Harmonic Trem Help

[Christoba]

Hello,

I have a Cherry Pie Tap Tempo from Delyk based on the Twin Peaks Tap Tempo Harmonic Tremolo.  The system worked for about a year (minimal use)without issue but now has a problem. If I let the pedal sit overnight and turn it on, it will work fine for about 4 minutes.  Then, the tremolo will no longer function (no oscillation).  I get sound, tone knob still works.  However, the vactrols stop oscillating and the "rate indicator LED" goes to a constant light glow. I've checked all connections and joints. Below are the numbers:
Link to manufacturer Project
https://www.delykpcb.com/product/cherry-pie-tremolo-pcb/

Schematic


IC1 (TL074)
1 - 4.65
2 - 4.65
3 - 4.28
4 - 9.38
5 - 4.67
6 - 4.68
7 - 4.68
8- 4.68
9 - 4.68
10 - 4.68
11- 0
12 - 4.68
13 - 4.68
14 - 4.68

IC2 (NE5532)
1 - 4.07
2 - 3.4
3 - 3.4
4 - 0
5 - 3.4
6 - 3.4
7 - 2.7
8 - 9.3

UC2 (Tapflo)
1 - 5
2 - 2.2
3 - 2.3
4 - 5
5 - 4.98
6 - 0
7 - 0
8 - 5
9 - 2.62
10 - 5
11 - 2.2
12 - 3
13 - 2.45
14 - 0

U1 (78L05)
1 - 5
2 - 0
3 - 9

U2 (78L05)  [Seems strange for pin 2 to have voltage)
1 - 6.8
2 - 1.8
3 - 9.2

Thank you for your time and I will take any and all suggestions.

Thanks,
Chris

[Mark Hammer]

That sort of behaviour usually is associated with a bad electrolytic cap.

[bluebunny]

U2 (78L05)  [Seems strange for pin 2 to have voltage)

That's because the GND pin is elevated by the LED V_F (1.8V in this case), thereby also elevating the output to 6.8V instead of 5V.

[ElectricDruid]

Rate indicator LED (D1, "Tempo LED", right?) going to constantly on suggests that the TapLFO isn't happy. A bad +5V supply is the first place to start, but it doesn't like having input pins left open either, so a pot gone bad or a connection on one fo the pot inputs going open could cause this too.

Finally, this circuit is set up for the original TapLFO 2D, not the more recent TapLFO3. That's the chip you've got, right?
If not, no massive problem - it can be adapted easily enough. There's just a few parts that aren't required any more.

HTH,
Tom

[Christoba]

Tom,

Thank you for the info.

-Yes, Correct D1

-I read a solid 5v at the power supply to the chip.

-The markings are PIC16F684, so 2D?

-Sounds like you are telling me to go through the offboard components to check for anything strange. Will do.

[Christoba]

Rate indicator LED (D1, "Tempo LED", right?) going to constantly on suggests that the TapLFO isn't happy. A bad +5V supply is the first place to start, but it doesn't like having input pins left open either, so a pot gone bad or a connection on one fo the pot inputs going open could cause this too.

I went through everything and the problem persists.  Also, D1 isn't full on.  It's at about 30 percent. 

Thank you blue bunny on the note.  I should have figured that out. Doh!

[ElectricDruid]

I went through everything and the problem persists.  Also, D1 isn't full on.  It's at about 30 percent. 

Ok, 30% on is quite different. That sounds like the PWM output is working, but you'd need an oscilloscope to check for sure. And if the PWM is working then the PIC is running. And yes, 16F684 is TapLFO 2D, so that's good.

BTW, does anything happen when you move the pots? Any change on the LED?

Are the Vb and Vd voltages correct? Mark is right that "works for a little while and then stops working" is typically something you'd associate with a bad electrolytic cap. That suggests the power section - hence the question about the Vb and Vd voltages.

[Christoba]

I pulled everything apart again, replaced all electrolytic caps and and verified values. Same problem.

