gain stage question

[iknowjohnny]

I'm don't have much theory/understanding of the theoretical aspects of a 12Ax7 gain stage. But i HAVE noticed that there seem to be 2 opposite schools of thought on how to shape tone in cascaded preamps. In mine i usually go the marshall route with 100k plate, 820R to 1.5k cathode bypassed with .68uf. Then a coupling cap thats usually .022. The other which is more along the lines of what fender does is to used big bypass caps then tiny coupling caps like 500pf to lose the excess lows frequencies from the large bypass cap. So i decided to try the more fender style values and replaced V1A's cathode with 1.5k/22uf and then the coupler going to V1B is now 500pf. I gotta say i think i like this, but i have one question since i don't have the theory to answer it myself. Whenever i see cascaded preamps using these values or similar i notice they tend to use a 220k plate instead of 100k. Should i change mine to 220k, and if so what is the reasoning behind it is any?

[tubeswell]

... So i decided to try the more fender style values and replaced V1A's cathode with 1.5k/22uf and then the coupler going to V1B is now 500pf. ...

FWIW, 'standard' Fender coupling caps in V1 are more like 0.022uF (22nF)

[Synchu]

Go there:
https://www.ampbooks.com/mobile/amplifi ... alculator/

put your values and see the plot and the calculated voltage gain.

220k/3.3k (if we stick to the golden ratio) provides more overall gain and impacts the stage response and next stage response as well (as you're feeding more signal into it).
Once you start overdriving stages, it is more efficient to experiment (or borrow proven designs) rather than doing calculations and plotting.
Admittedly, I did a number of LTSpice analysis for some of the more experimental builds of mine, which were more of general interest on how it works and attempting to foresee impact.
While it was great educational experience and somewhat useful when taking build decisions - given the real life components tolerances and the actual "hearing it" rather then "stearing at it" experience - a 30 second change of capacitor or resistor here and there is far more productive (while keeping an eye on the schematic)

Niki

[iknowjohnny]

that graph doesn't really tell me much as far as 100 vs 220k other than a bit more gain. Is that the only reason I see 220 so much more often in designs with large cathode/small coupling? Is more gain the only reason? Because it makes little sense looking at the graph. If all it does is add gain then why is it i see that used a lot more with those cathode/coupling values while marshall values seem to more often use 100k plates? Just seems like there must be other tonal implications that i dont see in those graphs, and if it were just about gain they could get that elsewhere. Maybe it adds or subtracts something other than gain like dynamics or compression, maybe complexity? Thats what im trying to find out....why do they lean towards a 220k plate in that design so often if it's not for gain?

[TUBEDUDE]

Adjusting Ra and Rk will change gain and possibly the symmetry of the output signal when overdriven, as the bias point can be shifted toward saturation or cutoff. This may be intentional to create a desired distortion quality or not.

With changes in gain, there are other response changes as well as increased noise with higher resistances.
Signal also changes by adjusting Ck and the coupling cap. larger coupling cap gives higher gain to a broader band of frequencies. The amount of that gain increase being dependent on the Rk value.

Rk to Ra sets the bias point. Ck and coupling cap set bandwidth. All affect tone to some degree. if Rk is electrolytic, and the coupling cap is foil or metalized film, their influences affect tone also.

The small Ck, high Rk you see minimizes the caps influence on bandwidth.

[Synchu]

My bad - I meant looking at the graph with various values of bypass capacitors - this provides a good visual on how the frequency response changes.

...and what TUBEDUDE said.

Higher gain may cause (besides desirable overdrive) - blocking and/or crossover distortion and some of the component values are selected (coupling caps, grid stoppers among others) to counteract these as well.

Niki

[iknowjohnny]

I get how the coupling cap and cathode caps work, in fact thats why i changed them to this very different scenario. I always felt big cathode bypass caps made the amp totally unusable but that small coupling caps removed all that extra low end the big bypass cap allowed. It seemed to me like the same goal as the marshall design but accomplishes it in a opposite way. the marshall cuts all the lows in the gain stage then uses bigger couplers so as to allow a lost the full range thru. the fender style design allows gain thru the entire bandwidth them cuts out the extra lows after the stage going into the second stage. And whenever i see 2 very different ways to accomplish the same thing i always have to try both because even tho they will accomplish the same goal, there are always going to be differences. And when i am looking to improve tone i certain ways i like toi try the "other" method and see if along with it comes a better tone in the details i'm looking for. This did just that.

i guess today i'll try a bigger plate value but it sounds great right now and unless it makes an obvious change for the good i'll go back to 100k. But when i try 220k vs 100k i usually find the difference is i end up preferring the 100k. But with this very different gain stage i thought there might be a reason designs with a gain stage like this more often seem to use 220k and wanted to ask because i wanted to know what to listen for.

