Help me debug a DigiDelay

[soggybag]

I built two DigiDelay PCBs. They both have the same problem.

There are two 3v3 regulators. Which didn't seem to be working correctly with an initial power check before installing the ICs. I was worried I put them in backwards so I pulled them out. I think the L78L33ACZ I got from Tayda have a reverse pin out from what is shown on the PCB.

Once I got these out checked a few things with my multimeter. Both boards measure ~20k resistance across the +9V and Gnd! Which seems like a problem.

All the parts look correct, there are no chips just resistors, caps and diodes, I can't see any solder bridges.

Seems weird that they are both the same. If it's a solder bridge or mechanical mishap it would be odd to make the same mistake on both.

I'm looking for debugging tips. I'm stumped.

[Slowpoke101]

If this is the Electric Druid's DigiDelay then 20K from 9V to GND is fine.
The resistors used for the Vref (4.5V ) voltage divider are 10K each and are in series so 20K is fine.
Are you getting 3.3V (with all other ICs not installed ) ?

[soggybag]

Thanks for the reply. I was getting 3v3 from the first regulator but the second was showing ~8V5 for some reason. This was where I though they might be backwards. So I looked up the data sheet for the parts I got from Tayda.

https://www.taydaelectronics.com/datasheets/A-492.pdf

Looking at the PDF it shows pin 1 is Vout. The DigiDelay schematic shows pin 1 as in.

https://electricdruid.net/wp-content/uploads/2016/10/DigiDelay-Schematic-Pg3.jpg

Maybe I just had these backwards the first time?

[Slowpoke101]

Both the datasheet and the DigiDelay's silkscreen overlay do match.
So both regulators go in the same way. Put them in and see if you are getting 3.3V or 8.5V
If you get 8.5V then one of the regulators is in backwards or has failed.

[soggybag]

Thanks for looking at that confirming. I'll give it try tomorrow.

[soggybag]

Thanks again! Everything is looking like it should now.

[soggybag]

Finally got around to finishing a pair of Digidelays. They both work but they both have the same problem. There is a digital hash/noise when the delay is engaged. It goes away after a while and things sound normal. It seems like the noise is recorded in the delay line when the effect is turned on. It passes through some repeat cycles and then is gone.

[ElectricDruid]

Finally got around to finishing a pair of Digidelays. They both work but they both have the same problem. There is a digital hash/noise when the delay is engaged. It goes away after a while and things sound normal. It seems like the noise is recorded in the delay line when the effect is turned on. It passes through some repeat cycles and then is gone.

Yep, that's because that's pretty much exactly what happens.

The firmware doesn't wipe the memory at start-up. Instead it just starts up with the delay bypassed. If you have the delay set to a long setting and then engage it immediately, you hear the end of the random contents of the RAM being fed out to you (noise).

I've been meaning to go back and correct this, along with one other slightly annoying bug, which is that the DigiDelay chip boots up in Bypassed mode. If it booted up into Effect On mode, it'd make it much easier to build the circuit with a true bypass foot switch if required.

But there are so many more fundamental issues with it that it demands a full redesign, not a couple of tweaks, hence why it has never happened.

Rest assured it's not you though - that _is_ what it does. Just give it a few seconds to settle before you engage it. Leaving it with the Repeats and or Level turned down minimises the problem. I realise that's a work-around not a solution.

[soggybag]

Thanks for clearing that up.

[ElectricDruid]

Very nice builds, btw. Those look sweet.

[soggybag]

I was playing with the Digidelay today and I notice two things.

First there is a faint high pitched oscillation in the background, it's constant and doesn't change with settings.

Second, there is a little pulse at the delay tempo.

Is there a away to get rid of this?

[ElectricDruid]

High pitched oscillation isn't right. I'd check around the filters and the other op-amps to make sure nothing's oscillating. One builder had oscillations in one of the filters - turned out it was a bad cap, and it turned the resonance of that stage up to eleven.
It could also potentially be heterodyning from the sample rate if you're feeding the signal into something else that has a high frequency clock.

A little pulse at the delay tempo..not sure. There is an clock output pulse at the delay tempo from one of the pins on the digidelay processor, but it doesn't go anywhere (unless you've wired it up?) and I've never heard of it causing any ticking problems.