Vb=2.5
Vd=3.3

I noticed something strange though, when I measure voltage at pin 9 of the taplfo (U2), the led and vactrol begin to oscillate.  This oscillation is unaffected by turning any pots.

I don't have a oscilloscope unfortunately. Thanks again for all of the help so far!

[ElectricDruid]

Could it be a simple bad joint on Pin 9? Then when you press it with a probe, you make the connection and it starts working?

The PIC doesn't like to have pins left open, and all sorts of weirdness ensues if you try it, so a bad connection there would explain what you're seeing. Although it's not the only possibly explanation by any means.

[Mark Hammer]

That's what I get for trying to look at the schem with a small tablet screen.  Didn't even realize there was a PIC involved. 

[ElectricDruid]

TBH it's not that obvious on the schematic at any size. It doesn't say "PIC 16F684" or anything useful like that on the drawing. I only know because it's the buy-a-board-and-build-one version of David Rolo's Twin Peaks Tremolo, and David and I have exchanged email about PICs and pedals and PICs _in_ pedals and so on. He builds some lovely stuff. I'm a fan, I admit.

[pinkjimiphoton]

sounds to me like a bad electro as mark suggested, or a bad solder joint. try checking voltages at every junction.. check all the resistors first, then check the caps. any place you find a dc voltage on the - side of an electro, suspect it. it may be bad or may not, it depends if its used for coupling, decoupling. a coupling cap may have voltages on each side, but usually one side will be a lot higher than the other. a decoupling cap to ground should never show any dc on both sides. often, its a bad solder joint.

another clue to it being an electro, is to a certain extent, electros are "self healing" in some cases, which can be maddening.

how long does it take til it messes up? if it takes a long time, then messes up, then comes back on, its almost always an electro.

on a side note, have ya'll see this?

https://zeppelindesignlabs.com/product/quaverato-harmonic-tremolo-pedal/

that is really the one harmonic trem to rule them all. i bought the kit on ebay for like, 99 bux, and it's absolutely insane.

out of the harmonic trems i built, starting with jon patton's first one the cardinal, this is the one that was the keeper.

[ElectricDruid]

Yeah, I had seen that Zeppelin Labs trem. It looks very nice. Someone had a problem with one that they were asking about here not that long ago. Before then I'd never heard of it. It's very similar in concept to the Twin Peaks. The LFO even has the same multipliers and the same wave distortion function, as far as I can tell. It's based on an AVR not a PIC though, so the guts are a little different technically if not in function. The Twin Peaks has more different waveforms, although whether or not you really need more choices is an open question. I like random stuff though, so I'm always in favour of a nice random waveform. Adding a MIDI implementation is a nice touch too - That's pretty cool on a DIY pedal.

[pinkjimiphoton]

i didn't add the midi board to mine, but it seems cool if its useful..as a guitar player, not so much. it's only real downfall i've found is i'd like it to be just an rch more intense. but it really nails the fender well...and most others i've tried.
it'll be replacing my TL (barry's guitarpcb version) next time i redo my board.

i only endorse it being cool , for MY needs, others mileage may vary. really good detailed build doc*, even a beginner could probably nail it first time!





*almost overkill i should have read the whole doc first, lol

[Sonic_Fields]

Hi guys,

Although I frequently come across this wonderful forum, this is is my first post here...

Working on a Harmonic Trem also right now. Got some inspiration from the Cardinal, the Twin Peaks en the Electric Druid datasheet. Many thanks to everyone who already did some had work on this matter!

Anyways, I was wondering what the function of the LED at the ground pin of the 2nd voltage regulator is. Any ideas on this?

Thanks!

[ElectricDruid]

This is the answer! -

U2 (78L05)  [Seems strange for pin 2 to have voltage)

That's because the GND pin is elevated by the LED V_F (1.8V in this case), thereby also elevating the output to 6.8V instead of 5V.

It's a cheap/alternative way of getting a higher voltage out of the voltage regulators you've already got around. It makes sense to use two 5V regulators, rather than add a 6.8V regulator to the BOM, even if you could find such a thing!

[Sonic_Fields]

Ok, makes sense indeed! Thanks.