My Overdrive isn't clipping

[Electric-Gecko]

I am making my first "serious" attempt at making an effects pedal, which is an overdrive for my bass based on the Rose Screamer with a dual (bipolar) power supply.  But I'm having a strange bug that I can't figure out; it's that I can't get the signal to clip through the diodes.  Here is a schematic that I made of the overdrive section of the circuit that I soldered together:

But the strange thing is, when I have my 500k potentiometer up to full gain, my multimeter measures only 35 kilohms between the op-amp's output and inverting input.  However, if I bring it down to zero gain, and I measure the resistance from the potentiometers pins 1 & 3, then it shows 500k as it should.

[smallbearelec]

This thing is like a DOD 250 with a TS-style tone stack on the output. Something is surely not wired right, and I can't see why the designer suggested powering it bipolar. Take a look here:

http://diystompboxes.com/projects/intro.html

on the top left. The voltage divider arrangement is the way most 9-volt pedal circuits are powered from a single battery. I'd suggest taking apart what you have and breadboarding just the first stage. You'll want a small cap on the output, say .1 mf. When you have one stage working, add the tone stack.

[PRR]

> 35 kilohms between the op-amp's output and inverting input.

In-Circuit, the 500K is shunted by the diodes. (Also opamp guts.) Depending on meter parameters, 35K may be the "right" answer. (Diodes are not resistors so the actual reading depends on the volts and amps the meter uses to measure Ohms.)

And "35K" against the 2K minimum (with 10K pot shorted) is gain of over 16V. With strong guitar, -something- should be clipping.

Have you found the Debugging sticky-post?

[Cozybuilder]

Solder joints should be shiny, many of these are suspect.

[greaser_au]

It's hard to tell for sure, but I think it would be worth a look at how the diodes are connected - two of them look to me like they are possibly connected backwards.

david

[Kipper4]

Shouldn't the 470k go to V/2?

[Groovenut]

Shouldn't the 470k go to V/2?

Not in a bipolar power supply. Same with the 27k. Ground is the center point between V+ and V-

[robthequiet]

Worth going through with a magnifying glass before doing anything drastic. If it were my project, I think I would suck all the solder off the board and recheck the components, then re-solder with a bit more heat for better bonding.

[rutabaga bob]

There appears to be a component on the solder side, and a solder bridge nearby...possibly more than one under there. 

[LightSoundGeometry]

Solder joints should be shiny, many of these are suspect.

the copper looks oxidized to. I had several failed builds of an 11 part boost I have made 50 times, and thinking it might be the tip not heating right on the gun in combo with some bad strip or possible laziness on my part of not cleaning the components properly. I know it sounds crazy but small things like properly cleaning the copper and right amount of heat sometimes will make or break a build.

[greaser_au]

the copper looks oxidized to. I had several failed builds of an 11 part boost I have made 50 times, and thinking it might be the tip not heating right on the gun in combo with some bad strip or possible laziness on my part of not cleaning the components properly. I know it sounds crazy but small things like properly cleaning the copper and right amount of heat sometimes will make or break a build.

A ballpoint pen/ink eraser (the kind that contains an abrasive) is useful here. use the corner of one end for spot cleaning of the PCB.  Cut a slit in the other end for about 6mm (or 1/4") for cleaning component leads. Obviously, bulk cleaning of the board (just before assembly) is best done with something like scotchbrite or an abrasive cleaning powder like Ajax/Comet.

david

[Vitrolin]

But the strange thing is, when I have my 500k potentiometer up to full gain, my multimeter measures only 35 kilohms between the op-amp's output and inverting input.  However, if I bring it down to zero gain, and I measure the resistance from the potentiometers pins 1 & 3, then it shows 500k as it should.

if you want to compare measurements take them at the same place, measure R between out and inv.in at max and min gain.

[Electric-Gecko]

By the way, the solder joints of the diodes are actually on the far side of the board in the third picture, unfortunately out of focus.  That was a mistake when taking the picture.

Worth going through with a magnifying glass before doing anything drastic. If it were my project, I think I would suck all the solder off the board and recheck the components, then re-solder with a bit more heat for better bonding.

