Yusynth EMS diode filter

[dummyplugconspiracy]

Hello! Today I finished this module and attempted to calibrate it. The problem I'm having is that the emphasis/resonance knob doesn't seem to work correctly. The potentiometer is good and I can prove that with a meter. I can even hear the knob working properly when I cycle through 0 to 7 V with my V/OCT generator connected to the emphasis control input. But when I'm just trying to hear it work with no CV it does nothing at all. My guess is that it is getting a constant CV even when I'm introducing no CV control purposely and that's overriding the changes I attempt to make with the emphasis/resonance knob. My wiring looks fine and I am checking my component placement now, but I already checked that before I plugged anything in. I did make note of the four erratum items on this board and I believe I did everything correctly. Everything else seems to be working great. Any advice for me?

[dummyplugconspiracy]

I did some more testing. All points I tested exhibited no obvious crazy wrong voltage. My -9 and +12 seem to read right. My grounds are grounded. My pot is a 10 Lin as it should be. No shorts around the wiring. My resonance knob is making voltage change at R41, R39, the base of Q12, even at pin 4 of U4. When I apply my cv test voltages to the resonance mod I get 1v - 7v as I should at pin 7 of U1b and all the way back to the pot. As before, with cv the resonance works fine. It's just the resonance knob that doesn't work. I thought my problem may be the 680k at R41 as documented. I switched to 220k and it had the expected impact of sending it straight into self oscillation, but the resonance knob still didn't affect it. And since it was way too much oscillation, I went back to 680k. I'm stumped. Since the filter technically works, this has to be something relatively simple. Help!

[Paul Cooper]

When I apply my cv test voltages to the resonance mod I get 1v - 7v as I should at pin 7 of U1b

I assume you mean -1 to -7V?

My resonance knob is making voltage change at R41, R39, the base of Q12, even at pin 4 of U4

Yes but what voltages are you seeing and how does this compare to the voltages measured at the same points when just using the resonance CV input?

[dummyplugconspiracy]

I won’t be back at the bench until this weekend, but I will get you that information. I’m not sure whether I mentioned it explicitly in the other emails, but when I apply resonance mod CV the resonance knob does have an affect on the range. That seems to prove that the knob works, the wiring goes to the right place, and it is capable of controlling resonance, but it’s not controlling it unless I am Introducing resonance mod CV. Without CV input the knob has no effect. Every other knob and jack appears to be working properly. Thanks!

[dummyplugconspiracy]

I had a chance to take some readings today comparing voltage with and without CV input. But to respond to your first question . . . YES I meant to say CV was cycling from 0 to -7v at pin 7 of U1b. Sorry about that.

I did this by testing with no CV and then with 4v CV first sweeping the Emphasis knob from lowest to highest each time to compare readings. I tried to test every point that was related to the CV input or the Emphasis knob. Let me know if I missed anything or got the labeling wrong:

_____________ No CV _________ 4v CV
R43_______ .12 - 0.14 ______ 4 - 4
U1b pin 5 ____0 - 0 _________ 0 - 0
U1b pin 6 ___ .08 - .08 _______ 0.1 - 0.07
U1b pin 7_____0.2 - 0.2_______3.9 - 3.9
R44________ -0.2 - -0.2 ______ 3.84 - 3.84
R45_______ -1.2 - -1.2 ______ -1.6 - -2.8
C13_______ -1.2 - -1.2______ -2.85 - -2.85
R41_________ 0 - -9.0_______ 0 - -9.0
R39_________ 0 - -1.2 _______ -1.6 - -2.8
Q12 base____ 0 - -1.2_______ -1.6 - -2.8
C14 neg____1.55 - 1.6________0.4 - 0.2
C14 pos____skipped this________1.6 - 1.7
R42_______1.54 - 1.58_______ 1.6 - 1.7
R40_______1.54 - 1.6________ 1.6 - 1.74
U4 pin 4___1.55 - 1.59_______1.63 - 1.73
R42_______1.54 - 1.58_______1.6 - 1.7
R33_______10.47 - 10.47______10.49 - 10.5

The only other thing I noted today was that there were a couple pins on the 3046 chip that didn't seem to be behaving as the schematic would suggest. Pin 8 was reading 9v, but it appears to go straight to the 12V rail. Pin 12 was reading 0.6v, but it appears to be going straight to ground. Not sure that means anything, but I figured I should mention it.

