gain of 10,000?

[C Bradley]

I've got a few 10M 1/4 watt resistors, and the idea of using them in a opamp distortion circuit in the feedback loop for a very high gain effect. Is this possible. Will the circuit be stable?

Chris B

[ExpAnonColin]

If I understand you correctly, you're talking about the gain in and outs of op amps?  In which case, adding some resistance can tone down the amount of gain a bit, depending, of course.

But I don't think I  have any idea of what you're talking about.

-Colin

[drew]

First of all- if you have 10M resistors and put one or more in the feedback loop, that is pretty close to open-loop gain (unless your input resistor is very high too.) It will probably not sound too spectacular on its own.

You can search on google for "op-amp basics" or a similar phrase... I found this guide which should explain the basics of using an opamp as an inverting/non-inverting amplifier:

http://ourworld.compuserve.com/homepages/Bill_Bowden/opamp.htm


If you want to get a book, I would suggest one of Walter Jung's "Op Amp Cookbook" guides, as it has a bunch of practical circuits in it which can serve as starting points for designing your own circuits around op-amps.


drew
toothpastefordinner.com

PS: Colin, nope

[C Bradley]

I have a pedal I built with a non-inverting 741 opamp circuit. 100k Rfb and 100 ohm resistor to ground through a .47uF cap. A 1k pot is between the 100 ohm and the 100k for gain adjustment.

I just switched the 100 ohm resistor to a 10 ohm, so now my gain is (100k/10)+1 = 10,001. It starts to oscilate a little towards maximum gain, but it's quiet with the guitar plugged in.  

So I guess it is possible. It sounds OK, but I'm still experimenting. I'd like to get a tone similar to what Billy Corgan got on the Siamese Dream album. Total destruction of guitar tone into a bright sounding buzz.

Chris B

[Tim Escobedo]

If you want to get a book, I would suggest one of Walter Jung's "Op Amp Cookbook" guides, as it has a bunch of practical circuits in it which can serve as starting points for designing your own circuits around op-amps.

Get the first edition, from, like, 1974 or so. It has lots of really good audio stuff inside. Later editions cut the audio stuff out, placed it in the "Audio IC Op Amp Applications" book, which is sadly out of print. It's contents never seemed to be rolled back into the "Op Amp Cookbook".

Those big resistors can be used, but you may have problems with stability if indeed your circuits are configured for very high gain. Not to mention noise. Generally, it seems JFET-input op amps are best suited to these large resistances.

[C Bradley]

Tim,

So, would the 10 ohm be OK for the 741? Is the impedance of the circuit too low? It seems to sound OK right now, and I might have to try one of the FET opamps with my 10M's for maybe a dual gain stage design with 10,000 per gain stage! :twisted:

Chris B

[idlefaction]

heh.  billy's tone has a name - 'big muff pi'    

if you listen close to siamese, the 'racing car' guitar tone is always coupled with riffs that don't have any stops in them...  except for 'hummer' where it lapses into sudden and complete feedback when he stops playing.  i reckon it was two big muffs in series!!!!

so if you'd like to get billy's racing car sound, build a big muff -  
http://www.generalguitargadgets.com/v2/index.php?option=displaypage&Itemid=52&op=page&SubMenu=

HTH!  (btw, i love this pedal and the people here can help you mod the "original" circuit to your tastes!!!!)

[Tim Escobedo]

Tim,

So, would the 10 ohm be OK for the 741? Is the impedance of the circuit too low? It seems to sound OK right now, and I might have to try one of the FET opamps with my 10M's for maybe a dual gain stage design with 10,000 per gain stage! :twisted:

Chris B

Hey, if it works, it works. Hard to argue with success.  :wink:

For a non-inverting amp, the gain setting resistors don't have a influence on input impedance (if that's what you mean).

[Paul Perry (Frostwave)]

I find it is easier to get 'ridiculous' amounts of gain by putting two lower gain op amp circuits in series.

