MXR Distortion+ Build Report

[frequencycentral]

I just built one of these from this schematic, which is from the sites schematic section:



Now, if you read the notes, bottom left, you could be mistaken for thinking that the 1M resistor between the 741's output and negative input is a mod to prevent click prevention. I was!

After an unsuccessful debug I checked the schematic against the DOD OD250 schematic: http://www.montagar.com/~patj/dodoverd.gif
This confirmed that said 1M is in fact necessary and not a mod, so I popped the resistor in and it works fine.

If you read this Aron, R.G. or Jack, you might like to amend the schematics wording - though now I got it working I feel pretty dumb for misunderstanding those notes!

Not having any 1n34 diodes, I'm running mine with two different diode arrays:
1. 1n4148's - buzzy and fuzzy like the schemo says
2. red LED's - more bottom end and more volume.

Just for fun, I tried replacing the 741 with an LF351, then a CA3130 both of which have the same pinout. Neither worked as a direct replacement. I'm not technical enough to know why, though both the 351 and the 3130 have the 'offset null' feature which the 741 lacks. If anyone can tell me why that would be great!

I managed to squeeze the circuit onto an 8 hole by 18 hole piece of perf! I'll be glad when I get back to working on my modular synth so I can 'spread out' a bit!

It sounds pretty good - and certainly nails those late 70's punk tones!

[sshrugg]

Hey, that's pretty cool!  I have one with a coupla germaniums.  I'm building my silicon one today.

*high five* glad you dig it.

[Mark Hammer]

The number of duplicate redraws of this circuit that are posted is enormous.  Damn shame you stumbled onto, and worked from, one that was confusing to you.

Do note that when people mirror, or worse, re-post schematics, that does not mean they have somehow signed a contract that requires them to check back with the source to see if there were errors.  This is one of the reasons why those who do take the time to draw circuits get a little annoyed when they are reposted somewhere else, instead of linked to.

The Dist+ will "work" with a wide variety of op-amps.  I can verify this from pulling them out and popping them in by the fistful.  However, the biasing circuit (specifically the 3 x 1M network that feeds the + input pin) may be inappropriate for some other op-amps and deliver less than optimal performance.  Note as well that the CA3130, a good op-amp, needs a compensating cap to work properly.

The circuit will provide considerably more output with use of a 50k-100k output volume pot instead of 10k, and with a pair of silicon diodes (1N014 or 1N4148) instead of the germanium ones shown.  Also, if you find the loss of bass at high gain settings to be annoying, change the .047uf cap in series with the gain pot for .22uf.

[frequencycentral]

The circuit will provide considerably more output with use of a 50k-100k output volume pot instead of 10k, and with a pair of silicon diodes (1N014 or 1N4148) instead of the germanium ones shown.  Also, if you find the loss of bass at high gain settings to be annoying, change the .047uf cap in series with the gain pot for .22uf.

Yes, I used a 100k lin pot for the output volume, and a 1M lin for the gain. The output is definately higher with LEDs. Incidentally, this circuit even distorts with no clipping diodes connected, which I didn't expect - the difference between no diodes and LEDs is subtle - with the 1n4148's its grainier, with less bottom end and a small volume drop. I will try the cap bass mod you suggested!

Note as well that the CA3130, a good op-amp, needs a compensating cap to work properly.

I have a stash of vintage CA3130s, CA3080's and CA3094's - with the round metal cases. I can use the CA3080's and CA3094's for VCFs and VCAs, but am yet to find a suitable circuit for the CA3130's.

I hope to get it boxed tomorrow - using my first attempt at waterslide decal - so pictures to follow!

[frequencycentral]

Finished!

[gigimarga]

It is possible to use 250K for gain (it's the single rev pot that i have)?

[Mark Hammer]

It is possible to use 250K for gain (it's the single rev pot that i have)?

