circuitbenders uk harmonic engine (E&MM harmony generator) malfunction

[pinkjimiphoton]

so a while ago, i stumbled upon this:
https://www.circuitbenders.co.uk/forsale/harmonic/harmonic.html
i'm a glutton for punishment, so i bought one for shits n grins
here's the schematic


so i ordered it and all the parts, took a while to track down the 4526 chips, but i got a few of them.

i expected it not to work when i fired it up, but i was wrong. it KINDA works.

it passes signal. it definitely latches on to whatever note i input.. you can hear the synthesized signal trigger on and off, and i added an led to show when its latched.
the rotary switch appears to work and shoots it thru several octaves,

but....

all it does is produce a "g" note. it doesn't track any other notes input, the output is just a g. the switch for 3rd/none/5th doesn't work either, tho you can hear it change something.

a quick voltage check last nite showed expected voltages just about everywhere, but 2 pins that shouldn't be connected on one of the chips are showing a similar voltage... i checked for shorts between ic pins, but found nothing at all.

i will get the voltages today... it was 3 am and i was too lazy to write stuff down then.

voltages also are right on the 3 way switch. i traced the switch to one of the 4526's, U7, and the switch is working properly as is the voltage to the chip itself, i get 9v to pin 5 or 11 or no voltage at all in the middle as expected.

i tried fuzzing it up with a fuzzface and a big muff, tried rolling off all the highs to feed it as little harmonic as possible, no dice. every note i play, it reacts with a g. it won't track anything at all.

being a 10-toed freak with opposable thumbs, a fuzz-monkey with a breadboard, i'm kinda at a loss.... i don't understand 90% of it anyways, but i can follow directions and solder neatly.  so.... where would i start looking with this circuit?

its passing signal, its latching on to the input, its producing a tone, the mix and sensisitivity pots as well as the octave select switch are all working...

i DO have multiples of most of the chips, but don't wanna pull them and replace blindly.

i DID notice that it latches on better with a 9v wall wart. i tried it originally with a battery, and it worked, but not as well.

i built it without the 9v reg, as i figured it has onboard polarity protection and filtering.

so... any ideas? gonna be a little bit before i get down to the dungeon to work on it and get the voltage readings.

oh yeah, and fwiw, i managed just barely to cram the whole mess into a 1590bb took a little bit of creativity and i had to whack off a couple corners of the circuit board to get around the lumbergs, but... if i can get it working, will be a lot of fun i think.

anyways... where to start looking, and how?

i DO have a signal gen now that i built, i DO have a small o-scope too if i need it, an audio probe, etc etc... but i'm not losing audio. it seems to me like the pll is oscillating, but not being driven by whatever it is that makes it track the pitch of the incoming signal.

appreciate the advice!!

check in later... thanks guys!!!

[anotherjim]

A lot depends on the envelope detector IC2. One way, it controls the level of the generated harmony out of the switch chip IC9b pin 11.

The other thing, and maybe most relevant it does, is the digital control signals out of IC4 b & c control the pitch tracking. IC5 is the clever bit.

IC5 compares the Signal in (pin14) with the counter outputs back into pin3. CIN = comparator input! This isn't an analogue voltage comparator, it compares the timing of the square wave edges between Sigin and Cin or the relative phase of one versus the other. Any difference in those edges and it briefly switches pin13 PC2 either high or low. If there is no difference, pin13 is disconnected because it's happy.
That control out of PC2 controls the VCO frequency. It has to set a DC voltage to pin9 that controls the VCO frequency. You can think of PC2 as a "nudge" button. If the comparator thinks the VCO is too slow compared to the signal, it will give pin13 a positive nudge to speed it up. If it goes too fast the nudge is negative to slow the VCO down.
Those nudges are too on/off and jittery, so an RC filter is used to smooth the changes PC2 has to make. The RC filter here is in 2 parts. R16/C11 and R17/C12.

To catch and track the note of the guitar, the 3 switches in IC9a need to be working. Those switches should let PC2 control the voltage on the RC filters and get to the VCO control when the envelope picks up the note. When the note starts to fall off, the switches let the RC filters remember and supply the VCO control voltage while the note fades too low to track anymore.

So, you need to make sure the "C" inputs of the 4016 switches are being switched 9v or 0v when you play into it. If the VCO in IC5 doesn't change pitch it probably means it's at its minimum pitch (set by the resistor on pin12) and not getting nudges through from pin13.

