Yet another 3340 tuning issue

DIY STUFF
1

I have read through many other threads (also Advice for troubleshooting 3340 based VCO) but I feel like I am having some weird issue which requires some advice from experts.

I built the CEM 3340 VCO (mine based on AS3340) with all the bells and whistles and also double checked the annotated schematics from Frederik (Annotated Simple 1V/Oct Oscillator Stripboard) but I am unable to find the error.

The circuit works, I get nice triangle/saw/pulse when I check them with the oscilloscope and also the PWM does what it should, but I am unable to tune it, I mean: not at all.

Here are my observations:

  1. The coarse tuning works fine with very low frequencies up to high frequencies but when I turn it up and it reaches at about 90%, the frequency suddenly drops between ~90%-95% and stays the same from 95%-100%, round about.
  2. I play C0, C1, C2 in a loop but whatever I do with the 10k trimmer, I cannot reach an octave difference. It’s more like fifth or so… The whole frequencies shift up and down and the largest gap is between C0 and C1 but not an octave.
  3. If I go to higher CV, the frequency drops suddenly, e.g. C3 appears lower than C0.

I am not sure what I am missing here, so any advice is more than welcome! Obviously the scaling is messed up and there is this weird drop, but it doesn’t make a sense to me.

2

Very strange. Do you have multiple AS3340’s and do they all show the same problem? Do you provide a proper 12V and -12V?

I would especially look into the part of the circuit surrounding the trimpot since I would suspect the error to be there. I have had issues with potentiometers that do not function properly, so a replacement might be an option. I’m often double checking the component values I am using, using a component tester like this one, to make sure the values are as advertised.

I always breadboard it before soldering so I can validate the circuit itself and while soldering I have a working version ready, just in case the soldered version has problems, so I would suggest to do the same.

1 Like
3

Thanks @nschagen I will replace the trimpot and also the AS3340. Should have done it already but no idea where my brain was.

Currently I provide the 12V and +12V from two powersupplies and set them to exact values (three digits after the decimal).

Yes regarding the breadboard hint: I made the mistake that I have not tried to tune it on the breadboard. I was happy to see clean saw/triangle/PWM_squares and thought that the tuning will work nicely once I have everything soldered on the stripboard…

OK, so I’ll check everything around the trimpot then. I also use a component tester btw…

Actually the -7.2V on pin 3 as written on the annotated stripboard from Frederik is different from mine, I measure -7.36V. That’s the Zener diode and connects the trimpot to pin 1 but I thought that 24kOhm + (0Ohm to 10kOhm) gives around 50% of range and it should be negligible. In the AS3340 datasheet it’s a 7.4V Zener Diode, in the CEM3340 it says 6.5V. Nevertheless, I think that even if my 24k resistor is a bit too off (I only had standard one), that should be fine.

But yes, I’ll now replace the trim pot and measure the new one before I mount it. I’ll also replace the AS3340.

Btw. here are three sample measurements of the triangle cycles of different C-tones with different trimpot and coarse tune settings. The first one is the closest but definitely too far off from octaves:

C-1  38.20ms
C 0  19.85ms
C 1  10.42ms
C 2   5.78ms

C-1   9.32ms
C 0   5.07ms
C 1   2.80ms
C 2   3.30ms

C-1  10.00ms
C 0  10.80ms
C 1   6.02ms
C 2   3.63ms
4

Actually the -7.2V on pin 3 as written on the annotated stripboard from Frederik is different from mine, I measure -7.36V. That’s the Zener diode and connects the trimpot to pin 1 but I thought that 24kOhm + (0Ohm to 10kOhm) gives around 50% of range and it should be negligible. In the AS3340 datasheet it’s a 7.4V Zener Diode, in the CEM3340 it says 6.5V. Nevertheless, I think that even if my 24k resistor is a bit too off (I only had standard one), that should be fine.

I think I’m actually using a 22k resistor there and also using a 680 ohm instead of 620 ohm, since I didn’t have those values. It worked well for me so I don’t think thats the problem. By tuning the pot correctly, I think having a slightly bigger/smaller value there doesn’t really matter.

1 Like
5

OK found it! The trimpot was OK but pin 4 and 5 were shorted via a tiny little solder bridge. I found it under the microscope since I got suspicious after checking the resistors with the removed trimpot and AS3340.

Thanks for the input Nathan! Now the octaves are pretty close :smiley: I’ll tune tomorrow, I need some sleep…

2 Likes
6

No worries. I’m pretty paranoid about my solder work and always double-check. I have pretty bad eyesight so I rely on my multimeter to find little mistakes like these using the connectivity tester and it saved my ass a few times, even in cases where I couldn’t really see where it was connecting. I’m doing such checks a few times during my work since I know that catching a simple mistake early can save me hours of frustration, so its worth it :+1:

2 Likes
7

for what its worth saying i had a weird issue where if i socketed the octave switch it wouldnt hold a tune but soldering it directly fixed that. guess its a good thing to check for others