Any Good LFO shecmatics

[ThePastRecedes]

I've be searching for some good lfo schematics for the last few hours, I've even use the search engine 'million short'. But I've had no luck in what I'm looking for.

The LFO is going to be used for modulation in a delay pedal. I'm looking for something with a few peramiters not just two. The majority that I have found just have controls for rate and depth.

I'm looking for it to have a wave shape control too. I'd like it to cycle through sine, triangle, sawtooth and square. Maybe in between wave shapes too.
The other way I could do it is to build 4 separate LFOs and switch between them, but that will probably result in a very high parts count.

thanks!

[Pyr0]

try searching for function generator circuits.

Or have you thought about using the old ICL8038 function generator chip. Obsolete but plenty still available. It can make sine, triangle and square waves with just a few external components.

[R.G.]

One of the difficult but not commonly recognized issues with LFOs is that each LFO application is a custom design that has to be fitted to what it makes sweep.

The circuits which are swept by LFOs have varying needs for the DC level of the LFO, the DC level shift as it changes size (which may be zero shift or not), the min and max sizes of the LFO waveform, and whether the LFO drives whatever it drives by being a current source or a voltage source.

Picking a good LFO oscillator gets you part of the way there. The next 80% of the job is to get the LFO adapted to the circuit it controls.

[mth5044]

Check out the Electric Druid's tap LFO.

http://www.electricdruid.net/datasheets/TAPLFO2Datasheet.pdf

[artifus]

also: http://www.geofex.com/article_folders/lfos/psuedorandom.htm

[slacker]

You could have a look at the LFO in my Echo Base this is a standard design that gives triangle and square, and can be modded to give saw and sine without too much effort.

If you want something fancier have a look at some synth DIY websites, they'll need some work to convert them to stompbox use but there's loads of designs. There's a bunch of simple ones here, the variable skew one will do everything you want.

http://www.musicfromouterspace.com/index.php?MAINTAB=SYNTHDIY&VPW=1075&VPH=516

Or you could have a look at the electronic druids digital LFOs, they do everything you want and much more.

http://www.electricdruid.net/index.php?page=projects.lfo9
http://www.electricdruid.net/index.php?page=projects.taplfo

Like R.G. said building the LFO is often the easy part.

[WhenBoredomPeaks]

These are maybe a bit over the top:
http://yusynth.net/Modular/EN/LFO/index.html
http://www.cgs.synth.net/modules/index.html#cgs58
http://www.musicfromouterspace.com/index.php?MAINTAB=SYNTHDIY&PROJARG=ELECTRONICS/synthdiywebsites.html&VPW=1356&VPH=515

[Processaurus]

One of the difficult but not commonly recognized issues with LFOs is that each LFO application is a custom design that has to be fitted to what it makes sweep.

The circuits which are swept by LFOs have varying needs for the DC level of the LFO, the DC level shift as it changes size (which may be zero shift or not), the min and max sizes of the LFO waveform, and whether the LFO drives whatever it drives by being a current source or a voltage source.

Picking a good LFO oscillator gets you part of the way there. The next 80% of the job is to get the LFO adapted to the circuit it controls.

Another hitch with LFO's is that depending on the circuit they are being hooked up to, they may or may not sound like the waveform going in.  In modulated delay circuits, like choruses, a triangle LFO can end up sounding square-ish, because it doesn't ramp the pitch up and down like one might think, but by ramping the delay time up and down at a linear rate it can make the pitch alternate in an abrupt way between steadily slightly sharp to steadily slightly flat.  A square LFO can just sound weird, depending on how much audio is inside the delay element, it could end up sounding like the pitch of the audio coming in, but with quick little burbles on the rising and falling edges of the square wave.

A tremolo, as another instance, modulates the volume, but as the human ear hears volume changes in a logarithmic way (and hence volume controls are logarithmic), for the wave to translate like it is turning the volume knob up and down it would ideally want to have some sort of logarithmic conversion, otherwise it would sound like all of the action was in the lower half of the waveform.  Synth filters can have linear to logarithmic converters too, for control voltage inputs.

This isn't to discourage you in experimentation, just to point out there can be more to implementing an LFO musically, beyond generating the waves.

[Processaurus]

One of the difficult but not commonly recognized issues with LFOs is that each LFO application is a custom design that has to be fitted to what it makes sweep.

The circuits which are swept by LFOs have varying needs for the DC level of the LFO, the DC level shift as it changes size (which may be zero shift or not), the min and max sizes of the LFO waveform, and whether the LFO drives whatever it drives by being a current source or a voltage source.

Picking a good LFO oscillator gets you part of the way there. The next 80% of the job is to get the LFO adapted to the circuit it controls.

Another hitch with LFO's is that depending on the circuit they are being hooked up to, they may or may not sound like the waveform going in.  In modulated delay circuits, like choruses, a triangle LFO can end up sounding square-ish, because it doesn't ramp the pitch up and down like one might think, but by ramping the delay time up and down at a linear rate it can make the pitch alternate in an abrupt way between steadily slightly sharp to steadily slightly flat.  A square LFO can just sound weird, depending on how much audio is inside the delay element, it could end up sounding like the pitch of the audio coming in, but with quick little burbles on the rising and falling edges of the square wave.

A tremolo, as another instance, modulates the volume, but as the human ear hears volume changes in a logarithmic way (and hence volume controls are logarithmic), for the wave to translate like it is turning the volume knob up and down it would ideally want to have some sort of logarithmic conversion, otherwise it would sound like all of the action was in the lower half of the waveform.  Synth filters can have linear to logarithmic converters too, for control voltage inputs.

This isn't to discourage you in experimentation, just to point out there can be more to implementing an LFO musically, beyond generating the waves.

[duck_arse]

so, I've been messing with the "puny-phase" from another place, and needed an oscillator, one that didn't clang, yang, tik or suck. here is where I am so far.



the shark-tooth output of the standard single integrator osc looked a real mess after hyper-ing, so I went back to the schmitt and integrator, which had been tikking. only one circuit I found included the cap and resistor (R3 / C2) on the inverting input of the schmitt. I tried it and it seems to round the very tips of the triangle, and slightly slow the osc rate.

after messing with all the hyper-ing methods I could find, the "standard synth sine-shaper" circuit appeared. seeing as I still had half an lm324, I tried it and a rail-splitter. with one transistor unbalanced, some resistor changes and careful tuning of the sine-level trim, I was getting 6 to 7.5V output, with about 2V offset. playing with the v-ref level (shape), the output changed from hyper-sine to ugly sine-tri to hyper-triangle.

just what I wanted. with the addition of one switch and two diodes, you also get a ramp, left or right, which can be tuned with the shape control for curves and points where you like them. with this heating the ldr's of the puny-phase, I was getting oohs, ahhhs and wahs from depth setting variations. no long, slow jet-type sounds, maybe the wrong phase circuit for this. with some of the standard mods at the output, this may be useful for fet-based circuits as well.

of course, with careful layout and wiring, it can be made to tick like a clock if that's what you want. it needs range-switching for the rate cap, as ever, probably the rail-splitter is overkill and I'm sure you engineering types can improve/correct any number of flaws, but here it is anyway. any input, improvements or feedback appreciated.