Local Feedback in a Mu Amp Stage Question

[deadastronaut]

ive always loved the FET (J201/2N5457 muamp setups for heavy dirts,  cleans up nicely too, and very quiet...

no annoying trims on drains etc, ..and they do have a certain amp like tone and response...to my ears anyway

something you can never get through opamp/clipping etc...imo.

if you want bite, and definition fet it up ,

unless you prefer fuzz...of course...

just my 2p.....as you were...

[amz-fx]

Hey Jack, thx, those circuits are regularly inhaled by me. Somehow when I'm breadboarding the blues away, I always return to this kind of circuit - this diode-clipping world is not really for me. I like the hint of raw power behind these design and love to taunt it with various fet combinations: weaker ones in the first stage to keep them from fuzzing out, big trolls like 201s in the second to give the output (and the amp after it) something to work with. An inexhaustable source of different flavours in those

One quick tip... use any fet that you have a plenty of for the top jfet. It is less important to the sound than the bottom one, so use what ya got!

Best regards, Jack

[KarenColumbo]

Hey Jack, thx, those circuits are regularly inhaled by me. Somehow when I'm breadboarding the blues away, I always return to this kind of circuit - this diode-clipping world is not really for me. I like the hint of raw power behind these design and love to taunt it with various fet combinations: weaker ones in the first stage to keep them from fuzzing out, big trolls like 201s in the second to give the output (and the amp after it) something to work with. An inexhaustable source of different flavours in those

One quick tip... use any fet that you have a plenty of for the top jfet. It is less important to the sound than the bottom one, so use what ya got!

Best regards, Jack

Cool! You mean they don't have to be the same type? That's even more cool Should have similar Idss though, don't they?

[amz-fx]

Cool! You mean they don't have to be the same type? That's even more cool Should have similar Idss though, don't they?

Don't have to be the same type or similar Idss for the circuit to work. You can even use a mosfet or bipolar on the top (with some tweaks).

regards, Jack

[PRR]

> they don't have to be the same type?

Matched pair reduces 'errors'. Many such errors you don't care, and some errors you may LIKE.

I would expect two very different devices would not find a happy operating point, but go ahead and try.

[amz-fx]

I would expect two very different devices would not find a happy operating point, but go ahead and try.

That is not the case. It makes little difference what device is used for the top jfet, and matching is not needed at all. The circuit will always self-bias, which is one of the attractions of it. In fact, the original NatSemi app note had different devices for the two transistors.



The performance of the circuit does change some with the specs of the top devices, but generally speaking, any jfet works for the top tranny (or either one for that matter).

Best regards, Jack

[Eb7+9]

Jack is right, the top device acts as an active load and provides level shifting ... its Vgs voltage determines amount of up-shift from Gate to Source ... all this provided it's Vgs is not chosen to be too large relative to supply ... that's just as far as biasing goes

the OP wants to know how to deliver measured amounts of clipping per pound of input signal ... for that to happen, to taste, it's the Vgs or bottom devices that need to be chosen appropriately ...

as far as cascaded Mu-stage designs go :

measure your devices' Vgs voltage and swap them in terms of hi/low first last and see/hear what happens ... you can ignore Idss in all this

[KarenColumbo]

Yeah, well, the circuit worked out fine. Thx to all - I just took my time and carefully "voiced" the whole thing from low gain fets to higher gain ones.

But I digressed quite a bit halfways and thought about what jack said. Got a hint about Doug Hammonds "CCS Drive" and tried out a BJT totem pole construct just for fun, to see if I get the loads of gain and the "fresh" and "dynamic" colouring using bipolars.

I remembered those "Vulcan" and other overdrives by Doug, deceptively simple circuit snippets with a sh*tload of benefit using readily available BJTs, that sound awesome.

So i started with a gain stage à la CCS Drive and toned it down 'til it almost did what I wanted. Except when it clips, it clips ungracefully harsh and sounds bad.

But there is a lot of promise there: crisp harmonics, loads of easy gain and even spectrum amplification without taking away the character of the guitar signal.



I guess I can tone down the amplification factor by increasing R2 to, dunno, 4k7. But There's always this ugly clipping behaviour waiting behind the corner.

