Potentially ignorant split load question :)

[daz]

Ok, thanks. Haven't gotten around to it but i'll post back w/results. Thanks so much !

[daz]

Well, good news and bad news. The good...no pop even w/o 1M R. I mean there IS a pop, but it's super low, just barely audible. The bad it it works great but for the lowest gain setting which works fine till you rank the gain on the amp. Then i hear what sounds like blocking where the sound just sorta falls away. I guess it's because theres only 33k between the B+ and the cap? Maybe it can only be so small and might need maybe 47k. But then theres not enough difference to make 3 settings necassary. And yes, rather than use the lowest setting with gain up full i can use the middle setting and roll the gain back. But it's a matter of the circuit having a certain feel and sound with the gain knob at a given setting. 2 gain levels would probably be enough if not for that. Anyways, it otherwise works great as long as i don't turn the gain up full on the low gain setting.

EDIT: on final thought, i'm just gonna use two settings, a 56k and a 47k. As i said earlier, probably over kill and it is.

[daz]

Spent hours playing it after going back to a 2 position split load and then finally back to the basic 100k and cap, no split. Theres something i can't put my finger on with the split load I started to hear that bothered me in the top end. Just wasn't right even when using it with the switch in the normal position with cap at the plate. Makes no sense but when i put it back to the way it was b4 the split mod it just sounds more right. So back to the V2A switchable cathode R as gain reduction.

[mhuss]

With a split load you have an additional R in the signal path, which combined with the input capacitance of the following stage will reduce the ultra-HF a bit.

[Helmholtz]

With a split load you have an additional R in the signal path, which combined with the input capacitance of the following stage will reduce the ultra-HF a bit.

Not true.

With a typical fully bypassed 12AX7 gain stage using a 100k plate resistor output impedance is 39k.
With 50k/50k split load output impedance calculates as 35k.

[daz]

Whatever it is it did sound a bit different so i have gone back to the V2A cathode for gain reduction method. But instead of switch i wanted to be able to lower it anywhere from the stock 820R to as low as i would ever want the gain, so i put a 5k pot and paralleled it with a 4.7k to get around 2.5k and put it in series with the set 820R. i was thinking i'd get a loud scratchy sound when adjusting it but nope. Seems to work great so far.

[Helmholtz]

With a non-linear system like a tube amp any change of signal level has an inluence on harmonics content and thus on sound.
Also makes a difference where the gain is changed.

Particularly varying the V2a cathode resistor changes way more than gain.

[daz]

I know, but i'm out of ideas to have a switchable lower gain. The split load changed things too and this way seems like it changes things less. I'm open to suggestions if u have any. Already have a 68k load on V2A.

[uneumann]

Since no one has mentioned it (unless I missed it) you can decrease the impedance the plate drives to cut gain. It's just an alternate approach.
The current 0.022u cap w 68k has a 100 Hz rolloff. Leave that alone to keep that response. Add a new coupling cap at the plate in series with a new resistor (say 47k for example) to ground. The new cap should be relatively large (say .1u) to minimize impact to your current response curve and to allow you to vary the resistor values to ground. Your can try other resistor values to get the cut you want. The lower the R value, the lower the gain.

BTW - this will impact the nonlinear behavior of the stage slightly, but then almost anything you change will impact that. Maybe you will like it.

[Helmholtz]

Not quite clear to me what the purpose of signal attenuation is here.
Why no just lower the vol. pot setting?

If the attenuation is intended to increase clean headroom of the amp (as mentioned in the other thread) it won't work as also available clean signal gets attenuated.

[daz]

It's because when i play the amp at high volumes the output and preamp drive together is too much unless i turn the gain way down and that ruins the distortion tone's complexity. Even at lower volumes i like the gain knob set around 2:00-3:00 but would prefer the gain there to me more like it is at noon. I can turn it to noon but then it's different and again, loses too much complexity and becomes sorta dull. I realize thatsthe nature of lowering gain but theres more to it that that....it's just hard to put into words.

[eschertron]

Don't some Marshall models have a 470K resistor in front of the 500K vol pot to cut down the max signal reaching the next stage? There's also a parallel cap for tone shaping options too, which might be nice...

Even at lower volumes i like the gain knob set around 2:00-3:00 but would prefer the gain there to be more like it is at noon.

Sounds like what Daz is asking for?

[Helmholtz]

Don't some Marshall models have a 470K resistor in front of the 500K vol pot to cut down the max signal reaching the next stage?

Not sure which circuits you mean, but the 2 "mixing" resistor in a typical Marshall (or Fender BM) are after the vol. pots and are needed to make the pots somewhat independent from each other. Without the resistors no signal would pass if the pot of the other channel is at zero.

Adding a resistor from grid to ground (like a grid leak) would give additional attenuation.

But to be able to keep a certain vol. setting the attenuation should be before the vol. pot. Easy using a voltage divider.

[daz]

Not sure which circuits you mean, but the 2 "mixing" resistor in a typical Marshall (or Fender BM) are after the vol. pots and are needed to make the pots somewhat independent from each other. Without the resistors no signal would pass if the pot of the other channel is at zero.

Adding a resistor from grid to ground (like a grid leak) would give additional attenuation.

But to be able to keep a certain vol. setting the attenuation should be before the vol. pot. Easy using a voltage divider.

I've done that. It's strapping it across the pot and changing the value. I tried a couple, one that results in i think a 300+k pot something k, and another that gave 180k. Like everything i've tried so far it works but tone and/or feel gets lost to one degree or another.

[Helmholtz]

I've done that....

Which of the methods mentioned do you mean? A drawing would be best.

Another interesting option would be using a lower gain input tube like a 12AY7.

BTW, does your amp have a MV? If so, is it fully up?