Help with this LFO, I don't like the waveshape...

[Slade]

Hey,
I've built a flanger and I have this LFO in the circuit:



I don't like the shape I'm getting from it as it's not "hypertriangular" and it's almost the inverse of that... I don't have an oscilloscope, but from what I hear from the effect I can describe the waveshape by this image:



I tried adding an inverter on it's output once but didn't worked.

I know I can just try a different LFO but first I want to know if I can do something to make this one behave like I want it to.

Is there anyway I can modify this type of LFO to get the hypertriangular waveshape that I'm looking for?

Thanks in advance.
Fernando.-

[jonny]

get rid of the .01u cap. move one end of the 82K,  the side  on IC2b - input, move it so it goes to the IC2b +  input. Then connect the - input of IC2b to the +input of IC1a. You may have to ditch the depth control and take the output just from the 1st op amp. If you take the output from the 2nd op amp you get a square wave. You can put a pot in place of the 5k1 resistors and it will adjust the symmetry(sawtooth).

Edit: when i said
Then connect the - input of IC2b to the +input of IC1a. You just need to short the 100k resistor between them.

[Slade]

My bad... 82k resistor was already connected to IC2B (+) Input.

So, here's what you recommend, am I right?

[jonny]

Yip i'm pretty sure that's right. I've built one like that before and it worked great. Only difference being I didn't have that depth pot.

[Slade]

Thank you for your help
What you mean with taking the output from the first opamp? How would you connect Depth control here?

[jonny]

No problem. Take the output from between the output of IC1a and the 33k resistor but I think the depth control should work, just haven't tested it myself. Hope it works!

[Mark Hammer]

That .01uf cap may well be to prevent audible ticking.  Take a read through the Stompboxology chapter on tremolo that Charlie Barth has posted: http://www.moosapotamus.net/IDEAS/stompboxology/MoTremLo.htm

And take a look here for how to generate a hypertriangular LFO/clock: http://hammer.ampage.org/files/hypertriangleclock.gif

[cpm]

i got to use a logarithmic transformation for the LFO for a phaser. Its going into OTAs, so with other devices it may need modification

half of a sine is log-scaled, the other half is linear
its done with a log-amp opamp configuration, one germanium diode. Plain text book example...



the best thing about it is that amplitude is modified before the transformation, so the final OTA response is linear when varying the pot

[Slade]

That .01uf cap may well be to prevent audible ticking.  Take a read through the Stompboxology chapter on tremolo that Charlie Barth has posted: http://www.moosapotamus.net/IDEAS/stompboxology/MoTremLo.htm

And take a look here for how to generate a hypertriangular LFO/clock: http://hammer.ampage.org/files/hypertriangleclock.gif

Thanks, Mark. Last night I was looking to that gif at your site. Does it use bipolar power supply? Is it necessary for the circuit to work or may it be used with a single PS?


I tried the changes (on breadboard) for the LFO, but it didn't make any change. My guess, still having no oscilloscope, is that the waveshape might be almost triangular at both sides and it only needs to be less triangular at the bottom, so at the end it might be necessary to just add a different LFO with a module into the box.

[Mark Hammer]

Thanks, Mark. Last night I was looking to that gif at your site. Does it use bipolar power supply? Is it necessary for the circuit to work or may it be used with a single PS?

Well, the JFET and CMOS part doesn't seem to need a bipolar supply, and the op-amp part is just a standard 2-op-amp LFO that can be easily adapted to work off a single-ended supply.

[Slade]

Thanks Mark.

While I was waiting for your response I've been trying Ross Phaser LFO:



It works perfectly on my BB but the problem I'm having with this one is that the output voltage of the hypertriangular waveshape moves in a very low voltage range. The current LFO of my flanger moves in a range from 3.5v to 7.5v while this one goes from 0.8v to 4.5v. I tried adding a non inverting amplifier with a gain of 2 to raise the output voltage range, and I have it moving from 1.8v to 8.2v (much more than I need) but now it doesn't have the hypertriangular waveshape anymore.

Is there anyway I can modify this LFO to get the voltage output I need without sacrifying its waveshape?

I'll be trying the one you've posted after this one.

[Slade]

Okay... I've been playing a little with Ross phaser's LFO and this is what I came up with. Values in red are the ones I've changed/added from the original:



As I said before, I don't have an oscilloscope to "see" the waveshape I'm getting but it's definitively better than before, here's a sample for you to listen:

http://dl.dropbox.com/u/3565959/Flanger%20Demo%2013%20%2813600%20LFOl%29.mp3

I'm still not totally convinced with it but it's getting better...
I'm just trying some different things for it on my breadboard, any advice would be appreciated

[PRR]

> The current LFO ...moves ...from 3.5v to 7.5v
> .....while this one goes from 0.8v to 4.6v.

7.5V-3.5V= 4V p-p centered on 5.5V
4.6V-0.8V= 3.8V p-p "centered" on 2.7V

3.8V is nearly 4V... you don't need more swing.

2.7V (or 2.4V due to lop-sided) is much less than the 5.5V center you want.

Where did that 2.4V come from? Look at R27 R28 270K+100K dividing a 9V supply. The node is at 2.43V. Via R29 it sets the DC level out of U4B; U4A follows this voltage, and the wobble swings both ways from there.



Raise that node to about 5.5V. Change R27 to 56K, 62K, or 68K.