Is this something worth following? I mean I'm aiming at the usual topology: Gain Stage 1 -> Gain Stage 2 -> FMV Stack -> Output Buffer

The first stage with the 5088s sounds beautiful as long as it doesn't clip

[KarenColumbo]

A little bit better:


Still: This thing should not clip - sounds awful Any chance to clamp this down?

[amz-fx]

Put a resistor in series with C1 (after the R1 pulldown). Start with 10k or 22k or 39k (it might take more resistance though).

Let us know how that works for you.

regards, Jack

[KarenColumbo]

Put a resistor in series with C1 (after the R1 pulldown). Start with 10k or 22k or 39k (it might take more resistance though).

Let us know how that works for you.

regards, Jack

Thx! Tried it, but doesn't work, gets worse, I'm afraid. Maybe I'm on the wrong track with this.
Maybe try some lower gain transistors here?

I know, this is fishing in the dark, as usual with me. There must be several things wrong. Please bear with me here if you can

- I put in a red LED instead of the germanium diode. Just because before I put in the classic 914/4148 and it was worse, so I thought, the higher the threshold for the current (?) "kicking back", the better. Got better a bit. But what does the diode actually do? In my inner unschooled eye it puts up a one-way wall against interaction between the local feedback R3 provides and eventual "counter-feedback" that goes the other way (input to collector of Q1). Maybe put a hefty Zener there? Like 9.1 Volts so it NEVER can go this way?
- IF it ever works as intended (or rather fantasized), there's still the problem that I hit it with  a Stratocaster - which has a pretty loud humbucker at the bridge. But there are way more powerful guitars that  just knuckle the thing over with pure Volts.
- Varying C3 has a huge impact on the overall operation. But I suspect it's only frequency-wise. i put in 100nF, sounds thin as american Budweiser, from 1 uF up it gives the whole spectrum. The 2u2 I have in the schemo do't differ from the 1 uF.

I'd hate to waste your precious time. Should I continue here? Or is it bound to fail?

Edit: Put in a 15.1 V Zener in the diode place - now the signal lost all treble content .... and still clips ugly

[KarenColumbo]

Since there ain't really  more envolved than a few minutes' time, I'll stick a second similar stage to it, the ubiquitous FMV stack and an output buffer. Tomorrow. Maybe the part won't speak for the whole. I will post the result

[amz-fx]

The 1N34a probably has some leakage so it is impacting the bias when in parallel with the 10M. When you sub in the LED, it had no leakage and the bias on Q1 changed.

You can prevent this by putting a cap in series with the diode... 0.1 to 1.0uF

It may mess with the tone so I'm not sure how well it will work for you, but it is worth a try.

Best regards, Jack

[amptramp]

Once again, it you use the circuit of reply #28, move the output to the emitter of the upper transistor.

[Rob Strand]

In fact, the original NatSemi app note had different devices for the two transistors.

Jack, you should be known as the guy that brought that circuit to the effects pedal world.

It's now a widely known circuit.   Pilfered and used in commercial pedals and all!

The NatSemi linear applications notes are something I'm very familiar with and not once did I ever think to use that circuit in an effects pedal.

[amz-fx]

Rob,

It should be even more widely known that most developments of the boutique pedal world started in this forum, due to the contributions of the many excellent members that have hung out here over the years.

The amp-in-a-box (or Marshall-in-a-box) started here and was developed due to the work of Aron, Doug, Ed and numerous others. So too was the analog-sounding PT2399 delays that are now found everywhere... and numerous other ideas!

Thanks for your kind comments.

Best regards, Jack

[PRR]

> most developments of the boutique pedal world started in this forum

Yeah, but someone curated diffuse thoughts/threads onto the well-known AMZ website. That's a real service also.

[Rob Strand]

The amp-in-a-box (or Marshall-in-a-box) started here and was developed due to the work of Aron, Doug, Ed and numerous others.

I think we're up to the 15 to 20yr mark now,  scarey stuff!

[KarenColumbo]

https://www.tubecad.com/2009/10/blog0174.htm
Food for thought ... I wonder what happens if I breadboard some of these

[Steben]

http://www.geofex.com/Article_Folders/modmuamp/modmuamp.htm



NFB is a very good thing but, IMHO, not a nostrum/panaceia..
(for a mild overdrive, there isn't any reason for huge gain and respectively huge NFB..)

Anyone tested the mu-Valve with very different R1 and R2?