What's next? Was this constructive at all?

THANKS!

[Paul Cooper]

Pin 8 was reading 9v, but it appears to go straight to the 12V rail. Pin 12 was reading 0.6v, but it appears to be going straight to ground. Not sure that means anything, but I figured I should mention it.

Before we look at the results of your testing, your supply voltages need to be right first. Pin 12 of the 3046 does go to 0V so seeing 0.6V there needs investigating. You are reporting only 9V from your 12V regulator equally needs looking into.

Checking power supply voltages and the current drawn by each rail should be done as a matter of course. Is U2 (78L12) getting hot? Note the schematic shows the rail currents as +30mA / -29mA, what are you seeing?

[dummyplugconspiracy]

False alarm on the 3046 pins. I must have turned the board as i was working and counted the pins off wrong. Pin 8 is almost exactly 12v and pin 12 is 0v ground. Pulling roughly the right current and the 78L12 isn’t hot. All is well. So I guess we can proceed to checking those other readings.
Thanks!

[dummyplugconspiracy]

I just noticed a possible discrepancy in my wiring and I wanted your advice before I experiment with it further. In Yves design, the three CV connections do not seem to be exactly the same. CV 3 is behind a 100K resistor at R15, CV 2 is behind a 220K resistor at R16, and CV 1 is behind a 100K resistor at R17. Since the soundtronics version of the kit includes the extra pot to accommodate two of the CV inputs, logically I would expect CV 3 and CV 1 to use the pots since CV 2 to has the different resistance. That makes me believe that CV 2 should not use a pot and should go directly to the board. That is not how I wired it, but it would be easy to Change if my assumption is correct. Right now I have CV 2 and CV 1 going to a pot and CV 3 going direct.
I don’t see how that would cause the problem I am having with the residence knob, but it seemed like it was worth exploring. I am having my resonance knob problem even when no control voltage is being input at those connectors, so it seems unlikely it could have any effect. What do you think?

[dummyplugconspiracy]

I assume Paul is either very busy or on well-deserved holiday right now, but just another update. I disconnected, reconnected, trimmed, examined, and traced. I’ve been over this module a million times and it’s built right. Hopefully those voltages I shared will reveal a clue. I’m starting to wonder if I have a bad 3046 chip. I understand they do have a fail rate and this one is reclaimed. But this project is dead in the water for the moment. I’ve reached the limits of my troubleshooting skills. Any advice is appreciated. Thanks!

[Paul Cooper]

As a matter of interest and with no signal input, at what point does the filter self oscillate when using just the emphasis control? At this point, measure the voltage at the wiper of P2.

Now repeat with the emphasis control at 0, apply a cv to the resonance input with its modulation knob at 100%. At what voltage does the filter self oscillate?

[dummyplugconspiracy]

I am away from my bench until Wednesday night, but I will definitely test this and report back to you. Thanks!

[dummyplugconspiracy]

I set both signal input knobs at zero, all CV modulator is at zero, set emphasis knob at 10, and swept the frequency cut off knob up from zero to test this. Nothing happened at all until the frequency cut off knob got to 8 at which point I just heard a medium pitched electrical hum that showed up abruptly. It did not fade in. As soon as I hit 8, I get the hum and increasing frequency cut off from 8 to 10 doesn’t change it.

Then I turned emphasis to zero, turned frequency cut off to zero, and turned the resonance modulation all the way up and plugged my 1 v/oct CV reference into the modulation input. I get no audible output at all at 1-7 volts. I tested with cutoff frequency at 5 and I can hear high pitched self oscillation only at 7v. With cutoff at 0 I get low humming oscillation at 6v and 7v. Much louder at 7v than at 6v.

Through it all, I still cannot make it self oscillate at all using the knobs without resonance cv input.

Are these useful clues?
Thanks!

[Paul Cooper]

Okay, yes useful. I don't have one of these to compare with so a little guesswork here.

