YuSynth Quad LFO issues

[diophantine]

Hi Yves, all,

Last weekend I finished building a Quadrature LFO, from a board that I had someone etch for me.

Everything seems OK with the board. I used 1% resistors throughout, so that isn't an issue. I matched the capacitors to 10%. I matched the transistors in the expo converter - it was my first time matching transistors (I used Ian Fritz' method), but these were the two closest. (And even if they weren't matched well, I don't think that could be causing the problem.) The wiring all seems to be correct, for the switch, pots, jacks, and LEDs.

Upon initial power-up (trim-pot @ 50%, CV @ 0, rate @ 10, level @ 10), it worked - all LEDs were lit simultaneously (apparently going too fast at ~100Hz), and the output in the oscilloscope was correct.

The switch worked, and did what it was supposed to do. The output level knob worked too, although the level seemed to be the same above 8.

Moving the frequency knob down, the oscillator response seemed sluggish... eventually, around 6 or 7, the amplitude started decreasing as well (independently), and eventually the output went flat. I tried adjusting the trimpot as well, but I got pretty much the same result... just the frequencies and fail-point would change.

Unfortunately, since then, I haven't even got even that much to work! With the frequency knob at 0, I get only the 270 and 0 degree LEDs lit; at 5 I get only the 0 degree LED lit, and at 10 I get only the 0 and 90 degree LEDs list. No oscillating whatsoever. It seems that the oscillator is stalled.

I've tried checking the voltages at various points. The voltages coming directly from the frequency pot and trimpot are fine (-15V to +15V, and -15V to 0V, respectively). At the input to the expo converter (base of Q1) I get an approximate voltage range of -.3V to +.1V. At pin 8 of U2 (output of U2c opamp, also one side of C6) I get a variation of approx. +3.4V to +6.3V depending on the trimpot & freq pot positions. It appears that there is some slow oscillating... the voltage will still change when no pots are being moved, and usually in a single direction, but I haven't been able to extrapolate any particular pattern.

I've also checked all the continuity surrounding opamp U2c: pin 8 is connected to C6; pin 9 is connected to the other side of C6 and pin 5 of the LM13700; and pin 10 is connected to ground. (Since the signals into, and out of, the LM13700 are currents, I didn't measure those... didn't want to cut traces, at least not yet.)

So, it seems that the oscillator is "stalled", i.e. that the capacitor (C6) isn't discharging. Today I finally decided to try shorting C6 with a small screwdriver to discharge it - well, it worked: the 0 degree LED went out, and the 180 degree LED lit! (On occasion I got the 90 and 270 degree LEDs to swtich too).

Anyhow, sorry to make this so long-winded, but I wanted to make sure I'd explained everything.

If anyone has suggestions regarding what may fix this, or what I should do to test this further, it would be much appreciated!

The only other thing I can think of mentioning, is that I'm using a TL074CN opamp... I don't think that this would be anything different from a TL074, correct?

Thanks!!

[yusynth]

Hello

Have you an oscilloscope ? If so you can start looking to the signal that are obtained at pins 8 and 7 of U2 (TL074).

[diophantine]

Hi Yves,

Yes, I have a scope. I ran it DC-coupled, since the changes were too slow to register when it was AC-coupled. All of these tests were done with the trimpot outputting 0V.

While warmed up (LFO had been on for ~15 minues)...
Pin 8 (Sine oscillator): With rate pot full CW I got a relatively steady +4.2V. I then turned the pot to full CCW quickly. The output quickly rose to +5.4V and then immediately dropped to +2.0V. It then slowly dropped (over a few minutes) to a relatively stable +1.4V.
Pin 7 (Cosine oscillator): With rate pot full CW I got a relatively steady -13.8V. I then turned the pot to full CCW quickly. The output quickly rose to -1.0V, and then (over a few minutes) slowly rose to reach a relatively stable +7.2V.

I also tried measuring the voltages at power-up (the oscillator had been off for ~15 minutes), with the rate pot full CW.
Pin 8 (Sine oscillator): 0V... rose quickly to +5.6V and immediate started to fall (slowly) to +3V.
Pin 7 (Cosine oscillator): 0V... dropped quickly to -8V and then continued to fall (but slowly) to -11.6V.

[yusynth]

With the scope do you see sinewaves at pins 8 and 7 of U2 (RATE pot fully clockwise). If so are these balanced around 0 and what is their peak to peak amplitude ?

[diophantine]

I don't get any sine waves (or anything cyclic) out of any opamps.

When I initially tested it a few weeks ago, I did get sine waves from the module outputs. They were clean looking sine waves... I don't remember exactly, but I think they were ~7Vpp and centered at zero. They were ~10Hz.

