preamp gain question

[daz]

I haven't been posting lately because i've been quite happy with my amp. But i lately have decided to add some gain to it and if you recall i had removed the voltage dividers between stages long ago and instead used 100k resistors from the 2nd and 3rd stage grids to ground, as you see in the schematic below. I can raise the gain considerably just by upping the value of either one. But my question is which would you change, the second stage or third? Or would you do both ? I can of course swap back and fourth and liten to what seems best, but is there a theoretical best way from looking at the circuit ?

[tubeswell]

Hi daz

I know you probably want a quick answer, but what you probably may want to look at is Merlin B's book. Chapter's 1 and 2 respectively cover the characteristics of a triode gain stage, and then coupling between gain stages. There's a lot of info in there, but in order to work out what will happen when you change the grid leak resistors, you probably need to understand:

1) the gain of each stage and how to work out the DC load line, biasing for desired results, what the input and output impedances are of each stage, the frequency response of each stage (taking into account the effect of miller capacitance, cathode bypass), as well as;

2) how the grid leak resistor interacts with the plate resistor of the previous stage and with the coupling cap between the stages, to produce the frequency response and gain (including working out the AC load line of each stage, and understanding how AC interacts with the output impedance of the previous stage and the input impedance of the next stage) - and there's a whole of lot formulas to come to terms with. So if you are anything like me you need to get the book and read it over and over and over until it begins to sink in...

M2CW - As you note, increasing the grid leak resistor will decrease the attenuation/increase the input AC voltage to that stage and vice versa. But this will interact with the coupling capacitance between the stages and the plate resistance of the previous stage to produce a particular frequency response. Increasing the grid leak resistor will also increase the noise floor.

[daz]

Well, i don't want to get that deep into it. As you know i already did my time on this amp, and i know i can raise the gain on either one to the degree i want with no problems. I just thought there might be some simple thoughts on which, such as raise the second one so as not drive as many remaining stages as much to reduce artifacts like oscillation or some such thoughts. But thats ok, i'll just put 250 or 500k trimmers in series with each resistor and tweak till i like it.

I've experimented raising both in the past, but i can't recall which seemed the best place to do it and i wanted to know the theoretical best way to avoid a lot of screwing around with it all over again. either place i raise it works fine as i recall, but one seemed slightly better after a time of listening.....just didn't want to spend a week evaluating it.

[BackwardsBoB]

I think that increasing the grid leak on V1B is the "purest" gain increase. You shouldn't have any grid conduction problems on this triode. The grid leak on V2A has to discharge the .022uF cap when grid conduction occurs at V2A (which is likely), so changing it's value will change the sound. Whether it gets better or worse is up to your ears.

A couple of trimmer pots and a tweaker would resolve the issue right quick.

Did I get it right?

[daz]

yes, thats the kind of answer i was looking for. So i will up that resistor with a trim pot till i like the amount, then if the amp sounds the same aside from the gain increase i'm done. Thank you sir.

[Steve Conner]

Well, I guess there is no "theoretical best way", but the two ways will have two different effects on the dynamics and feel, because they change the relative clipping points of V1 and V2.

[pdf64]

If you need more gain but want to retain the response characteristics of your present arrangement, the simplest and most effective way would be to put in an additional gain stage, in the form of an op-amp booster between the input and 1st triode.
See
Mission Amps Turbo Jax Amplified Input Jack
for a neat implementation. Peter.

[daz]

Well, I guess there is no "theoretical best way", but the two ways will have two different effects on the dynamics and feel, because they change the relative clipping points of V1 and V2.

Do you have any thoughts as to which to increase to retain the dynamics? By the way, those 100k leaks are actually 68k. That was an old schematic. Last nite i stuck 100k's on both and wasn't 100% happy. They have been higher than 100k in the past but i wanted less gain. But i changed the plate resistors a while back andthe gain went down some and lately i'm kinda missing it a bit. Tonite i think i will put the first one back to 68k and raise the second to about 470k and see how that does.

[daz]

Allow me to show my stupidity if you will.....i totally forgot i had a split load plate on V1B. I had done that back before i did my last round of mods a few months ago that had the amp sounding just as i wanted. Before that i had the split load because i has a 220k plate on V1A and some other differences that gave me too much gain, so therefore the split load.

Anyways, stupid me....forgot all about that. So i just left the grid leaks as they were and replaced the split load which was 56k/56k with a 100k and all is friggin perfect now. I was just going by memory as to the circuit when i posted this, but i just looked at the schematic and saw that and promptly slapped my forehead.