Karplus Strong Opamp Circuit

dougseidel · Joined: Feb 10, 2010 Posts: 74 Location: NJ

I agree with Gary also - I love reading your ideas and learning from your thought processes Les - it has helped me figure a lot of stuff out.

Doug

Inventor · Posted: Mon Oct 08, 2012 8:19 am Post subject:

Thanks for your perspectives guys, I've had some time away and I feel calmed now. I don't understand why life has to be so difficult, why people have hurt me so badly. Awww, whatever. I've got nothing to say.

Les

elmegil · Posted: Mon Oct 08, 2012 8:25 am Post subject:

Glad to see you back around, Les.

analog_backlash · Posted: Mon Oct 08, 2012 8:48 am Post subject: Welcome Back

Ditto.

telbonic · Joined: Jan 08, 2010 Posts: 39 Location: uk

elmegil · Posted: Tue Oct 09, 2012 5:11 am Post subject:

Really telbonic? That was necessary?

I think it's up to Scott to insist on an apology or not, as he wishes. Maybe you can re-read the thread yourself and take note of a couple salient points that make this particular situation more complex than your garden variety being a dickhead moment, and not tried to stir the pot after it was more or less calmed down.

Inventor · Posted: Tue Oct 09, 2012 7:16 am Post subject:

Well, let's not get all worked up about offenses and apologies and ague and flame and all that. At electro-music.com we like to put out the fires of the flames pretty quick if only to keep the site nice and friendly, and i am STRONGLY guilty of not doing that with my childish outbursts.

I don't apologize ever because I believe that a change of behavior is far more significant than a statement that you regret your actions. In this case in particular, my behavior is not going to change until I better resolve my anger issues against all those people who hurt me so bad.

Oh, and that phrase you mention may be largely true, but not always. There are situations in society in which a person becomes a target of others simply by virtue of culture clash, superior intellect, lack of sensitivity to the feelings of others, or any of a thousand other reasons. In my case it may have been a mix of the above three which cause a group of people to incorrectly identify me as a bad person.

But the mistake was theirs, for the Lord says "Vengeance is Mine", meaning that you don't go around in life attacking people you don't like. It's wrong, and I was wronged. I'm still bitter about it but not that much. Don't judge me by reading this thread alone, I'm not that much of a jerk, lol!

Les

telbonic · Joined: Jan 08, 2010 Posts: 39 Location: uk

Hollywood · Joined: May 22, 2012 Posts: 17 Location: Ireland

To bring this back on thread, would the differences between a strict KS model and what Les has come up with here mean that Les's approach could be used differently? And can you do it with a 741?

JovianPyx · Posted: Wed Dec 05, 2012 7:00 am Post subject:

JingleJoe · Posted: Fri Dec 07, 2012 3:20 am Post subject:

I've used vactrols in filters before, however the main problem is matching them, a slight mis-match in resistance causes the filter to be extra-resonant or less resonant (it changes the Q factor and thus the filter type; chebyshev/bessel/butterworth). I don't know about SVF but I tried this with a butterworth, a twin T and some other twin T variations. It can also cause the filter to oscillate madly when you don't want it to. I did all sorts to get around this, I'll post the circuits eventually. Anyways just thought I'd chime in with my observed vactrol problems.

elmegil · Posted: Fri Dec 07, 2012 6:01 am Post subject:

So Joe, matching....

Just the total resistance, or are there other characteristics which have to be matched as well?

JingleJoe · Posted: Fri Dec 07, 2012 12:04 pm Post subject:

**WARNING THREADJACK IN PROCESS**

Yes, that's it, but I found that slight changes in LED forward voltage caused them to be brighter at different voltages, which leads to different resistances for two vactrols at the same voltage.
Also LEDs aren't really precision instruments, so I think thier light level will vary even at the same forward voltage. I get the feeling a lot of effort is put into making the light wavelengths of all the LEDs in one batch the same, but not the forward voltage / brightness.

Engineering can be defined as problem solving so there's allways a solution:
I put trimpots in series with the LEDs and a 1k resistor and just tirmmed them untill my filters didn't oscillate madly and they behaved how I wanted them to Smile

Also you can simply make one vactrol higher resistance than the other by making the LED resistor twice the value of the other vactrol's resistor. Weighting them in this way will kill a lot of the resonance without fiddling, but may attenuate it too much for your liking or alter the filter type to one which you do not desire.

JovianPyx · Posted: Fri Dec 14, 2012 7:37 am Post subject:

The idea of a vactrol I posted was a more Lunetta style approach, and yeah, I would think that some matching might improve their tracking of each other. It would be a possibly fun benchtop experiment on a solderless breadboard that could be used at least as a learning experience.

I also mentioned OTAs which are already somewhat matched (quite more at least than randomly chosen vactrols).

There are also other types of filters that can be used for resonant drum type sounds, state variable in bandpass mode for one.

Cynosure · Posted: Wed Oct 16, 2013 7:32 am Post subject:

I tried testing this out on breadboard but can't get it to work. I changed it to a +/-5V power supply so used 0V where you used virtual ground. I also used a TL074 as the opamp. I don't have any 330k resistors, so I tried 470k instead.

When it didn't work I tried changing the 470k resistors to 100k, and I tried changing the 0.1uf caps to both 1uf and 0.01uf.

Any suggestions for how to get it to work?

Cynosure · Posted: Wed Oct 16, 2013 11:54 am Post subject:

I made C2 a lot smaller (470pF) and added a resistor from the left side of C2 to the output of the opamp and that gave me really high oscillations.

I think the problem was that it was oscillating too slowly and the click was gone before it had time to oscillate.

Unfortunately, the click goes away so quickly that it is only usable for a clap or wood block type of sound.

I tried using a slow rate squarewave osc instead of clicks, but that just makes two clicks instead (one on the rise and one on the fall).

Any suggestions for how to increase the duration of the sound?

How do i coax it into oscillating more easily?

JovianPyx · Posted: Wed Oct 16, 2013 12:40 pm Post subject:

First, I'd suggest using a bandpass filter.

The answer to how to get longer resonance is higher Q. Too much Q could give you self oscillation though.