What steps to take to get rid of LFO ticking?

[Ettore_M]

Hey guys!

I just built an optical tremolo circuit based on an old tremolo design (I'll give hints later..  ) on breadboard.

At the start, it didn't have ticking at all! But after some changes in the circuit, I got a huge amount of ticking.
Although I returned it in the previous settings, the ticking remained.

It sounds pretty interesting and it's really simple. If only I could take the ticking away.. 

So what steps do you guys take to remove LFO ticking when designing a circuit? 

Hector

[Ettore_M]

Replying to my own thread... 

I found this interesting article over at Flo FX. It quotes some of Mark Hammer's ideas on how to remove ticking.
https://sites.google.com/site/flofxdiy/deticking-lfo

But, still, I'd like to hear your techniques too. 

Hector

[culturejam]

Can you post a link to the schematic? That would help spot trouble areas.

Generally, I use a separate op amp package for the LFO (if it's IC-based, that is). And I also generally add extra low-pass filtering on the power supply pins of the op amps (or collectors if BJT-driven). A simple series 100R + 100uF between the rails will do it.

[alanp]

Adding a 100n cap as close as possible to the LFO op amp's power pin can also help.

[Ettore_M]

Can you post a link to the schematic? That would help spot trouble areas.

Generally, I use a separate op amp package for the LFO (if it's IC-based, that is). And I also generally add extra low-pass filtering on the power supply pins of the op amps (or collectors if BJT-driven). A simple series 100R + 100uF between the rails will do it.

Hey CJ. The man I was thinking that would the answer. 

The circuit is an idea of mine (if I can tell it that  ), so I don't have a schematic. It's basically a CMOS tremolo, that's based on Craig Anderton's tremolo, using a CD4049. Only one chip, doing the LFO and amplifying. 

But, maybe that's the problem. Maybe you can't have the same chip doing both things. 
I say that, because I tried the 100R and 100uF to Vcc's pin, but it only cut down the ticking in half. But it didn't dissapear. 

Is there anyting else to do? It sounds pretty awesome and it's super simple, so it's a pity if we can't make this work!

Hector

EDIT: I just drew a quick schematic. 

[culturejam]

It's basically a CMOS tremolo, that's based on Craig Anderton's tremolo, using a CD4049. Only one chip, doing the LFO and amplifying. 

Ah, the old CMOS gate IC used as a relaxation osciallator trick, eh? 

I've not played around with that one, but I suspect that having the audio and LFO portions sharing the same internal power/gnd pins will make overcoming the ticking more difficult.

But, maybe that's the problem. Maybe you can't have the same chip doing both things. 
I say that, because I tried the 100R and 100uF to Vcc's pin, but it only cut down the ticking in half. But it didn't dissapear. 

I would first try what I call The American Approach, which is simply this: "If X is good, then 2X simply HAS TO BE twice as good." 

So go with 220R and 220u. I once had 470R and 1000u on a circuit to make it shut up.  But maybe you want a steeper corner frequency, so perhaps literally doubling up with a two-pole filter (instead of doubling the values of a single-pole filter) would be better. So maybe try a 100R/100u, and then another one in series, and then the power pin.

Of course, the easiest answer is to use the audio path of the Tremulus Lune and be done with it. Granted, that's not as fun as trying to make it all work with one IC, so by all means keep on tinkering. You may well just figure it out.

[micromegas]

Is that a sheet with gain/central frecuency calculus under the schematic? I wish I had enough time to study circuits again... not the telematic stuff I'm submerged in now.. (all these protocols, Markov queues and standards are the worst thing I've ever had to study...)

I hope you solve this. I'm going to breadboard it as soon as I finish my exams, I happen to have a bunch of 4049s I bought on december....

[micromegas]

by the way: if the calculus are related to the circuit it would be a cool thing to share them. It made me think that we should share that kind of info more often so others could learn from the work and not only use the circuits shared (some circuits are easy to understand to almost everyone, but others don't).

I promised myself that I won't build anything this year whose behaving I don't completely understand.

[Ettore_M]

Is that a sheet with gain/central frecuency calculus under the schematic? I wish I had enough time to study circuits again... not the telematic stuff I'm submerged in now.. (all these protocols, Markov queues and standards are the worst thing I've ever had to study...)

I hope you solve this. I'm going to breadboard it as soon as I finish my exams, I happen to have a bunch of 4049s I bought on december....

Haha! Nice spotting right there! 

I'm studying for my tomorrow exams on Medical Instrumentation, and that's the analysis of an inverting amplifier with a T feedback network.

Good luck to your exams!

Hector

[Ettore_M]

by the way: if the calculus are related to the circuit it would be a cool thing to share them. It made me think that we should share that kind of info more often so others could learn from the work and not only use the circuits shared (some circuits are easy to understand to almost everyone, but others don't).

I promised myself that I won't build anything this year whose behaving I don't completely understand.

That would be a nice idea, if we had the right spot on the forum to post them. Maybe if there's interest..

Hector

[micromegas]

Is that a sheet with gain/central frecuency calculus under the schematic? I wish I had enough time to study circuits again... not the telematic stuff I'm submerged in now.. (all these protocols, Markov queues and standards are the worst thing I've ever had to study...)

I hope you solve this. I'm going to breadboard it as soon as I finish my exams, I happen to have a bunch of 4049s I bought on december....

Haha! Nice spotting right there! 

I'm studying for my tomorrow exams on Medical Instrumentation, and that's the analysis of an inverting amplifier with a T feedback network.

Good luck to your exams!

Hector

Good luck to you too
That subject sounds cool. I got my "Electronic Instrumentation" exam last week and passed it, hope you have the same luck.

[Ettore_M]

CJ, I knew you're my man! That kind of reply I wanted!

I would first try what I call The American Approach, which is simply this: "If X is good, then 2X simply HAS TO BE twice as good." 

Firstly, I'm just stealing this for my signature! 

So go with 220R and 220u. I once had 470R and 1000u on a circuit to make it shut up.  But maybe you want a steeper corner frequency, so perhaps literally doubling up with a two-pole filter (instead of doubling the values of a single-pole filter) would be better. So maybe try a 100R/100u, and then another one in series, and then the power pin.

That's what I'll do!  I'll start increasing the values until I get the thing to shut up! 
But what do you mean "another one in series"?

Of course, the easiest answer is to use the audio path of the Tremulus Lune and be done with it. Granted, that's not as fun as trying to make it all work with one IC, so by all means keep on tinkering. You may well just figure it out.

And that's why you're my man! Isn't it more fun to try (and fail) to get what you had in your mind at first? No compromises! It's the challenge! 
And if I go for the extra IC that would be just a CA tremolo, right? 

Thanks!

Hector

[micromegas]

But what do you mean "another one in series"?

I guess another filter stage to make a 2nd order filter?

[Ettore_M]

But what do you mean "another one in series"?

I guess another filter stage to make a 2nd order filter?

Indeed, you're right! Now I got it! 2nd order filter = steeper cut-off. Thanks!

Hector

[culturejam]

But what do you mean "another one in series"?

A passive second-order low-pass filter. Like this: