Gentlemen, don your deerstalkers - the game is afoot!

[moid]

This is rather odd, I have a clear (well I thought it was clear) memory or replying to all these posts, but I there's no sign of what I wrote. I popped back to ask another question (see below), and couldn't see any evidence of what I'd written at all! So I apologise to you three chaps who it looks like I never replied to!

Very nice-looking builds!  Certainly puts ME to shame!

I made a layout and built a board according to what you posted earlier, with the addition of a .1uF cap just before the blend pot.  Haven't tried it out yet.  Hope it works like the original effect.

Thanks very much! And don't be hard on yourself; I rarely get to decorate any of my builds; the majority end up in bare aluminium boxes with bits of tape stuck to them that I write the names of the controls on! The box for my friend was a commission; and the one for my son got decorated because he complained that he can't remember which bare metal box does what, and could I put some colour on it to help him? Anything I build for myself rarely gets any artwork.

I hope your build was successful. I may have to build one myself; my son says he still prefers the sound of the original rogue leader with the oscillation bleed through... I may even attempt it on perf board... maybe... I've never built anything on perf board that actually worked, but the circuit on vero board will be quite large I think, so when I get paid at the end of the month I might buy some perf board. And then remind myself about how much I hate perf board and how everything I make on perf is always dead.

that knob is hilarious. I thought it had gone all wrong in the knobless pics, but now I see it. and yes, might as well be 1uF, it's in the right place.

I don't know the 4046 at all, and I can't hear anything in your vid except the whines of this crappy lappy, but it's possible you were getting unwanted digital feedthrough from the 4046 coupling into the output of that version.

I would never leave a pedal knobless! It would be a crime! I mean how could I stand here, in this forum; knobless? You're very accepting chaps, but I think that takes the knoblesse oblige to the limits

unwanted digital feedthrough sounds likely - it does cut off a little while after the guitar notes decay; it's not a constant sound thankfully.

That type of blend control pot can't silence the unwanted side. You can disable a 4046 with a switch to pin 5 but you can get away without it.

OK I didn't know that - so even if I used a higher resistance pot (like a 1M), even at maximum setting it would still allow some of the 4046 to bleed through? That's odd. Well fitting an SPST is no difficulty to control it.

The 4046 VCO should drop down to subsonic speed when the signal on pin14 stops and that's effectively silent. I think that diode is part of a noise gate and I do wonder if the other diode to pin3 shouldn't be in reverse across the one already going to pin14 and pin 4 directly links to pin 3. The 386 output self biases to 1/2 supply volts. By coincidence, the 4046 signal input is also biased to 1/2 supply volts. So the 386 output would have no effect except when the signal out of it swings above or below a diode volt drop. Simple noise gate.
Pin4 is the VCO output going thru a cap to the blend pot. Pin 4 should also link to pin3 which is the clever bit.
Pins5&6 have a cap between them. That's the oscillator timing which works with the resistor on pin 11 with the pitch pot to 0v.
Pin 2 is the output of the brain that works out if the signal into pin14 is the same frequency as the VCO into pin3. It goes thru the Depth pot to pin9 which is the VCO control voltage input. A cap is added to smooth out the control corrections from the brain together with the resistance of the Depth pot.
If the input signal stops, then the brain thinks it's running the VCO too fast, so it tries to drop the VCO frequency control volts to match which it can't do more than letting it go all the way to 0v. The VCO frequency becomes glacial - you might hear a tick occasionally as it toggles.

Thanks for explaining that! The 4046 does stop making noise a little while after a guitar note rings out, so there is some sort of noise gate in there - this circuit is obviously a bit more thought out than I gave the original builder credit for (still disagree with his decision to coat the circuit with blobs of hot glue though!). It sounds like it would be worth me socketing that diode that you think might need flipping if I build my own version.

Onto my new question!

