looking for multiple LFO chip

[birt]

I'm looking for a chip i can build six LFO's with. they don't have to be great sounding with a lot of options, simple LFO with choppy/smooth control will do just fine.
are there any chips that can do this?

[R.G.]

The CD40106 hex schmitt trigger inverter will do that. It takes one cap and one resistor to make one LFO. You will have to buffer the triangle outputs, though. Another CD40106 could do that. The triangle outputs are 1/3 of the power supply in size and the square waves are full power supply in size, so you'll have to use a resistor divider to get the square waves down to a matching size or amplify the triangles up.

Otherwise, you're looking at
(a) a microcontroller chip and programming
(b) at least one opamp section per LFO output, probably two
(c) there could be some other chips I'm not aware of.

[birt]

so one part of the chip needs a feedbackresistor and a cap from input to ground.
i don't really understand how to get the triangle but if they go down in voltage when they get less choppy it's probably possible to add adual pot to control the chop. as things get choppier by one half of the pot, the other half reduces the output. or do i get it completely wrong?

[slacker]

Have a look at the LFO from my slackfilter. http://www.eskimo.plus.com/fxstuff/newslackfilter.png
That uses a 40106 to give triangle and square waves. If you copy the section with the pot in it that would give you 6 LFOs. You'd need to buffer the triangle ouputs either with simple transistor buffers or opamps like on the schematic.
I've got another version of this that I'm experimenting with at the minute. It's even simpler and depending on the use it doesn't need buffering. I'll post it later.

RG - How do you use a 40106 to buffer the triangle outputs? Wouldn't that just give you square outputs. I know you could do it with a 4069 biased as a linear amplifier, but I can't see how you can do it with a schmitt trigger.

[slacker]

Here's the variable speed section from the new LFO I'm working on. It gives a triangle or square output and can drive high impedance loads without a buffer.



The bottom schematic is how it's set up in my design, the top one shows it with a transistor buffer which would be needed if you wanted it to drive an LED. You could replace the transistor with an opamp buffer.
The LFO normally has a square wave output but when the 10uF cap is connected to ground by the triangle/square switch it smooths the waveform into a triangle. The resistor marked XXX is optional, without it the square output will be bigger than the triangle. Adding the resistor when in square mode makes a voltage divider, like R.G suggested, and lowers the volume of the square output. You'd need to play around with values to get one that suits you, I haven't bothered with it because the depth control can take care of the differences in volume.
I'm using my LFO to drive JC's Nyquist Aliaser with suitably mental results 
If you want some more ideas the Daddy as far as these kind of LFOs go is Ken Stone's Super Psycho LFO

[birt]

great stuff slacker!
i thiks it's ideal for what i want to do. i want to make an effect that divides the input into 6different frequency bands that have seperate tremolos and are mixed back together after those. i just need to find a way to control all of the speed settings at the same time in a way they never go togheter.

i guess i could use a different feedback resistor on every LFO. with a dual gang pot i can control the speed of 2 of them and then use the output from one to drive a led wich controls an LDR wich sets the speed of another (but it will also be different because of the different feedback resistor in series with the LDR). if this idea works i could have 1 knob to control the speed of all 6 LFO's.

maybe i could even let out the switch for smooth or choppy and just go with the best sounding way. if the effect is usable that is.

if you have an idea to split the signal into 6 different frequency bands (with 6 outputs) let me know.

[R.G.]

The LFO normally has a square wave output but when the 10uF cap is connected to ground by the triangle/square switch it smooths the waveform into a triangle.

Actually, a cap to ground rounds off the leading edges of the square wave until it starts limiting the height, then it approximates a triangle. But the triangle waveform height goes down as frequency goes up because the capacitor only integrates at one speed.

i want to make an effect that divides the input into 6different frequency bands that have seperate tremolos and are mixed back together after those. i just need to find a way to control all of the speed settings at the same time in a way they never go togheter.

Use different value caps and replace the resistors with resistors in series with a JFET or CMOS switch. Control the switch with a PWM waveform and all the resistors will vary at the same time. Different cap values would give different frequencies.

Or use one LED and six LDRs clustered around it. The LDRs will change together, although not by the same amount.

if you have an idea to split the signal into 6 different frequency bands (with 6 outputs) let me know.

See "crossovers" at Elliot Sound Products or the "Quadrafuzz" by Craig Anderton.

[slacker]

Actually, a cap to ground rounds off the leading edges of the square wave until it starts limiting the height, then it approximates a triangle. But the triangle waveform height goes down as frequency goes up because the capacitor only integrates at one speed.

Yes you're right the triangle output isn't perfect, the original way i did it by tapping off the junction of the input and the cap to ground and buffering it works better. For my purposes though this way does the job and simplifies the design.

[R.G.]

Simple's good, and if it does what you need, great!

[birt]

i guess it'll do what i need too. and i think i'll go for the multiple LDR approach to keep the design simple. i still have to make 6 tremolos around the LFO's and the equalizing part too so it's gonna be quite big i think.

plus there's the fact that i really don't have a clue how to implement a CMOS switch :p