[C Bradley]

heh.  billy's tone has a name - 'big muff pi'    

Yeah, I've got one. Mine's a great pedal but he must have an older model or has his modded. Of course his guitar helps a lot too. I've got a JB jr in my Strat where he's got Lace Sensors.

Billy C used a lot more than just a Big Muff, he has a huge assortment of fuzzes. I'm working on expanding my fuzz collection with a few of my own designs and adaptations.  :wink:

Chris B

[C Bradley]

if you listen close to siamese, the 'racing car' guitar tone is always coupled with riffs that don't have any stops in them...  except for 'hummer' where it lapses into sudden and complete feedback when he stops playing.  i reckon it was two big muffs in series!!!!

I tried my 10,001 gain pedal in front of my Big Muff Pi today and lo and behold, I had the "Hummer" tone!!!    I think I'll keep my 10,001 gain pedal as is, except for the addition of an LED to indicate when it's on.

Chris B

[Ansil]

i have never been one to say no to gain, but 10,000 isn't this a little ridiculous, i mean you could get dimebag crunch out of a tesco delray guitar with that. personally i have never really needed anymore than 300 to 400 usually jamming around 220 or so for smokin rhythmns.  only time i use more is if i am trying to jam the local radiostations lol... or build a sustain feedback device.  but just my two cents worth

want that really singin sound on siamese dream take a 386 with no input or output caps and bridge the gain from pins 1 to 8 put a 5mm led in parallel to the output, it works best if you put it on a amplifier that has a little grind to it, ie think ac dc's rock and roll ain't noise pollution.  where you know he has the guitar knob turned down but where it sets it would break up if you hit it hard but still pretty clean.  that is what i did to get the sound clips that i did like a year ago in the studio, we used an old peavy head i believe that was based on a fender into a 5150 cab with that pedal i described pushing it,  the bottom end is a little mushy so it doesn't work for real fast heavy metal but it is a staple that i use for that great strummy overdistorted sound, that is on the siamese album, well at least the songs i listened to as i don't own the album.  

[Mark Hammer]

How much gain is too much?

It depends on the amplifying device, the supply voltage and the input signal to be amplified.  A 100mv input signal can't be made more than 90x "bigger" with a 9v supply, even under the most ideal of circumstances, without some "corruption" of the signal.  Of course, in the case of distortion units that corruption is the endgoal, but it still remains an open question as to how much more one can/should go beyond that before the point of diminishing returns or even sonic costs.  I should point out that many op-amps will not swing "rail-to-rail" so that maximum output level prohibits anything more than maybe even a 5-6v output before clipping sets in.

The other thing, hinted at but not really openly discussed, is the question of where all this gain is situated.  EVERY amplifying stage contributes noise from its input.  Stick all your gain in one stage and 0.01mv noise turns into 1mv noise with a gain of 100, and 10mv noise with a gain of 1000.  Play some notes and a gain of 1000 manages to mask that noise, but take your hands off the guitar and that 10mv noise signal keeps on going.  If you *distribute* the gain over multiple stages, though, and stick in some lowpass filtering at each stage to keep the hiss under control, you can achieve large amounts of *cumulative* signal amplification without contributing much noise.

Example:  Two cascaded stages with a gain of 50 will get you a total signal gain of 2500, however the input noise from the second stage is only increased 50x.  If the input noise from the first stage (which is also increased 50x) is treated (and this can be as simple as a cap in the feedback loop of an op-amp to keep bandwidth reasonable), then whatever noise manages to make it from gain stage 1 to gain stage 2 is sufficiently attenuated that hiking it up by 50x still doesn't add nearly as much as amplifying that initial input noise itself by 2500.

A second very different reason to want to use cascaded stages is that giving a clipping stage a harmonically-rich signal to work with delivers an even more harmonically complex tone.  The Big Muff Pi is an example of that in action, and the "Chaos" Fuzz I posted on my site is another simpler version of that principle.  Situating just enough gain i one stage to achieve clipping, then sending that clipped signal to another gain/clipping stage gets you mucho harmonics, maybe even more than packing all the gain into one solitary stage, and with less of the nasty intermodulation.

