Is there a way to freeze an LFO?

[kvb]

Building an electronic drum taught me that an LFO can be triggered and that when the power at the front of the op amp is removed the LFO will fade/stop.

So then getting the LFO to rise to a certain level and then stop . . .

Is that a sample and hold? I don't know anything about those circuits yet.

[kvb]

I took a look in musicfromouterspace and decided that I really did not want to build a sample and hold for a simple task.

I think I answered my question in the "blend" thread.

[gez]

Sample and hold would be one way, but when you 'start' the LFO again ('release' the hold and start sampling again) there will probably be a jump to a new voltage as the LFO has been cycling away as normal during the hold. 

Perhaps a better approach would be to stop the clock in a "stepped LFO".  There's more complexity to the LFO, however.  Unless your PIC skills are up to scratch.

[Paul Perry (Frostwave)]

Only a microprocessor based solution will let you 'freeze' a LFO at part of a cycle, if that is what you want.
Otherwise you are at the mercy of capacitors that will slowly leak away.
But, for a short time, you can just disconnect one end of the timing cap in a LFO (if it is the integrator type of circuit) and it will stay where it is, for a bit.

[dxm1]

...or, use a FET or 4016 to switch the LFO output based on level...

[Processaurus]

This is kinda like what paul and dxm said, I believe you can do what you want in a common one or two opamp LFO by using a FET input (for the high high input impedance) opamp for the LFO (like tl072), and using a switch or FET or 4016 or 4066 to disconnect the pot that feeds (and sucks current on the negative cycle) current into the big electrolytic timing cap to ground, but leaving the cap connected to the opamp input.  It wouldn't hold it forever, but it would be close enough for rock an rollll.

Good concept by the way, I always wish I could stop hold lfo's on phasers and the like when the get to the point were they resonate with the notes you're playing.

[salocin]

I actually stumbled upon a way of doing this with a dual opamp lfo, such as that in the tremulus lune. disconnecting on of the pot connections did it, but can't remember which pot

[R.G.]

Freezing an analog LFO can be done by careful notice of what to disconnect when you want to freeze. However, as with all analog sample-and-holds (there's gotta be one of those, even if it's implicit) you must deal with the problems of sample aperture and leakage. Getting it to freeze for a second is easy. Longer is harder. Much longer is much harder.

uController LFOs - which can be 8-pin DIPs, $1.50 each - are clearly the way to go with this. After that it's ASMOP.

[Mark Hammer]

Seems to me that the most straightforward analog method of doing this would be to use an interruptable-sequencer/counter approach.  A pair of 4017s would give you 18 steps from what I gather.  Smooth those out with some (variable or fixed) slewing, and as long as the sweep rate is slow enough for you to catch the point of preference in the sweep cycle, you should be able to stop things by preventing a clock from making the 4017 take any more steps (for the moment) and then continue where you left off.

The difference between this and a S&H approach is that the step output from the 4017 is fixed at whatever logic high is, and the user divides it down to whatever they want below that.  In contrast, a S&H approach has no fixed "default voltage", and droops with time.  As far as I know, the counter output remains stable unless you continue counting.

[tommy.genes]

Maybe not exactly what you want, but R.G. also gave me some good advice in THIS THREAD.

-- T. G. --

[kvb]

The logic idea sounds neat. Eliminating the 'steps' would be the tough part.

You could get an effect to sweep right to the top (flanger?) and then hold it, play some notes, and then have it come swooshing down only when you told it to.

And, once again its PICs to the rescue.

I understand that micro controllers are the modulation/switching handy-man in this realm.

(aside) - The trouble (for me) is that it took me two years of building other peoples circuits to decide what it was that I wanted to create for myself - and that thing ended up taking me what is going on nine months of development. It took me many separate months of looking at logic circuits / chapters on logic until my brain went from zero to 1 (+5V). I've got something cooking here too.

I figure that for me to learn about the PIC thing it would take about two years.