can you help me with my 566 VCO?

[rjd2]

hi folks, i just finished building my 1st of 3 of these, and ive got some issues:

1-im only getting audible output at the square and ramp outputs. the triangle output is VERY faint, but exists, and the pulse and sine are completely inaudible. i know i should be trimming on a scope(which i will be once i know im getting outputs properly), but i tried to sweep through the triangle offset trim pot, and i got no more or less of an audible signal.

2-square and ramp outputs, at first glance, sound identical.

ive triple checked my wiring, and it is all per the build notes(the only thing i changed was that i group all my jack grounds and run them to a ground pad at the power input harness pins).

any ideas? thanks alot for the help, cant wait to get all 3 up and set em to stun.....

[davebr]

Sine and Pulse are derived from the triangle waveform so it makes sense with no triangle the other outputs are faulty.

Ramp is also derived from the triangle. There is a "flipper" circuit that inverts the waveform every half cycle. Without analysis, it's conceivable that with no triangle input, the ramp circuit might be generating a square wave output by simply "flipping" a DC level so they would sound the same.

You need to get the triangle waveform working. This is a simple level shift and DC offset all in one op amp.

Check all the part values around U1A: R18, 19, 20, 21, and 22.

Check / resolder all the parts around U1A

If you have a voltmeter, you can check the output of U1A while you adjust the triangle offset. It should change the DC value.

You can listen to the triangle waveform output from the VCO. Connect pin 4 of the 566 through a capacitor (to block the DC offset) to you amplifier. I think the waveform should be about 3.3 volts pk-pk so it will be lower in volume. This will verify the 566 is working and isolate the problem to U1A.

For the pulse, did you remember to jumper TRI to PWM IN? The pulse comparator needs this jumper for an input.

Give this a try and report back.

Dave

[rjd2]

thanks for the help, dave. i made some progress; 2 steps forward, 1 step back....

your tip about tying pwm in to tri did it. i've now got signal from all outputs, i just need to calibrate.

i wired up 566 VCO #2 to compare to VCO #1, and i am getting no output at any of the shapes. i havent listened to pin 4 of the 566, but i compared some voltages to my working unit, and noticed some slight differences:

non-working unit's pin 4 of 566 shows around -15. working was at -9 or so. a few other places i noticed higher voltages than the working unit.

guess i need to listen to the output of pin 4 maybe?

(one other question: on the non-working unit, some thermal compound dripped down onto the legs of the transistors. this shouldnt be a problem, right?)

[blue hell]

[davebr]

There should be a negative voltage on pin 4 as the waveform is centered around some negative voltage (not sure what). However, it shouldn't be -15 volts. That would mean it's clamped at the negative rail.

If you have sockets, pull the 566 and see what the voltage is. It should read 0 volts since it is referenced to the virtual ground at pin 2 of U1A. If it's not, then there is your clue. You could also measure the DC voltage at pin 2 of U1A. It should be 0 volts.

Since you have multiple VCOs can you swap the 566? That would tell you if there is a problem with the part. I wouldn't swap it until I confirmed there was 0 volts on pin 4 of the 566 socket (with the part removed). If you have -15 volts because of a solder bridge then that might damage the 566.

Other thoughts? Look for bad solder joints or a solder bridge. Try swapping U1.

[rjd2]

thanks alot for the help, guys. got a little more info:

using my crystal earpiece for diagnosis, im not getting any audio at pin 4 of the 566 of good unit #1, but i AM getting audio at pin 1 of U1. pin 1 of U1 on the bad unit #2 gives me nothing. the 566's are can-types, so they arent socketed, but i can desolder and swap them(think i have a few leftover). i'll clean my transistor pair while im at it, swap out U1, check pin 4 while its out, and report back in a day or two. have at least one more question, but i'll wait til im out of the weeds on these issues. thanks guys.

[davebr]

It's getting a little confusing between your multiple boards ...

I wouldn't unsolder the 566 yet. Instead I'd lift one end of R18. Much easier. Then check the voltage on pin 4. If it's at -15 volts, I'd guess you either have a solder bridge or your 566 is bad.

With this resistor lifted, U1A simply becomes a voltage source based on the setting of R22. Put a voltmeter on the output and see if you can adjust the ouput from - something close to the rail (-14.5 or so) to + the same amount. If so, then U1A is functioning correctly. There is only 1 resistor between this and pin 4 of the 566.