Yes.  You may want to use a parallel resistor to alter the taper, though, since there usually isn't much of interest happening to the tone until you get down to around 50k of pot resistance (gain goes up as pot resistance goes down).

[gigimarga]

Thx Mark again !

[frequencycentral]

It is possible to use 250K for gain (it's the single rev pot that i have)?

Yes.  You may want to use a parallel resistor to alter the taper, though, since there usually isn't much of interest happening to the tone until you get down to around 50k of pot resistance (gain goes up as pot resistance goes down).

By 'rev' do you mean antilog? I've only ever used log and lin. A 1M log definataly wasn't going to work - I tried it - thats why I used a 1M lin. What was the gain pot on the original? If it was antilog is there a simple mod to make the lin smoother across its range?

[frequencycentral]

........and a gutshot!

[Mark Hammer]

The original used a 1M antilog.  It did so in order to get through the first 900k-950k of pot resistance as expediently as possible because gain would not change appreciably until the resistance started to drop.  People cling to that value out of some sense of authenticity.  However, I can pretty much assure you that monkeys will fly out of your butt more frequently than the number of times you are likely to set the gain pot to a resistance >100k.  If what you wanted was a modest clean boost of x3 or something, then maybe.  But there are a large number of notes posted on this forum from people remarking that you can simply NEVER get a Dist+ to be clean...ever...even without clipping diodes.  So, you will use other means to achieve yuor clean boost.  Ergo, the first 900k is absolutely useless.

[frequencycentral]

The original used a 1M antilog.  It did so in order to get through the first 900k-950k of pot resistance as expediently as possible because gain would not change appreciably until the resistance started to drop.  People cling to that value out of some sense of authenticity.  However, I can pretty much assure you that monkeys will fly out of your butt more frequently than the number of times you are likely to set the gain pot to a resistance >100k.  If what you wanted was a modest clean boost of x3 or something, then maybe.  But there are a large number of notes posted on this forum from people remarking that you can simply NEVER get a Dist+ to be clean...ever...even without clipping diodes.  So, you will use other means to achieve yuor clean boost.  Ergo, the first 900k is absolutely useless.

That explains the uneven sweep on my 1M lin - practically it's just fine though, I used it at band practice last night - though when first built, I did initially think I had a duff pot!

[drewl]

Has anyone gone lower than 500K for the gain pot?
I had a 1M in a new build and you guessed it, no good dirt until the very end so I went back to a 500K.
Just wondering if dropping it to a 250K or even 100K would lower the overall volume or boost.

[Mark Hammer]

I use 100k for one, and I think 50k for another.  Keep in mind that the gain here will be [feedback resistor + pot resistance + 4k7]/[pot resistance + 4k7], and that most distortion pedals will generally require a gain of at least 50x or more to produce discernible clipping.  You would think that if the pickups put out 200mv, and the GE diodes clip at 250mv, that even a gain of 10 ought to get you serious clipping, right?  However, that 200mv only occurs for slammed chords, and rarely for unwound individual strings, and even when you slam a chord, the supposed 200mv only exists on that first quick transient, with a fairly rapid drop down to maybe 20-50mv shortly after the pick attack.  If you want to hear distortion the string has to provide 250mv or more for at least a half a second.  That means adding more gain than you thought you might need.

With a 1M feedback resistor, and a 100k gain pot set to max resistance, your gain is 9.5x.  If you can get any clipping from the diodes with a gain of 10, you have my profound respect.  Drop the pot resistance by half to 50k and your gain is now 18.2x.  Still only a hint of grit.  Drop it to 10k and now we're at gain = 68x. Now we're getting somewhere.  Max gain of the stock Dist+, regardless of what pot value you use is [1M + 4k7]/4k7 = 213.  So even with a 10k pot resistance, we're still quite a ways from the max gain.  You can begin to see why you only really need 50k or so to tinker with.  Hell, I'd say that for a great many of you, all the stuff you're interested in lies betwen 25k and 0k.