PT2399 Repeats Noise Problem, Need Some Help...

[Paul Marossy]

So I'm messing around with a PT2399 based delay that uses two of them in series, and it's also using a NE571 compander chip. I'm more or less satisfied with it except for one problem - depending on the delay time setting, when the repeats control is turned up past one or two repeats I get this original series Star Trek phaser like noise after I stop playing and let it be quiet for a moment. It just spontaneously happens. I like having the maximum repeats I can get and it's OK while playing, but the noise happening when I stop playing is not cool.

The repeats control is on the NE571 chip, it's like a feedback loop with a pot on it. I surmise that maybe it's the compander chip that is freaking out? Here is essentially how it's wired:



Any ideas on how to make it stop that? I was thinking that maybe a cap needs to be added someplace, but I don't know much about these PT2399 circuits. I don't really have any noise problems otherwise, no quant. noise that I can hear.

[Mark Hammer]

I bought myself a Korg Monotron Delay recently, and its noise source IS the audio crap produced by the onboard PT2399!

Seems to me you have a couple of simple options available:

1) Cut back on the delay time a bit.  When the 1st and 2nd generation analog delays came out, and manufacturers were pushing humble MN3005 and SAD4096 chips as far as they could (and only one per pedal, thank you very much), some would include "liberal" lowpass filtering with rolloffs just below where they needed to be for the longest delay times.  You'd get audible whine at max delay, but decent bandwidth at shorter delays; the manufacturer assuming that you could do something at the amp to reduce the audibility of that whining when it happened.  Others would design for no whining at ANY delay, with the result that the unit would have less delay bandwidth, and/or shorter delay times, than competing products.

2) Make your lowpass filtering stricter for the feedback loop.  Quantization noise and other audio grime accumulates over repeats.  Figre out where to insert some additional lowpass filtering in the feedback loop.  It doesn't have to be severe, just there when you need it.   F'rinstance, the Rebote 2.5 has a 10k-15k fixed resistor in the feedback path.  If a person replaced that with a pair of 5k1 in series, and ran a cap to ground from their junction, you'd add one more pole of lowpass filtering to what is already there.  Pick your corner frequency wisely, and you'll have a treble rolloff that gets juuusssstttt stringent enough to keep the accumulating hiss/noise at bay over multiple repeats.  Note that this is entirely seperate from whatever lowpass filtering is imposed on the wet signal en route to the output mixing stage.

[Paul Marossy]

Thanks Mark, I'm thinking your suggestion on #2 is the place to start.

I guess maybe it is quant. noise, but it wasn't like the noise I listened to in a soundclip linked in another PT2399 topic.

I'm currently getting about a 1 second max. delay, which I like because it can do some fast picked "frippertronics" kind sound kind of well at that delay time (I know Fripp used between 3 and 5 seconds of delay but I've him use a shorter delay too). Well, maybe it's more of a King Crimson kind of thing. Anyway, I don't want to go beyond that amount of delay time (1 second), but I don't want to go shorter than that either.

[psychedelicfish]

I was playing with a ~1s delay. I used 1uF CC0 and CC1 caps, and 1uF for mod and demod caps too, and noise wasn't bad

[cortezthekiller]

I had made a rebote delay a while back which had an issue similar to what Paul is describing. It turned out I had one of the electrolytic signal caps in backwards. It worked, though would make random glitchy noises at times. In the diagram on the first post it shows 1uF caps and no polarity. Did you use electrolytic or unpolarized caps in those spots?

[Paul Marossy]

I had made a rebote delay a while back which had an issue similar to what Paul is describing. It turned out I had one of the electrolytic signal caps in backwards. It worked, though would make random glitchy noises at times. In the diagram on the first post it shows 1uF caps and no polarity. Did you use electrolytic or unpolarized caps in those spots?

I used electrolytic caps only where indicated. Mark's idea of putting filter on there worked, I have it to where I can live with it now. Not perfect, but what PT2399 chip is?  

[~arph]

Did you try a couple of different PT2399's as well?  Some just sound worse then others

[Paul Marossy]

Did you try a couple of different PT2399's as well?  Some just sound worse then others

No, but I have four more to try.