Show threadC.Aug 30, 2025

I haven't tried it at higher clock rates yet. I probably will, but it's a case of diminishing returns. To go from 4ms to 3ms, I would have to push the clock from 6MHz to 8MHz. It's a big jump.

But it's much easier to control the delay with the M65831. In its "easy" mode (not using an MCU to control it) there are four inputs to select one of 16 different delay ranges, and then you only have to vary the clock over a few megahertz to cover that range. The longest delay with a 2MHZ clock is ~197ms, while the same clock gets you ~12ms in the shortest range. And if you drop to 1 MHz, you can get it to ~400ms. And it still sounds fine, to my ear.

So this one chip can go from metallic flanging and phasing, through various kinds of chorus, plate reverb, early reflection, room / hall / cave echo, spring reverb, all the way to true echo / delay.

The only downside I can see is that it's a 24-pin IC, rather than the 16-pin PT2399, so it's not as compact. But it's available in SOIC packages, too, so that shouldn't be too bad a limitation.

It's originally a Mitsubishi IC, but clones of it are everywhere. Anyone else tried this IC? Thoughts on it?

2/2

#M65831 #flanger #phaser #chorus #reverb #echo #delay #guitar #effect #pedal

C.Aug 30, 2025

This post is probably only for the intersection of the #Venn diagram with circles labelled "#guitar player" and "#electronics hobbyist". I'm not on fora like #diystompboxes and such, so here it will have to be.

Like many others, I've built delay / echo #pedals / circuits using the ubiquitous #PT2399 digital delay IC. They're dirt cheap and simple to use. But the lowest delay period you can use them for is too long for some things, like #flanger / #phaser effects, or even some types of #chorus effects.

But I haven't seen much discussion of alternatives. I've been playing around with one I've built using the #M65831 digital #delay chip. It's not quite as cheap as the PT2399, but you don't have to look very hard to be able to get them for a couple of bucks (Canadian). But it seems to be so much better in this application.

First, it's designed and documented so that you can use a crystal to set its clock OR drive it directly with a CMOS-type square wave. No fiddling with drain currents and on-reset latchup problems.

Secondly, it sounds (to my ear) far better than the PT2399. The #datasheets I have for the PT2399 don't state a sampling rate; the M65831 uses 500 ksamples/s for lower delay times.

Thirdly, the range of delay available is larger. The shortest documented for PT2399 is ~31ms. For the M65831, it's 12.3ms using the default clock of 2 MHz. But the one sample I've tried is happy to run at 6 #MHz, for a delay of ~4ms. That's short enough for #flanging!

1/2