anyone interested in my new overdrive?..

[Johan]

I drew this up on my lunchbreak today and have not had time to test yet, so you've been warned..   ..but I think it should work as intended..
it's three stages of clipping with a clipping succecion like this. starting from a very low level signal and increasing to full drive. the first clipping occurs at the end of the circuit with symetric softclipp mimicing a push/pull output.
increasing the signal, the second stage hard clipps assymetricaly as would happen in a typical amplifier. increase signal more and the first stage hard clipps assymetrically.

the opamp bias is offset to 6 volt instead of 4.5volt to allow for more assymetri and the gainpot adjusts gain in both opamp stages
the goal is to have a distortion/overdrive that responds well to picking dynamics and also goes in and out of clipping a little more gracefully than what I'm used to from solid state amps and stompboxes...since I havent built it yet, I cant make any claims about how well it does, field testing necessery...perhaps it sucks.. ..

anyway. if anyone is interested in trying it before I get a chance...here it is...http://www.aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=44203&g2_serialNumber=1

any thought on this?...now I'm off to band practise..

J

[Mark Hammer]

You're using an LED in stage one, and you expect to acheieve some sort of asymmetrical clipping with a max gain of 22.3x?  Dream on, my friend.

You will likely get some clipping in stage 2, because 10x22.3 puts you up in a range where some clippipng will occur with the second LED (though not that much, unless stage 1 is maxed).

I'm thinking that you probably need to shift some additional gain to stage 1, and away from stage 2.  Say, make max gain of stage 1 something like 200x (make that 4k7 a 470R instead, and increase the cap to 1uf), and make stage 2 gain 3x (3k3 input resistor).  That will allow you to get enough gain to clip both LEDs if you want, or asymmetrical clipping (LED2 but not LED1) if desired.

You also need to add some feedback caps to stages 1 and 2 to shave off those ugly harmonics of harmonics of harmonics.  Multistage clippers generally always need to have the treble gradually spooled out to sound smooth.

[ayayay!]

Reminds me of a Daddy O.  

[jrod]

Simple! I love it!

Can't wait to see a final version.

[dschwartz]

i have been so away from designing and building these days (i just became a daddy). But in my latest experiments i was taking the same approach..
the assymetric clipping thru stages using just 1 diode on the feedback and let the opamp clip the other side. but i used 4 stages with mild gain, and ended up with tons of gain with very nice touch sensitive feeling, but lots of noise and some biasing issues , too..
i´d recommend you to go further with this..just consider filtering and voicing on the design..IME, sound quality is 10% clipping device, 20% playing techniques, and 70% pre and post filtering...

[Johan]

thank you all.
Mark.
you might have a point about not geting clipping in the first stage, but I think you're missing the point somewhat(with all due respect). with a max voltage gain of ~200, there will be enough gain to have the output clipper working with even very weak signals. the intermediate stage adds harmonics/clipping to a large portion of the duration of the cords or single strings ringing and the first stage should be able to add gritt to the pick attacks. at least thats the idea.
My burtsbuckers put out~0.25 just after the pick attack when struming on my les paul, so even with low gain it should add a little something with the gain up. the main distortion is still coming from the clamping at the end. 
just like on your little princeton, the first stage doesnt add much, but take it away and the rest becomes bland.

again, I havn't built it yet and you may wery well be proven right, just wanted to bring it up.

and about filtering, yes most likely something needs to be added but it will take some experimenting..a have an idea for that too, but first things first..

ayayay.
..Daddy-O...isnt that one based on Marshall bluesbreaker pedal?...the dual function gainpot is defenetly borowed from the BB. I believe one of Nobels dirtboxes and at least one Fender solidstate amp uses the same idea too..

I think I'll get some time over later tonight or tomorow to start breadboarding and experimenting
J

[Mark Hammer]

Well, certainly don't take the 200x as some sort of requirement.  My point was simply that the small amount of gain in the first stage made the feedback LED completely superfluous.  That's not necessarily a bad thing, merely a waste of an LED.  The recommendation for more gain in stage 1 was simply to make actual use of that LED, since I assumed that was your reason for including it in the first place.