[AllyP]

Bare in Mind Billy C used a marshall JCM800 100watt for that record:

He used it in a fairly unique way as well....see what you think:wink:

Turn the volume of your amp up to 10 baby!! Yeah read it....10 woohoo...

Use your pre-gain pot on the amp for volume.

I believe those amps had high and low inputs...he used the low one.

[C Bradley]

Hey Mark,

My design doesn't have the intermodulation distortion (I think that's what it's called. In tube amps most folks know it as farty bass.  :lol: ) that the Big Muff Pi has. I think this is due to the relatively small 4.7uF cap I used in the non-inverting amp for the ground on the inverting input / feedback loop side. I had a 0.47uF in there, but it was more like a metal / blues type distortion. Not bad, but I think I like the current design better.

I was thinking that if I designed a two stage opamp design, I could get the gain up into the millions!! :twisted:

Chris B

[drew]

To be honest, when you drive the gain up past a reasonable point, it just doesn't sound much different. I would suspect that a gain of 10,000 sounds similar to a gain of 100,000 and 1M, given the same circuit topology... high gains, to me, only get interesting when you have several stages and you screw with the sound in between each one.

Oh... and no matter how you filter it... once you get past a certain point, the noise is going to be as loud as your guitar signal...

[Mark Hammer]

Not trying to rain on your parade, but exactly how much need do you have for the sound of your fingers gently sliding on or being lifted off the frets being heavily clipped?  Perhaps more to the point, exactly how much hum-proofing does your signal chain have?  For my part, as much as I love fuzz, sick or thick, or hard as my.....( :oops: got a little off track there), my first inclination is "Hmm, I wonder how many dogs I'd send into therapy if I plugged a single-coil equipped guitar into that anywhere within a mile of a flourescent light?"

Once again, I invoke the proximity-to-clip principle.  One needs as much gain as is required to bring an existing signal (and the desired portions of it) up to the level at which the given circuit will clip.  If the signal is weak and the threshold high then more gain is needed.  If the signal is fine and the need only for clipping peaks rather than any playing level, then less gain is needed.  If you plan to have all useful segments of the signal transformed into square waves, then some gain and some threshold-lowering may be sufficient.  It's the gap between where the signal is NOW and where it needs to be in order to clip that yields the tone and the design requirements.

At a certain point, one ends up being able to draw a line between the signal floor and noise ceiling; i.e., everything above this is absolutely part of the signal, and everything below this stands an excellent chance of being just noise, whether thermal or mechanical or whatnot.  Applying large amounts of gain can end up treating everything as if it is signal.  If that is one's intent, well fine, but if there is some intention of being able to provide pitched notes in the midst of the thickness and sickness, then some restraint is called for.

At the very least make a special effort to use semiconductors known to be very low noise and metal film resistors.

[C Bradley]

I guess my purpose is experimentation. I want to see what those levels of gain sound like. There is a very real difference between gains of 1,000 and 10,000, and I suspect that there's a difference between what I've got now and a gain of 100,000.

Chris B

[Doug H]

Not trying to rain on your parade, but exactly how much need do you have for the sound of your fingers gently sliding on or being lifted off the frets being heavily clipped?  Perhaps more to the point, exactly how much hum-proofing does your signal chain have?  For my part, as much as I love fuzz, sick or thick, or hard as my.....( :oops: got a little off track there), my first inclination is "Hmm, I wonder how many dogs I'd send into therapy if I plugged a single-coil equipped guitar into that anywhere within a mile of a flourescent light?"

Mark, I nearly fell out of my chair laughing when I read this! Funny stuff!

I'm not sure if I prefer heavy even or odd order harmonics on my finger squeaks though...

Doug

[Mark Hammer]

If my little jibes deliver half as much pleasure as your mini-amp posting yesterday, then I've done my work for the day.  Glad to put a skip in your step, Doug.