But, it stopped oscillating that same day, when I was testing it more. I never tested the U2 outputs until after the thing stopped working.

[yusynth]

Ok then it looks like there is either a dead component or a bad soldering somewhere. Did you change the LM13700 and TL074 already ?

[diophantine]

Alright, I will try to replace those components! (I don't have any extras on hand.) I guess I've been lucky, and never had an IC die on me before.

I'll also go over the traces and soldering joints again. I tested most of them, but perhaps I missed something.

[Cardinen]

Hi,

I've just finished a Quad Lfo but i have the same problems written above in first post.
I get only the 0 and 180 degree LEDs lit.
The oscillator seems "stalled" as you describe, the rate pot have a "fall point" between 6 an 8 where the oscillator out goes flat. Moving the trimmer is the same, only the "fail-point" change.
I've tried different IC (both TL074 and lm13700) and is the same, so i guess the problem is not there.
If i short C7 with a screwdriver i get the the 90 and 270 LEDs on.
I've noticed the frequency of the lfo is higher than what is written on Yusynth site, even with T1 fully CCW.

From Yusynth site :
"set the frequency pot to the minimal rate (fully counter clockwise) then adjust the trimmer T1 in order to set the lower rate (the lowest rate achievable is around 30s per cycle)."

I only get something like 3-4 sec. at minimum rate, so i think is not fine.
Advices ?

Thanks.

[yusynth]

Post detailed photos of your board (component side and track side).

[Cardinen]

Ok

[yusynth]

OK Cardinen

On the photo I have a doubt concerning the trimmer that does not look to be correctly soldered to the pads, check this.

Next the circuit etching is far from perfect, check with a continuity tester that there is no microbreak for all tracks.

[Cardinen]

Hi Yves,

I've re-checked all tracks and they are fine, no microbreak.
The trimmer is fine too.
The problem seem to be around the second OTA (U1b);
I'll continue my experiment and i'll let you know asap.
thanks

[Cardinen]

Ok, i've found the problem around the second OTA, it was a microbreak or a bad solder from one side of capacitor C7.
Now it work as it should.
Thanks Ives !!

[diophantine]

Hi Yves,

I replaced the 3 ICs in my Quad LFO. Now, it oscillates quadrature sine waves! But, it still has some odd behavior going at slower speeds.

Here are my tests. I connected it to my scope and ran it in DC coupled mode, so that it would read fine at low frequencies.

All the waves are sine and ~15Vpp, unless mentioned otherwise.

===

Trimpot: CCW
Freq Pot: CW (10)
Output period: ~0.19ms.

Keeping the trimpot the same, I moved the freq pot down to CCW (0): the frequency decreased as expected, and the point-to-point voltage remained constant.

Trimpot: CCW
Freq Pot: CW (0)
Output period: ~19.0ms

===

Trimpot: 50%
Freq Pot: CW (10)
Output period: ~1.0ms

Keeping the trimpot the same, I moved the freq pot down towards CCW (0): the frequency decreased as expected, but when the freq pot reached 1, the point-to-point voltage started decreasing on its own (without touching the freq pot) until it reached 0Vpp (i.e. no oscillation - nothing on the scope & nothing on the LEDs).

Turning the freq pot back up resumed oscillation.

===

Trimpot: CW
Freq Pot: CW (10)
Output period: ~40.0ms

Keeping the trimpot the same, I moved the freq pot down towards CCW (0). It performed like the experiment above, except that the uncontrollable decrease in peak-to-peak voltage started at around 7 on the freq pot.

===

Also, with trimpot CW and freq pot CW, I tried the level pot. It works, but the voltage is the same from 7-10. It decreases going from 7 to 2.5. At 2.5 the peak-to-peak voltage reaches 0V, and remains the same as the pot is moved to 0.

===

If you have any ideas on these issues, I'd love to hear! Looking forward to get this properly working in my system, alongside my other YuSynth modules.

Thanks!

--
Kyle

[yusynth]

Change Q1 and Q2.

[diophantine]

[bc547c]

hello yves, I've finished a your Quad LFO module; everything is ok but when I turn the Output Level Knob to the maximum position the circuit stop oscillate...I tried different pot but I can't find the solution. Can you help me?

[yusynth]

Are you sure that you wired the level pot correctly ?

[slowroom]

I had a same issue with yours.
Even though all the solutions mentioned above, it still didn't work at all.

However, once I replaced the LM13700 with another from a different seller, it worked well all of sudden. the LM13700 ordered from Aliexpress, didn't work with Yusynth Quad LFO.

I hope it works.