Distortion+ question ??

[Marty]

I just got finished building a Distortion+.  It sounds exactly the same whether it has diodes or not.  Why is that?

[earthtonesaudio]

Sounds like it's troubleshooting time.  It should sound different with the diodes.

[Marty]

It's funny because it seems to be working normal.  I've never played a real one so I don't really know

[ppatchmods]

check the resistor right before the clipping section. if it's not soldered right it can cause a boost that can get distorted and make the diodes sound different.

[aron]

At the very least, the output volume should drop when the diodes are connected.

[Marty]

The volume does drop a hair. but it sounds exactly the same

[Joe Hart]

Are you trying it through a clean amp? That way you can really hear the pedal. Sometimes the amp's distortion can mask what the pedal does (or doesn't do).
-Joe Hart

[Ice-9]

One thing to remember with the Dist+ circuit is that even with out the diodes the opamp is driven into clipping, but you should beable to hear a difference to the opamp and diode clipping.

[petemoore]

  No diodes should = boosted output.
  Debugging, what to do when it doesn't work...a sticky thread..
  Let's you find out if it's working for one thing, for all other things it is the best reference for fixing and finding things out about circuits I know of.

[Mark Hammer]

The Dist+ IS like that.  I haven't done the legwork, but my observation of the differences between the D+ and virtually every other similar type of clipper suggest that the bias to the non-inverting pin is "starved".  Look at the Vref source in virtually any other pedal (including the DOD250, which is very close) and you'll see that Vref comes from a rsistor pair that is generally 47k/47k or less (e.g., 22k, or 10k).  The D+ uses a 1M/1M pair and then feeds that Vref through yet another 1M resistor.  So, just how much bias current does the noninverting input get?  Not a heckuva lot.  And keep in mind that the 741 is not exactly a state-of-the-art chip, so it likely requires more than that for any sort of reasonable performance.  In effect then, the D+ provides a sort of double clipping that results from the starved biasing, followed by the more obvious GE diode pair.  Lift the diode pair, and you still get the one source of signal perversion.

Let's do a collective experiment, though, to see if I'm right.  Those of you who have a DOD250, with the 22k/22k+470k biasing scheme, lift the diodes and tell me if it distorts at all when set to reasonable gain (i.e., less than half-max gain).  My prediction is that it won't.

This sort of thing must drive novice builders crazy.  The D+ is often recommended as an instructive and useful project for first time builders, and when the diodes don't seem to make a big difference they must end up scratching bald spots into their scalps!

[ayayay!]

Marty, you're not using clear LEDs for clipping are you?

[Marty]

So that is the nature of the beast?  so the clipping is coming from two different locations?  Another question.  what kind of pots should I use?  I've seen a few different recommendations.  I'm currently  using a 1M for the drive and a !0K for the Volume.

[Caferacernoc]

Before I built anything from scratch the first thing I did was mod an old Ross distortion pedal. It is real similar to Dist+. I was suprised to learn how much clipping comes from the op amp. The diodes and cap to ground "shape" the final product but it clips either way. However, you should hear a big volume drop when you add the diodes.

[Mark Hammer]

So that is the nature of the beast?  so the clipping is coming from two different locations?  Another question.  what kind of pots should I use?  I've seen a few different recommendations.  I'm currently  using a 1M for the drive and a !0K for the Volume.

You really only need 50k-100k for the Drive, since the first 900k is wasted on getting slight gradations of unclipped (by the diodes, anyways) boost.  It is those slight-to-modest gain settings where you'll notice the presence of clipping even without diodes.  Once you crank the gain up full, it is easy to expect clipping from the op-amp simply because of voltage-swing restrictions.  The diodes simply add to that.    The distortion produced by what I believe to be a starved bias (and please correct me if I'm way off here) is only obvious for those settings where the gain is no higher than 30 or so. 

For the Volume, you'll conserve much more of the inherent volume of the unit by using a 50k-100k volume pot.

[WGTP]

So Mark, your thinking with the starved bias that the chip is distorting at lower output and not putting out enough signal to get clipped much by the diodes???

[Mark Hammer]

So Mark, your thinking with the starved bias that the chip is distorting at lower output and not putting out enough signal to get clipped much by the diodes???

Oh no, it gets clipped plenty once the gain goes up, as it should.  It just doesn't behave itself as a low gain amplifier.

Now, I know that op-amps vary in terms of what they need for biasing in single-ended mode.  And I know that the Vref circuit will be a function of the total number of op-amps in the circuit and the total biasing current requirement.  But the dual 1M pair is pretty idiosyncratic in the world of op-amp circuits, so naturally that's where my attention drifts to.  I guess the other experiment to try is to replace the 1M pair with a 22k-47k pair (à la DOD250) and see if the unit runs clean in the absence of diodes.