This circuit uses a 386 opamp, which is only one channel. Would it work with a different dual channel op amp like a TL072, and would it then be possible to take the output of one channel of a TL072 and feed it into the input of another to get an even louder signal? I was looking through the datasheet for the TL072 and all the amplification examples just use one channel, so the other is presumably available? But then none of the examples showed a gain of more than 10, so presumably the 386 gets used because it can have a gain of 200 - even if the TL072 used both channels, that would still be quieter than using the LM386 with a gain of 200?

[anotherjim]

This is how I think the diodes might go (drawing from IainPunk IIRC)

The left-hand pair is what I mean. "Coring diodes" they are.

A TL072 amp is just an opamp - it can have any amount of gain, far higher than 200 - just make the feedback resistor value larger.
But the 386 clips nicer. 2 in series (with coring diodes in between for noise control) can give a ton of gain (stage gains multiply). See Acapulco gold...

[moid]

Thanks, I had no idea that the Acapulco Gold was so simple on the inside; the marketing said it was supposed to be a Sunn amp in a box! Something tells me that a Sunn amp would be a bit more complex... I will tell my friend about this, although he really loves the clean transparent overdrive sound of the circuit I built! To me it just sounds like normal guitar but louder (which is a bad thing for me because then you can hear how badly I play... whereas for him and my son, they just sound great through it!).

So you would put the coring diodes inbetween pin 5 of opamp 1 and say C3?

[anotherjim]

This is a similar scheme - just using a different amplifier idea...

Coring diodes go in the path to pin 14.
Note link pins 3 & 4.
If you take the 386 output to the blend pot from after the diodes (pin14 side) the gating effect of the diodes can make HM style palm mute chugging more defined.

[rutabaga bob]

Any of you tried to make the original circuit?  Just finished trying mine.  Just fuzz.  Zero sonic mayhem.

Edit: been going over this.  Built mine like Moid's layout.  I am speculating that something(s) that Jim writes about are where the problem lies.  I also have trouble wrapping my head around the fact that one side has output from an amp chip, but the other doesn't get that same level of boost.  You would think it should...or at least I do.  Hey...I never said I understood this stuff.  ;-)

[anotherjim]

I'm assuming the 4046 is used in frequency following mode, which is what this phase-locked-loop chip usually does.
It gets the input signal at pin14 and compares it with its own signal at pin3 and tunes its oscillator via a control voltage to pin9 to match. That's why pin4, the oscillator output, is usually connected directly to pin3.

If I also assume that the diode from the amp into pin14 is correct, might the other diode actually go from pin4 to pin3 to match the pin 14 input? That's what I think, although I can't see how it adds much.

There shouldn't be a problem with the blend as the output from pin4 is just as strong as a full clipping signal out of the LM386 & possibly stronger especially as the guitar decays.

[rutabaga bob]

So, you don't think that diode goes from pin 3 to grounded pin 5?
The swing on the blend pot is quite different...from strong gain to quiet.  Am probably going to cut the pin 1 - pin 8 connection and see if that brings a closer balance, and jumper pins 3 and 4 on the 4046.  Will report results.  Thanks!

[moid]

I haven't yet built the original circuit; once I get paid at the end of the month I will buy some parts (I don't have the ic for example) and have a go on breadboard to see what happens - but this will be a couple of weeks. Not that I know how these things work either; but perhaps a breadboard version will allow us to figure out why it works or doesn't work?

[anotherjim]

So, you don't think that diode goes from pin 3 to grounded pin 5?
The swing on the blend pot is quite different...from strong gain to quiet.  Am probably going to cut the pin 1 - pin 8 connection and see if that brings a closer balance, and jumper pins 3 and 4 on the 4046.  Will report results.  Thanks!

That side of the chip is just CMOS logic gates, so a logic output feeding a logic input doesn't need anything more than wire!
However, if pin3 isn't meant to connect to anything except that diode to ground and if the cathode connects to pin3, it will pull down to 0v via the diodes leakage in the same way the gate diode of a MOSFET Millenium Bypass pulls up - but I don't see why a wire to 0v wasn't used to do that.