That puts it back to a bad 566 or a solder bridge.

BTW, adjusting the output of U1A via R22 will operate the pulse circuit. You should be able to toggle the pulse output depending on the setting of R22 and the width control.

Dave

[rjd2]

[davebr]

That could simply be because pin 4 of the 566 is at -15 volts. With it at that extreme, the offset trimmer should bring the ouput of U1A back to 0 when it is at one of the extreme settings (the +15 side). The gain is the same for the voltage through R18 or R21. With the offset trimmer more in it's normal setting you could probably see the +4 volts (more negative voltage into an inverting amp results in more positive voltage output).

Dave

[rjd2]

lifting R18, i still have -15v at pin 4 of the 566. later on i will be tracing my soldering, and probably swapping a new 566 in there.



EDIT: SUCCESS! i've got it up and running. a new 566 and a flux wash fixed whatever it was. now just need to trim. thanks guys!

[rjd2]

ok, hopefully my last question on this thing....

one of these VCO's i've built will not scale properly for more than 3 octaves. its wierd. the best i can get is 3 solid octaves within a few cents of 0, and then, right above that 3rd octave, its like falling off cliff. it goes about 25 cents flat. ive never seen anything like this. the other VCO i built doesnt seem to exhibit this behavior.

any ideas? thanks again for all your knowledge....

[davebr]

I'm always suspicious of excess leakage. Could be the capacitor or could be contamination on the PCB. What kind of flux? Is the PCB clean?

Just a guess but I always make sure my PCBs are very clean. I use water-washable flux and touch up repairs typically with acetone. At least it eliminates a variable.

Dave

[rjd2]

[davebr]

Yes, the timing capacitor. Do you have both capacitors (e.g. the 4.7 uF and the 0.047 uF)? If so, is the short range an issue with both capacitors? If it's only an issue with the 4.7 uF I would suspect excess leakage, especially if it's an electrolytic. Recently I had an electrolytic of that same value not function in a 555 circuit because of leakage. I used polyester film capacitors in mine. I would swap the capacitor with a non-electrolytic and see how it performs. That would be a simple test.

Dave

[rjd2]

dave, thanks alot for your detailed diagnosis skills!!! part of what makes this site so damn helpful.

i have polypropylene caps at C1; looks just like your pic . i now have all 3 TH VCO's built out, and they all exhibit the exact same thing. the best i can get the tuning is like this:

C1-C2=slightly sharp, only a few cents.
C2-C5=bang on.
C5-C6=about 20 cents flat by the time i reach C6.

so ive got basically 4 octaves of useable tuneability. which will probably be fine, but it would be even better to have that last octave. if nothing can be done, so be it. i've cleaned my pcb's, checked all my soldering.

thanks for the help guys!

[davebr]

I didn't get the range out of mine that Thomas claims is possible. I got about 4.5 octaves of usable tuning. Maybe others can contribute what range they got.

The notes from my web site indicate mine was stable, symmetrical, and tracked well over ~4 volt range. At the low end the triangle and square waves skewed from a 50% duty cycle and the triangle linearity degraded but I think the tuning was on. At the high end it began to track flat.

It sounds like you got a little bit less of a range. I have no more insight as I never investigated this and I probably wouldn't have noticed a few cents on the low end.

There may be a possibility of differences between brands of 566's. I have a National which was supposed to be less performance than the Signetics. I don't know the details or if this is really a factor or not.

Dave

[wmonk]

I finished populating the pcb and run into the same issues as noted in the opening post.

The Square, sub squares and saw output all give a square like sound. The triangle, pulse and sine produce nothing. The noise is working correctly.

With the scope I can see a DC value at the triangle output. No wave or whatsoever. I got the LM566 from Magic Smoke recently.

I see mentioning of some shorting to get the pulse working. What should I short?

But first the triangle. Any ideas?
I did check the resistor values and those are correct.

[davebr]

[wmonk]

[davebr]

[davebr]

[wmonk]

[davebr]

[Tim Servo]

Hey Wmonk,

Hopefully Dave can help you get your board straightened out, but if you have a TH-101 board and you need the build docs, let me know (send me a note at magsmoke@gmail.com). The build docs include build / calibration notes, connector wiring guides, parts list and other goodies.

Tim (calibratin') Servo