Adjusting Ra and Rk will change gain and possibly the symmetry of the output signal when overdriven, as the bias point can be shifted toward saturation or cutoff. This may be intentional to create a desired distortion quality or not.

With changes in gain, there are other response changes as well as increased noise with higher resistances.
Signal also changes by adjusting Ck and the coupling cap. larger coupling cap gives higher gain to a broader band of frequencies. The amount of that gain increase being dependent on the Rk value.

Rk to Ra sets the bias point. Ck and coupling cap set bandwidth. All affect tone to some degree. if Rk is electrolytic, and the coupling cap is foil or metalized film, their influences affect tone also.

The small Ck, high Rk you see minimizes the caps influence on bandwidth.

[iknowjohnny]

By the way, why does this site get reported as a attack site when as far as i can see it;s not, and how can i get rid of that warning without turning it off for all sites in my browser?

[sluckey]

The other which is more along the lines of what fender does is to used big bypass caps then tiny coupling caps like 500pf to lose the excess lows frequencies from the large bypass cap.

Which Fender amp uses a tiny coupling cap for anything other than the input to the reverb circuit or the LTP PI?

[JazzGuitarGimp]

By the way, why does this site get reported as a attack site when as far as i can see it;s not, and how can i get rid of that warning without turning it off for all sites in my browser?

On my Mac, I use Safari for all my web browsing except amp garage. For that, I installed the firefox browser, disabled its security watchdog, and just to make sure I am not tempted to go other places in firefix, I made amp garage the defaul startup webpage so when I open ff, I am already at the desired destination.

[iknowjohnny]

The other which is more along the lines of what fender does is to used big bypass caps then tiny coupling caps like 500pf to lose the excess lows frequencies from the large bypass cap.

Which Fender amp uses a tiny coupling cap for anything other than the input to the reverb circuit or the LTP PI?

I know i've seen it but i'm not gonna go thru schematics to find one. But the bigger point was that fender uses big bypass caps a lot and there are a lot of more modern amps with cascaded gain stages that do the same then cut the flab with tiny couplers.

[iknowjohnny]

Another question....afaik something like 22uf allows the entire frequency band thru, so what is the point of using much larger caps as bypass? Do larger caps create more gain or what? I have seen as big as 320uf on a particular marshall, i think a 70's model.

[teemuk]

The following are naturally generalisations of the extremes but I hope it relays the idea:

High bypass capacitance and low coupling capacitance causes higher gain (most importantly) -throughout- the effective bandwidth of the gain stage and then distinct high-pass filtering of the result signal due to low coupling capacitance.

Harmonic distortion appearing throughout the effective bandwidth in equal amplitude means intermodulation distortion is also higher. The impact of that is often distinct especially towards lower frequencies. Distorted tone consists of more intermodulation products. Sometimes it's not so good considering "note separation" or overal comprehesibility of the signal.

Hi-pass filtering in coupling will then reduce amplitude of lower frequencies but it -will not- effect amplitude of intermodulation products of higher frequency that were generated previously.

Low bypass capacitance and high coupling capacitance on the other hand causes high gain at higher frequencies and lower gain at lower frequencies, a high-pass filter. Coupling does not introduce a distinct filtering effect. ...BUT having the high-pass filter implemented to the gain stage creates less intermodulation distortion at lower frequencies and probably less intermodulation products overall as well.

So effectively one may have two circuits that feature same frequency response (in total) and perhaps even similar pattern of harmonic distortion. But in practice they will sound rather different due to -intermodulation distortion- that is different in each circuit type.

[img:599:258]http://hyperphysics.phy-astr.gsu.edu/hb ... d/amp3.gif[/img]

[iknowjohnny]

Some of that sounds like what i'm hearing, tho some not so much. I take it intermodulation distortion is a bad thing? Anyways, with high bypass/low coupling capacitance it has a much tighter low end. The opposite gives some fuzz/flab in the lows, not a unbearable amount by any means but it's undesirable. I have tried both and tried to optimize other things to compensate for any undesirable changes and at this point i'm not sure which i prefer, but i'm leaning towards the standard marshall .68uf with .022uf couplers because tit has better complexity. The other way seems like a lot of harmonics are gone and it's much more fundamental. I just need to find a workable way to tighten the lows but so far nothing affects that and i have tried everything from PSU changes in voltage and filter values to peaking filters in the gain stage and all sorts of combinations of things.

[TUBEDUDE]

If distortion is the goal, harmonic is good, intermodulation is bad.