If we think it isn't a frequency follower then my ideas change. It may be a frequency modulator...
Let's assume the signal is meant to pass from pin14 to pin2 and directly modulate the 4046 oscillator pin9 control voltage. This is using Phase Comparator 1 in the chip which is only an exclusive OR gate. For that gate to pass anything, one input is tied to 0v which is pin3. 
The problem with this is it may not shut up when you aren't playing. Pin14 is not a straight logic gate input but an amplifier as shown in the block diagram. It's made of a chain of inverters biased to 1/2 supply voltage. To shut up, the output pin 2 needs to go to 0v. The diode in series with pin14 input may be meant to facilitate that -  but I don't see how that can be guaranteed.

[rutabaga bob]

Jim: just took another look at the photos posted at the beginning.  It doesn't appear that there is a bridging connection on pins 3 & 4, at least to me.

[anotherjim]

So it's looking like it's a modulator. If you have it wired as per Moid, you should have 0v on pin3 and when you play a wobble thing coming out of pin4.

[rutabaga bob]

With just power into the board, I'm getting .2 - .26 volts at pin 3.  All is unhooked now, so playing would take a while to wire.  Went ahead and cut out that 386, pin 1-8 bridge, as well.

Edit: rigged test leads a different way.  Reading now .08V @ pin3.

Edit 2: all things connected to test...same results as before - no sonic goodies.  Don't think I built it wrong, but...

[moid]

Hi chaps

I hope I haven't made a mistake (really sorry if I did). I looked through the folder of images I took of the original pedal and found a couple where I had removed most of the hot glue from the perf... not sure if they will help? The top of the large IC is at the bottom left of the perf in the photo - these are both quite big, click on them to enlarge them.



oh and here's one of the perf from the other side:

[anotherjim]

Pin3 is a logic gate input. It doesn't need exactly zero volts input to read it as a logic 0. Signal into pin14 should appear squared up out of pin2. If it doesn't, something's wrong. Actually, pin3 can be up at supply voltage (logic 1) too as long as it isn't floating about.
Does anything happen with pin9 voltage when you play into it? If pin9 is 0v the output pin4 will be practically silent. If pin9 is close to supply + voltage, it could make it so fast the pitch is too high to hear. The pot feeding pin 2 to pin9 must work right and the cap from pin9 to 0v not too high value.

[rutabaga bob]

No voltage @ pin 9, whether pot is all the way up or down.  .001V on the digital meter.

Edit: ooops!  Made a mistake regarding pin 9 measurement - it always helps to push the footswitch to the 'on' position. 

Anyway, voltage hits max 4V when string is plucked, then drops off.

[anotherjim]

Any audio out of pin2? Any audio into pin14?
What voltages do you have either side of the diode feeding pin14?

[rutabaga bob]

Yep.  Audio in at 14 and out at 2.  At idle, diode to 14 has ~4.4V on 386 side, ~4.2V on pin 14 side (reading fluctuates as it's hard to make a steady connection with a pointy probe).

[anotherjim]

So then, if pin9 volts rises, the pitch of the signal out of pin 4 should appear to rise too. Make sure pin5 (Inhibit) is grounded to 0v as if not, the VCO doesn't play (everything else still works).
The VCO control voltage on pin9 is being fed a square wave that is averaged to DC by the pot resistance and capacitor. If the signal out of the 386 was symmetric the pin9 voltage will tend to be the same and give the same pitch out of pin4 to the blend pot. Clipped guitar signal isn't often a textbook symmetrical square and the diode feeding pin14 stops it seeing the negative half of the waveform, so the pitch should wobble around as you play. When the feedback pot is minimum resistance, the VCO should be trying to follow the waveform voltage out of pin2 and sound like metallic ring modulation.
If it sounds like that -  it's working! See what happens if you remove the cap on pin9.

[rutabaga bob]

Pin 5 is grounded.  Will work on disconnecting pin 9's cap and see what happens.

I guess my question has really been: is the circuit/perf as shown viable?  Is what's diagrammed going to do what the demo presents?

[anotherjim]

I think it's probably right as per Moid. The difficulty verifying it is you need to know what it's supposed to do first!