IT WORKS! 4-pole filter from two LM13600 (audio demo+schem.)

beep · Joined: May 05, 2013 Posts: 105 Location: Germany

I recently built the ms20 lowpass. the 100k resistors at the control input need to be changed to 10k to increase the frequency range. It's a 12db filter.

I want 24db, so I just added two more filter stages. Will this work? Very Happy

sffs.gif

Description:

Filesize:

6.89 KB

Viewed:

5870 Time(s)

690_25337.gif

Description:

Filesize:

9.05 KB

Viewed:

354 Time(s)

This image has been reduced to fit the page. Click on it to enlarge.

beep · Joined: May 05, 2013 Posts: 105 Location: Germany

https://soundcloud.com/adolfchristus/dif

i breadboarded it today and it seems to work.

resonance is always a bit lower on breadboard,
when soldered the self oscillation goes insane.

should I change the 220k resistors? I found out, the higher the value,
the more self-oscillation occurs. too much.
or just add a resistor at the res pot in the feedback path?

the original description tells about clippers to
avoid distortion when oscillating. where to place them?
in the feedback loop? just a resistor and two diodes
(one to ground and one from ground) ??

had no 680pf caps, so I took 470p ones.

JingleJoe · Posted: Mon Sep 02, 2013 4:03 pm Post subject:

Just regarding the oscillation prevention clipping diodes; I used some in a filter I was building today in parallell in the feedback path in my circuit, however here's a similar circuit to yours which I found in a quick google for "roland vcf circuit" (because I knew I'd seen that arrangement you spoke of in roland VCFs) in which the diodes are connected in opposite directions from the wiper of the resonance pot to ground. Smile

Cynosure · Posted: Mon Sep 02, 2013 4:16 pm Post subject:

Have a look at the feedback path here:

http://www.schmitzbits.de/rs20.png

The green LEDs cause slight clipping in the feedback path.

In my experiments, I had to adjust the value of the 10k resistor that is parallel to the LEDs. I think I ended using 18k for my circuit, but it was a long time ago and I likely made other changes too.

The Monotron uses 330k for that resistor, but it looks like there are many things that are different with that filter.

beep · Joined: May 05, 2013 Posts: 105 Location: Germany

https://soundcloud.com/adolfchristus/24db-oct-lowpass-vcf

I made a more beautiful demo. it's really cool! but the distortion is annoying.

Thank you, JingleJoe and Cynosure!

@JingleJoe: Shall the diodes and the resistor go to vr or to real ground?

@Cynosure: Will the op amp clipper circuit work when I power the op amp from single supply?

edit:
link corrected...lol...tired... Very Happy

Cynosure · Posted: Mon Sep 02, 2013 4:37 pm Post subject:

beep · Joined: May 05, 2013 Posts: 105 Location: Germany

so the resistor to ground on the diode feedback loop will go to virtual ground, right?

A LM13700 in the feedback loop? voltage controlled resistor for resonance CV?

I think I'm gonna stock a few more OTAs before they become unavailable

Cynosure · Posted: Mon Sep 02, 2013 5:05 pm Post subject:

beep · Joined: May 05, 2013 Posts: 105 Location: Germany

Tried both, the passive diode clipper and the op amp one.
both didn't work. the passive clipper seems to do no more than a resistor
in the feedback loop, there is more headroom until it oscillates,
but the same amount of crappy distortion occurs.

The op amp circuit just boosted the oscillation, also with lots of distortion Sad

Cynosure · Posted: Tue Sep 03, 2013 8:57 am Post subject:

elmegil · Posted: Tue Sep 03, 2013 10:45 am Post subject:

I've also discussed using zeners to limit the amount of clipping in resonance, with Dave Dixon. Been too long since that conversation though, so I'm not sure I can reproduce the knowledge he tried to impart to me Very Happy

beep · Joined: May 05, 2013 Posts: 105 Location: Germany

I ended up making the feedback loop smaller
(only between the first two poles, adding a resistor between and a passive
diode clipper with a trimmer. It tames the amount of distortion, but also reduces the abilityfor self oscillation.

I adjusted it for a little bit of ringing when reso pot is at 100 percent,
at full oscillation it would still distort too much.

I also added a 1M resistor at the input to get a much lower input signal.
Now I have to amplify it and it gets noisy. I think the shitloads of
wire jumpers (some nice antennas there) are to blame and the oscillators
are on breadboard. I remember someone here told me about
stray capacitance, which is a disadvantage of breadboards and causes
noise/hum.

Tomorrow I will try the opamp circuit again, maybe also with a trimmer.
If nobody says "don't do it" to me, I always use the LM358, it's
my favourite op amp Very Happy

Is the 358 okay?

Cynosure · Posted: Tue Sep 03, 2013 6:48 pm Post subject:

LM358 is a noisy opamp. I usually use TL07x for audio.

Keep in mind though that the TL07x opamps do not go from rail to rail, so you might need to attenuate the sound on the input pin in order to prevent distortion.

JingleJoe · Posted: Fri Sep 06, 2013 7:57 am Post subject:

Don't let that put you off using the LM358 though Smile

I hear so many people complain about it but it's output swing at audio frequencies on a single rail supply is un-matched by any op amp of similar price! (but the lm324 of course which is just two lm358s)