At the same time, you are quite right that the brunt of any clipping is coming from the diode pair on the output (though the 10k/10k divider is going to deliberately make that clipping softer).  So, getting clipping out of ony the second LED may well be a good thing, insomuch as it introduces a certain degree of asymmetry.

[Johan]

...though the 10k/10k divider is going to deliberately make that clipping softer...

..yes. the idea is that if you would insert a strong ( say >3volt) sinewave on top of that string of components, you would certainly get flat tops on the diodes, but what do you get between the two 10k's, where the 100k pot is listening?...it has to be severely softened as compared to the top of those diodes...at least as long as the sinewave is provided by a low Z source.  by adjusting the ratio of those two resistors, you could adjust the softness of the clipping as long as the source isn't clipping, then it just becomes a string of resistors, but that's a different discussion...
J

[ayayay!]

ayayay.
..Daddy-O...isnt that one based on Marshall bluesbreaker pedal?...the dual function gainpot is defenetly borowed from the BB. I believe one of Nobels dirtboxes and at least one Fender solidstate amp uses the same idea too..

Guv'nor, IIRC.  FWIW, I actually like the Daddy O when properly modified.  Stock tone stack sucks, but it's got potential.  I think your circuit looks promising.  

[Johan]

here is a quick sample of "as drawn" no modifications yet.. it is a little bright, but not overwhelmingly so..the gain is at max. it cleans up in a nicer way if I back off the gain a little, but I wanted to show it at max gain if anyone wonders how much gain it has. there can be quite substantial boost with the volume up, but now its set to match the clean amp, so all distortion comes from the circuit and not from overdriving the amp

all changes in gain in the sample is from the volume pot on my guitar
http://www.aronnelson.com/DIYFiles/up/distbox.mp3   ...sorry about the playing..

any thoughts..
J

[ayayay!]

I really like it!  No really, I really do like it.  I usually don't like a lot of the recordings that end up here (sorry all) but it has some great things going on.  Reminds of me of the BluesBreaker II pedal I had a while back, but from the sounds of it I like this better.

Kinda reminds me of Nugent's "Stranglehold", with a little something more that his tone didn't quite have.

Hmm...  I'm about to start breadboarding again...  Is this an "as built" just like the schem?  

[slacker]

Sounds great, nothing wrong with the playing. I've always liked that Marshall building block.

[Johan]

...  Is this an "as built" just like the schem? 

yes, as drawn in the schematic in the first post..

[ayayay!]

Okay thanks.  Those last two resistors in the last stage, those are 10k, right? 

[Johan]

yes, its two 10K, but I dont think it's crucial, you could probably just as well use 5K or 15K. and the diodes at the end is regular 1n4148, the opamp in my build is a regular TL072. the bias resistor could probably be increased as long as you stay with a TL072 but that would give a brighter tone and limit your choises of opamp's more
J

[ayayay!]

Cool thanks.  I would probably use a TL072 anyway, I like them a lot. 

Nice job Johan! 

[BadIdeas]

That sounded great!
Some questions about the theory of operation...
It looks like turning the gain up on the first stage also increases the gain of the second stage, am I right? On the other hand, what would happen if the 10k input resistor on stage 2 were connected between the gain pot and the 4k7 resistor in stage 1's feedback network? If you used different diodes, it would sort of "pan" between the two "flavors" right?
Furthermore, if you replace that resistor and the stage 3 diode resistor with trimmers, you could have a very flexible circuit.
I will admit though, I get very confused with cascading gain stages because I always think that one will prevent the other from clipping.

I noticed that stage 2 is the only inverting stage in the circuit, so you may want an inverting input buffer, which will also ensure the low Z input that you mentioned.
BTW, I think you played more like a guitarist than a pedal maker.   

[Gordo]

all changes in gain in the sample is from the volume pot on my guitar

Sounds Great.  I really like the way it cleans up with the volume knob and has a bit of grit to it.

[BadIdeas]

Hey, don't know if you've done any work on this recently, but I just breadboarded it this evening and got it working after a few wrong connections. I followed the schematic exactly, but with an NE5532. My findings are that it needs more gain and some filtering, so I think I will change some things around tomorrow. 'Til then...
You used red LEDs right? It seems like a logical assumption, but you know what you do when you assume...

[caspercody]

Sounds nice! Thanks for sharing!