High gain question ??

[MartyMart]

Hi all, I'm trying for a very high gain design, several stages, starting low gain and getting
higher as it progresses through.
2 x Jfet stages then a muamp and two further stages with TC at the end.
I'm having trouble with the second stage, which is creating "hash and crackle" when biased up
correctly, so was wondering if someone could tell me if I've made a design "goof" here.
It's a 1m gain pot, top lug from stage 1, bottom lug to ground.
Wiper goes through a 330k with 470k to ground and into stage 2's Gate ( 2N5457)
This has a 2k2 from source with 680n cap also and is set to approx 4.7v using a 50k
trimpot from 9v.
There's no apparent "build mistake" here, but unless the bias voltage is taken up to 7v I
have this "crackly" noise .... but of course the gain is reduced and a bit of a hard "mis-biased"
tone then creeps in .....
I've used 2k2 source resistors for the first two stages and 680n source bypass caps too, hoping
to NOT push too hard at these early stages.
What's causing the "hash & crackle" ... even on numerous fets ??

Thanks,
Marty.

[zachomega]

Just a thought, but try replacing the drain pot with a fixed resistor.  Maybe something like 12k or 15k? 

Also, I assume you mean the noises persist even once everything is set up.  Or more specifically, the noises exist when you try and play through the two stages...Right? 

-Zach Omega

[MartyMart]

Thanks Zach, I did think of that, but it would set up "that particular" Jfet and at the
moment, I have sockets galore and am still testing !
I seem to have helped it with a 4n7 cap from trimpot top to ground, I think the Dr Boogey
uses this on one or two trimpots too, so "hash & crackle" is gone .... "hum" remains !

High gain = noise !!!

Trying a negative feedback path now, if it comes to anything then I'll post it up.

MM.

[zachomega]

I don't know, I still don't think you should be experiencing that much noise.  I'm also working on a similarish project where I have 3 2n5457 stages cascaded. 

I did it on my breadboard and the noise was extremely low.  The 2n5457 is a pretty low noise part overall. 

Are you using a power source or a battery? 

-Zach Omega

Thanks Zach, I did think of that, but it would set up "that particular" Jfet and at the
moment, I have sockets galore and am still testing !
I seem to have helped it with a 4n7 cap from trimpot top to ground, I think the Dr Boogey
uses this on one or two trimpots too, so "hash & crackle" is gone .... "hum" remains !

High gain = noise !!!

Trying a negative feedback path now, if it comes to anything then I'll post it up.

MM.

[MartyMart]

Battery, seems like a little "hiss" and some earth noise, it is VERY high gain, so
perhaps this needs a shielded input/gain pot and an enclosure  !!
It doesn't clean up well, so may be a "back to the drawing board" one ....

MM

[Izzy]

I would so love to see your schematics, as I am desigining one too. 

[John Lyons]

Marty

Does the crackle happen when the gain is turned up high as well as when it's mid to low?
It's possible you are hitting the second stage too hard.
I'm not sure what the first stage looks like... 

John

[MartyMart]

Marty

Does the crackle happen when the gain is turned up high as well as when it's mid to low?
It's possible you are hitting the second stage too hard.
I'm not sure what the first stage looks like... 

John

First stage is 68k in, 220p miller cap, 2k2 and 680n off source, 22n out 470k470pf to 1m pot
then 330k from wiper to stage 2 with 220p miller, 470k to ground at gate another 2k2/680n
and on to the MUamp via another 470k with 820k to ground ....

MM.

[John Lyons]

Hmmm values seem ok, When you say second stage do you mean litterally the second fet? If the crackle is at that point there shouldn't be too much gain there.
If the noise is a pretty gross sounding crackle then I'm not sure what It couled be. If it's just a bit too fizzy then a "snubber" cap across the tim pot can tame the fizz as boogie and soldano etc do. The use of the PF caps to ground from gate and the caps across the drain reisitor or Trim  will help the fizz but this is pretty subtle sometimes. Definetely not a crackle if that's what you have.

John'

[MartyMart]

John, I solved the "hash & crackle" see a few posts above 
My problems are :
1 - a bit overcompressed ( sounds like a dynacomp! )
2 -  It doean't clean up AT ALL !! ...? so perhaps there's just too much hitting the Muamp
3 -  Overall tone is "OK" but there's a "hard" mid sound to it
4 -  I've had better results overall with just two/three simple stages !!! this is a monster
containing 6 Jfets ... MPF102/2N5457 and four J201's   
That's what I deserve for "building on the fly" straight from mind to vero !

5 -  My "Mr Mostorto" still sounds better 

MM

[John Lyons]

Oh, ok. I guess the key words were "seemed to help the problem". Didn't know it was cured...

(Thinking out loud) So by adding a cap from to top of the Trim pot to ground you are adding decoupling filtering to the power supply at very high frequencies. In a tube amp the power supply is broken up into Reisistor capacitor sections which are sometimes decoupled to ground with small caps. But in pedals there is only one source of power, so adding caps to ground from the +9v rail is docoupling the whole supply right? Not just the stage where the cap is connected.

About the other issues.
Have you tried probing the sections to see how the gain is distributed? May be a way to see about a cleanup method bypassing a stage of two. Compression would show up with the prbe too possibly.

I wonder if some eq dip here and there would flatten out the strong mids? Or is it just that the feel is too stiff in the midrange (no joKes please!)

John

[MartyMart]

Thanks John, yes the "seems" was badly chosen !! sorry.
Well stage one/gain into stage two doesnt "look" too stange or overly high gain
so all I can think is that this hits the MUamp section waaay too hard, so can't clean
up at all, there's plenty of my first two stages around that work OK !
From stage two's output is a 4n7 cap/470k with 820k to ground at "Mu gate input"
so either here is a problem  ...or... output from MU needs dividing down some more.
At present that's a 6n8 from MU out to 220k/270k divider to stage 4's gate.
I'll try a 1M/2M2 and see if that reduces the gain/level some.
An interesting addition, was a feedback path from Q1 to Q5 via a 4n7c & 3M3r which
puts some "spank" ( can't think of a better word ) back in, a bit like "speaker in amp"
being driven hard ??
It's worth persuing and you're right, I should start probing to see what's what

Thanks for the suggestions, much appreciated
MM

[John Lyons]

Marty

You want to go smaller on the divider leg to ground to drop some gain.
Resistors to ground increase gain as they are raised, degrease gain as they are lowered.
Resistors in line lower gain if increased, and raise gain if decreased ( You know this I know...)

Hmm I would have thought that the feedback would have lowered the gain at the frequency of the cap..
Or is it that that it raises it like NFB or the Presence control in amps?

John

[MartyMart]

Of course John ,,,, Thanks .... I've made three changes and have increased gain 
It's the other way around, higher in line and lower to ground !!  DUH !
I mixed that one up nicely  eh ??

onwards ......

MM

[John Lyons]

Ahhh..no problem...it's all very confusing with these bits...

The clip is downloading now (remember dial up?)  looking forward to hearing it in the new thread.

John