Maybe I should do that as my next step.  I have a desoldering iron (which is a great & underrated tool), so I could remove those diodes and maybe the potentiometer too and then measure the resistance between the pads.
I actually do use my soldering iron at about 330° Celsius, which is definitely not lower than recommended for lead-free solder.  But having OCD, I tend to rework the joints more than I should.  That area is quite a mess, as I actually removed one pair of diodes and replaced them before posting here, as my first attempt to get rid of the leakage.

I did take a magnifying glass and looked between the pads.  I just found a little debris that I pushed out of the way, but the resistance stayed pretty similar.

I do have a theory as to how I made this "invisible resistor"; I used a tool to scratch the board between the traces there to remove some hidden solder joints that I the multimeter detected.  While I was actually trying to separate them, maybe some solder dust just got pushed into the substrate and mixed in, forming a resistive connection.

The copper looks oxidized to. I had several failed builds of an 11 part boost I have made 50 times, and thinking it might be the tip not heating right on the gun in combo with some bad strip or possible laziness on my part of not cleaning the components properly. I know it sounds crazy but small things like properly cleaning the copper and right amount of heat sometimes will make or break a build.

I actually put this board together over the coarse of a week.  I noticed the copper going bad, so I tried polishing it with an electric drill with a brass-wire brush attachment, and then cleaned it with cotton swabs and isopropyl alcohol, which appeared to help.

Maybe I should remove all the components and remake this on a new board.  I have the layout figured out now, so I could remake it much faster so that the copper doesn't oxidize.  Now that I understand the relationship between capacitance and impedance, I could use a smaller input capacitor to have more room for the diodes.  I could also use sockets for the diodes.


I actually decided to put together a similar pedal on a breadboard, and test it.  I used a 330k resistor instead of the 500k potentiometer, and only 2 1N914 diodes.  It actually clipped well, and the resistance between the op-amp output and inverting input remained over 300k.  I tested it with both a dual & single power supply.  They both worked, although the bipolar (dual) supply made it less noisy, although I discovered that I should have a "stronger" virtual ground than on the board that I soldered.

[PRR]

> lead-free solder.  But having OCD, I tend to rework the joints more than I should. ... polishing it with an electric drill with a brass-wire brush attachment, and then cleaned it with cotton swabs and isopropyl alcohol

Re-re-working bad joints is bad practice and often a waste of time. Maybe one clearly identified joint, but not when you don't know how many joints are bad.

Soldering Basics. CLEAN copper. Clean leads. Ample heat. Good flux. Good solder. There's tutorials on-line.

If you don't have children on the floor, get Lead solder (electrical, not plumbing, not roofing) if at all possible. Few of the Lead-Free solders are easy to work with, and at best they can "appear" to be a faulty lead solder.

If you must stretch-out a build over weeks, so you can't re-scrub the copper, then first-off use a hot iron and a quick hand to *lightly* tin all the pads without filling the holes.

[Vitrolin]

Lead-free solder fumes are very dangerous worse than leaded solder, because of the flux used.
Always use leaded solder if you dont have adequate fume extraction.

and it is easier to solder and needs lower temperature.
330C is hardly enough 350 at least to avoid cold joint.

[Electric-Gecko]

Lead-free solder fumes are very dangerous worse than leaded solder, because of the flux used.
Always use leaded solder if you dont have adequate fume extraction.

Really? I thought that the flux in Lead-free solder was very similar if not the same as in leaded; just twice the quantity of it.  I actually just got a fume extractor though.

I'm going to rebuild this project.  I have found tin-plated PCB's to be easier to solder than the bare copper ones.
Do you think I should make a custom PCB for this (which will not be tin-plated) or should I use a tin-plated perf board?  I'm going to make a custom PCB for a different project anyway, so it won't be any trouble to make a second one.
Would it be better to mount the audio jacks directly onto the board, or should I use flexible wire?

[antonis]

Would it be better to mount the audio jacks directly onto the board, or should I use flexible wire?

PCB mounted jacks are prone to "cracks" ...
Wired jacks are prone to "faults" due to double solder points..

[anotherjim]

Some IC sockets - the kind with thin strip "folded" contacts, are very loose fitting in regular perf/vero board holes. I've found it's very easy, particularly if using fine gauge solder, to push a bit of solder straight through the hole past the socket pin where it can cause hidden havoc. So, in hunting mystery shorts/resistance around IC sockets, you may have to remove the socket to get at it.
I think it is possible to ruin the insulation of SRBP board with contaminants + excessive heat, but be sure to remove all components from the suspect area before you can prove that is the problem.