinuating theres little dynamics in this circuit i didn't feel it. Seems fine in that regard and i'm quite into a dynamic tone. But the more the better so if you know how i can make it even more so i'll try it. Thats one thing i never get enough of and always figure it;s somewhat exclusive to cranked output tone, at least to the degree i would like it to be.

[Chuck H]

If you guys are insinuating theres little dynamics in this circuit i didn't feel it. Seems fine in that regard and i'm quite into a dynamic tone. But the more the better so if you know how i can make it even more so i'll try it. Thats one thing i never get enough of and always figure it;s somewhat exclusive to cranked output tone, at least to the degree i would like it to be.

Dynamic is a tricky word. The dynamic bias shift mentioned is actually a reduction in dynamics, but it IS a dynamic effect relative to attack. And, of course, dynamic is often used to describe an amps ability to respond to pick attack accurately. See? For some damn reason the word dynamic seems to mean either an effect that is relative to dynamics OR a lack of effected dynamics!?! So...

The circuit you're using now has more actual dynamics. That is, dynamics are less effected. Some players feel that an effected dynamic is more expressive and, so, might say that amp has better dynamics. When in fact it has less dynamics.

My head hurts.

[daz]

Dynamic is a tricky word. The dynamic bias shift mentioned is actually a reduction in dynamics, but it IS a dynamic effect relative to attack. And, of course, dynamic is often used to describe an amps ability to respond to pick attack accurately. See? For some damn reason the word dynamic seems to mean either an effect that is relative to dynamics OR a lack of effected dynamics!?! So...

The circuit you're using now has more actual dynamics. That is, dynamics are less effected. Some players feel that an effected dynamic is more expressive and, so, might say that amp has better dynamics. When in fact it has less dynamics.

My head hurts.

Ok, so it now has BETTER dynamics than before? If so i don't think i notice that but I would have noticed if there was less, at least to a degree that COULD be noticed. It just seems the same as before in that regard except that with the added clarity/articulation i suppose you could say that makes it more dynamic. I'll have to think about it as i play it and see if it's something i can notice compared to how it was, but one thing for sure it's no major difference in that regard. What i don't get is why anyone would want LESS playing dynamics? To me i have always felt it;s the biggest factor in great vs mediocre amps. Long ago i noticed it was amps with less touch sensitivity that were the ones i found hard to play and i don't think thats just me. I think everyone notices that even tho many probably don't know the reason behind why they have a much harder time playing amp A than amp B. Of course none of this matters to metal players who have gain beyond sanity all the time, but they generally like amps built for a ton of gain w/o little regard to dynamics.