[pinkjimiphoton]

thanks jim!
i am gonna re-read that 20-30 times and print it out for when its on my bench.

i THINK i f'd up and wired the rotary switch backwards.. the build doc is a little weird. i'm getting unison, octave up and 2 octave up oscillations, the other switch postions do pretty much nothing. so i'm rewiring it completely. i used an open wafer kind of 1p7t rotary switch, and pretty sure after re-reading the documentation, tho i followed their diagram, pretty sure i did so backwards. hopefully when i re-connect the whole mess, it will fire up.
i used ribbon cable. its great if ya never move it. move it once? ok. twice? shaky ground, 3 times? you asked for it. so going with individual wires to each connection this time, as it fell apart trying to open it up a little bit to get some pics.

first, voltages
power supply is a wall wart, 9.7v

from the cathode of the 4001 on, b+ = 9.06v
i did not use a regulator, leaving it jumpered to run at 9v

q1, bc557
e  9.05
b  8.49
c   0

u1, LF351
1  -4.80
2  0
3  0
4  -4.80
5  -4.80
6    0
7  9.06
8  0

u2, RC741t (metal can)
1  -4.80
2  0
3   0
4  -4.80
5  -4.80
6  .26
7   9.06
8  0.0

u3, RC741t
1  -4.80
2  0
3  0
4   -4.80
5  -4.80
6  0
7  9.06
8  0

u4, CD4093BE
1  3.07
2  3.07
3  9.06
4  9.06
5  .04
6  .04
7  0
8  9.05
9  9.05
10  0
11  0
12  9.05
13  9.06
14  9.06

u5, CD4046BE
1  0
2  .5
3  .5
4  4.62
5  0
6  1.50
7  1.42
8  0
9  0
10  0
11  0
12  6.09
13  0
14  0
15  0
16  9.06

U6, MC14526BCP
1   4.48
2   0
3   .05
4   0
5   9.06
6   4.62
7   4.63
8   0
9   4.50
10   0
11   0
12   .05
13   1.67
14   0
15  4.48
16  9.06

U7, MC14526BCP
1   0 
2   0
3   .05
4   0
5   9.06
6   4.48
7   3.81
8   0
9   3.57
10   0
11   0
12   1.67
13   2.11
14   0
15   4.54
16   9.06 i missed a pin writing stuff down, so these numbers are funky cuz the switches are disconnected. i believe the voltages were:
13   0
14   9.06
15  2.01
16  9.06
but one of them numbers may be wrong
when i get back from running errands and hook the switches back up, i'll re-take the voltages and compare.

U8, CD4040BE
1   4.52
2   4.52
3   4.51
4   4.51
5   4.49
6   4.49
7  4.11
8   0
9   4.51
10  4.61
11  0
12   4.52
13   4.52
14   4.52
15   4.52
16   9.06

U9, CD4016BE
1   0
2   0
3   0
4   0
5   0
6   9.06
7   0
8   0
9   0
10  .01
11  0
12  4.52
13  0
14  9.06

anyways, mostly, it seems fairly good to my limited understanding. i do believe i wired the rotary switch backwards from how it was supposed to be, but that should still work, i'd imagine either way... one common , 7 poles, correct?

here's pics too

[pinkjimiphoton]

ok,
i got it kinda working. forget a "mix" of dry and effect. it just wants to oscillate. i tried a plethora of different chips tonite of various persuasions and tried at least 2 other chips in each socket location, if no change, replaced with the originals.  no chips ended up being changed, altho u6 and 7 may have gotten swapped.

full wet, it tracks single notes and parallel 4ths fairly well, sounds kinda like an analog synth. i remember zappa saying he had a harmonizer that would track runs and bends, but sounded kinda like a farfisa organ. i'm guessing this is close.

seems the 7 pole switch is the culprit, i have to address the wiring tomorrow and make it much much shorter than it is at the moment.

i can't seem to get it useable at all if i try and blend the dry signal in, but wet, its pretty cool, and tracks most runs and bends really well. only at certain frequencies can i get the 5th to come out, and for that i have to overdrive the guitar fairly massively and boost the mids and cut the bass and treble to get it to really work best. in some places, as the note decays, then the overtones will swell in... or undertones. i've gotta figure it out.

once i solve the oscillation problem, i'll try n bring on the stupid for it in a major way. i'm betting i'm gonna have to shield the common and only connect the shield to the switch, so when its tightened down it will shield the wire to the chassis.

think i found a couple cool hacks along the way, too. gotta see once i get the goddamn tinitis part of it sussed. if i yank on the wires to the rotary just right, it stops, so its gonna be something stupid.

[anotherjim]

Maybe it's only leakage you are hearing anyway. That Octave switch has a lot of digital signal on its wires. The Harmony generator part never shuts up.
The "VCA", if you can call it that, is IC9b. If you have 0v on pin10, it should be silent out of pin11. The harmony signal on pin12 should probably always be there and a DMM could read that as about half of 9v. Fast logic chip square waves tend to average in the DMM to about half the chip supply.

I would be tempted to simplify for now. Replace the Octave switch with a wire link for one selection, say between switch pin 1 and 8.

When you play in, the voltage at IC9 pin10 should go above 0v. How high it goes determines how loud the harmony will be out of the pin12. You have to get this envelope control thing going, or you will never get it to shut up when you stop playing.