As before, no resonance modulation CV or any input signals. Focus first on getting self-oscillation using just the emphasis pot.

This is where I would clip a resistance decade box across R41 and find the value that given self-oscillation at around 8-9 on the emphasis pot. Assuming you don't have one of these and to save keep removing R41 and possibly damaging the PCB, I suggest temporarily soldering various test resistors across R41.
Start with a 1M, then try lower values until self-oscillation, you shouldn't need to go below 220k. You can then replace R41 and the test resistor with the nearest preferred value based the their equivalent parallel resistance.

[dummyplugconspiracy]

ok got it. But the test resistors go on in parallel . . not series. So I can attach the test resistor to the exposed legs of the R41 resistor and stack it on top until I get results, right?
THANKS

[dummyplugconspiracy]

Ok - tried this and 220k seems to be doing the trick. Very similar behavior to 330k but 220k seemed to have wider range. So I have that alligator clipped on.

When I introduce signal and try to use the module like normal, I'm not getting a smooth filter sweep with the cutoff knob. The emphasis knob still has no effect until it self oscillates at 8. Cutoff just seems to distort slightly when it goes above 1 and doesn't change much after that. You probably anticipated this. What's next?

Thanks

[Paul Cooper]

I may not be able to help much more with this module but here is some thoughts on the basis that you have populated the PCB correctly:

Now that you have changed R41 to 220k, measure the voltage at the gate of Q12 at fully CCW and at fully CW of the emphasis pot (no CV or signal).

Try the frequency response again with a saw or pulse input with the input signal attenuated to level 3 on the input pot.

Try replacing Q12

[dummyplugconspiracy]

OK… The gate is the middle leg of that Q12 transistor. When I test that I get 0 V when the emphasis knob is all the way counterclockwise at 0 and 3.6 V when the knob is all the way clockwise at 10. Does that reveal anything?

When I apply a 400 Hz Square wave and turn the signal input to 3, I am still not getting the right effect. When I turn the frequency cut off knob, I get some frequency response until I get to 3, but from 3 to 10 it’s just somewhat overdriven and no additional filter change takes place. The resonance knob still has no affect unless I turn it all the way to 8 and introduce self oscillation.

I will have to order some more of those transistors and see if replacing Q12 makes a difference. any other ideas you have in the meantime, would be appreciated. Thanks!

[dummyplugconspiracy]

Im attaching a picture of the Pcb. I am confident I populated it right, and I know you can’t tell much from this picture, but if you see anything that looks out of place, please let me know. I’ve checked at about 12 times. Thanks

[Paul Cooper]

When I test that I get 0 V when the emphasis knob is all the way counterclockwise at 0 and 3.6 V when the knob is all the way clockwise at 10. Does that reveal anything?

It does as the voltage should never be positive, only negative. Check your wiring to the emphasis pot as the CW terminal on the pot should be connected to -9V.

[dummyplugconspiracy]

Argh. Sorry -3.6v. Does that help?

[Paul Cooper]

That sounds right. I am now at a loss to understand why it is not working. Having not built one myself, I cannot do any bench tests. I am not aware of any issues from other builders so can only assume the design is fine so excluding any construction errors, that only leaves components.

The 3046 in your kit is probably not reclaimed but a new version called the AS3046 - check the markings to see if this is the case. It may be worth trying a different 2N3819 FET as the design may be susceptible to differing characteristics. The design is based on obsolete components which is the challenge with building clones of retro products dating back decades. We no longer have a supply for the original 3046 and not many kits have been sold with the AS version so perhaps that is an issue but this is only a guess.

[collinspr]

Looking at your pictures, you appear to have a number of wire links unpopulated on the PCB. Looking at Yves website, I’m pretty sure these are required,

Paul

[dummyplugconspiracy]

Oohhh ok. I thought those links weren’t needed on your PCBs so I’ve gotten into the habit of ignoring them. I’ll add them when I get a chance. Thanks!

[collinspr]

Not ‘my PCBs’ - different Paul!

[Paul Cooper]

The PCBs are double sided and those links already have traces there so links not needed. The links for the home made single sided PCBs only.