MXR Distortion plus bahaves strange!

[sajy_ho]

Thank you guys, I can't find any linear pots eighther! So I think I should go with Dshwart's idea about log pot in the feedback loop and a fixed resistor to ground.
But something that I don't understand here is how can type of the pot affect opamp clipping? I mean my only issue with my current arangement is when the gain pot is at max suddenly opamp transfers into severe clipping. I raised the supply voltage to 20V to see if hitting the rail voltage causing that problem, and surprisingly t wasn't! So that is because of the log pot?!
Sorry for being stupid

[dschwartz]

Youre not stupid..we all had similar issues and questions when we started..
Whats happening?

It is not the pot that is affecting the clipping, is the resistance ratio between the pot and the feedback resistor that sets the gain at any given pot position.
If you use a log pot, you get too litlle resistance even at minimal gain settings, so the gain is high on most of the pot rotation..with a linear pot, the gain change is more noticeable and useable..and 100k gives a decent range from 11 times gain to 240 times..

I prefer the use of a log pot at the feedback pecause they are easier to find, gives a natural range (gain from 1 to x) and makes a high pass filter when turned down, so it cleans up better

[Mark Hammer]

Pot taper never affects whether something works or not.  It only affects how easy it is to dial in settings you like, and how easy it is to repeat that setting in future.  If the taper of a particular pot places all the settings, and differences, you are interested in within a 5-degree arc, then it will be hard for you to get exactly the setting you want, and hard to repeat it...unless you have a very big knob and a very precise vernier around it.

[sajy_ho]

Thanks guys, now I finally got your point!
I also changed the resistor divider to 22k resistors like the DOD 250 and interesting news; the chip almost stoped clipping!
I think Mark's theory about the difference between Distortion+ and DOD 250 was right; because the distortion apparently is caused by starving the chip, not by hitting the rail voltage, because higher rail voltages didn't stopped it!
I actually find out I like the sound of Distortion plus(with 1M resistors in the divider), because as Mark said it is a dual clipping circuit and has a uniqe sound as it is, so I guess I'll get back to 1M resistors.
Anyway thanks for all your helps, learned alot from you guys..

Sajad

[Mark Hammer]

Many thanks for that.  I've always been too lazy to pursue the experiment myself, so I appreciate that you showed the initiative to follow through where I didn't.

The principle of "starving" the bias in some way is something that I don't think we've ever collectively explored.  If anyone of the more astute and informed EE types wants to take a stab at hypothesizing what it doies, I'd be similarly obliged.

[anotherjim]

The high bias divider resistance is interesting. I'm no expert, but lean to the idea that input bias current is allowed to cause some asymmetry by pulling the bias off from Vc/2 and this may be following the signal somehow - if the poor old 741 has some reaction between inputs (it's not a "perfect" op-amp). Otherwise, the bypass cap on the divider network should ensure that AC signal effectively sees only the 1M bias resistor and the high bias network values are irrelevant?

[sajy_ho]

I don't know man; but those resistors deffinitely affecting the chips headroom. With 22k resistors headroom improved but not too much and still there was ome clipping out of the chip. Maybe it's input impedance of the chip that causing asymmetry? Because using a high input impedance chip like tl071 also improved the headroom dramatically.

[anotherjim]

Found this scheme...

See the suggestion there to add a small capacitor on the feedback loop to prevent oscillation. Without using an oscilloscope to check the output, you can't say this is or isn't happening, but it could be a reason it sounds nasty at maximum gain.

Also, this is another old circuit that may have started out with battery power only, but with the added AC adapter option, has not added any local power supply decoupling capacitors to compensate for the loss of capacitance provided by the battery (battery is a pretty good capacitor).
A 100nF ceramic across the 741 power pins would help it's stability, plus a 100uF electro across the DC input to the pcb. - unless you intend to use battery only.

[sajy_ho]

Found this scheme...

See the suggestion there to add a small capacitor on the feedback loop to prevent oscillation. Without using an oscilloscope to check the output, you can't say this is or isn't happening, but it could be a reason it sounds nasty at maximum gain.

Also, this is another old circuit that may have started out with battery power only, but with the added AC adapter option, has not added any local power supply decoupling capacitors to compensate for the loss of capacitance provided by the battery (battery is a pretty good capacitor).
A 100nF ceramic across the 741 power pins would help it's stability, plus a 100uF electro across the DC input to the pcb. - unless you intend to use battery only.

Thanks man, Nice mod. I've seen that on ggg's schematic too. I'm gonna give it a try.

[Mark Hammer]

The Dist+ (but mostly the listener!) benefits from having a cap in the feedback loop to roll off the top end.  Values between 15 and 100pf will be helpful, depending on how much "bite" you like (100pf rolls off starting around 1.5khz).

For me (though it doesn't have to be for you) the objective is to have a sound from an overdrive that does not require adjustment of either the guitar's or amplifier's setting.  You get a throatier growlier louder sound, but nothing you have to compensate for. 

The Dist+ has no tone control but provides two points of influence where the top end can be rounded off to the user's taste: the feedback loop in the op-amp, and the cap in parallel with the diodes.  The 1000pf cap in parallel with the clipping diodes is not enough to do much good. 

Ideally, one would plan out the rolloff point of both the pre-volume cap and the feedback cap to achieve a pleasing 12db/oct rolloff, that maintains as much bite and upper mids as possible, but eliminates the fizz.  If you increase the 1000pf cap to 4700pf, and place a 47pf cap in the feedback loop, you get a (theoretical, given component tolerances) 2-pole lowpass filtering effect, starting just under 3.4khz, which gives just enough bite to be dangerous, but won't require you to fiddle with tone controls.  Naturally, your tastes and rig may be different than what sounds good to me, so feel free to experiment in an upwards or downwards fashon.

[sajy_ho]

Now I have to open the box again, my wife gonna kill me if she sees me around soldering gun this time
Thanks Mark for all your help...

[Mark Hammer]

Send her to me.  I'll set her straight.

Incidentally, all this kerfuffle is one of the reasons why I think the Distortion+ and DOD250 make an excellent learning platform for people just getting into this.  Many of the principles that will serve you well down the line, and allow you to know what you need to do with subsequent builds so that they sound exactly the way you want, first time, are illustrated in those pedals, in a nice uncomplicated fashion.

So consider this an investment towards saving time in the future!