long digital delay build log

[cloudscapes]

pretty much just "boring" stuff to do, now.  I've tweaked the headroom/feedback path to my satisfaction (got a couple extra bits out of that). stuff I have to do before I can start the PCB layout

- relay bypass logic (pretty easy, simple logic, some debounce)
- LEDs showing how its modulated (probably just a 3-4bit R2R ladder)
- interfacing the two toggle switches to the pic
- also finalizing the control layout

There is more to do after I design the PCB and have it fabbed, but it's all code stuff. things like tap tempo, remembering certain settings on power off, that kind of stuff.

[cloudscapes]

well, looks like I just fried my main board
accidental 9v to one of the micro's pins for a second. took out the 3v voltage regulator too, but that didn't save the micro. don't have any spare micros on hand, so looks like I have to wait a couple weeks to get some more.

really sucks, because I was nearly finished the hardware. another day or two and the fried board wouldn't have mattered. I could have gone ahead and started on the final pcb.

[~arph]

  that sucks man...

[SISKO]

Those things always happen in the prototyping/experimenting phase

[cloudscapes]

After sleeping on it, I thought of a way I might continue the hardware without my main board. I have some PIC32 stamps from Mikroe. They're slightly worse than the one I was using, and there are less pins broken-out, but I think they have what I need to finish the hardware. Remap some pins, maybe under-clock the whole delay, that sort of stuff. It should still work. The final hardware stuff I need to do is utility/control related, not related to the sound/quality of the delay itself.

[wavley]

This is an awesome project.

For most of the 90's until the Akai Headrush came out a Digitech RDS-900 with the delay time stretched out to about two seconds of bit crushed goodness was my only looper and I used the heck out of it.  The hold function is awesome.

[cloudscapes]

yeah a lot of those old loopers can be trimmed and stretched up to much more than what they were designed for. they sound like burning (which is awesome)!. sadly can't do that as easily with DSP loopers.

Managed to hack a pic32 stamp + dac + adc + ram together to temporarily replace my fried board. a bit underpowered, and the adc doesn't have the benifit of a ground plane, but it'll be enough for me to finish the hardware side of things. today, maybe!

[cloudscapes]

finished the hardware design!

8 ics. one microcontroller, one ADC, one DAC, one RAM ic, and 4 quad opamps (5 is pictured below, though only using half of 2).

let's see. 4 opamp stages are used as a PWM generator for one of the effects (tried in code, wasn't as smooth since it needed to be at near audio rates). 2 stages for the 4 pole DAC filter, and another 2 for the tone control filter. and 7 stages for the buffering, gain/pregain, feedback pathing, limiter, etc. 4 of the pots below will just be trimpots in the final build. stuff like ADC biasing, feedback trimming, and misc. gain control.



I think I've settled on this control layout, pretty happy with it. any tweaks will probably be to allow room for the parts or something.

[cloudscapes]

well, that was... hard

I also didn't do the power/decoupling on the micro itself on this schematic, since my lite edition of Diptrace ran out of pins.  There's a limit of 500.
I'm wondering if I'll have that problem doing the PCB layout. Probably..... and it's $235 to upgrade to 1000 pins.  
Was only a matter of time I suppose.

[Marcvv]

Wow, Etienne, impressive.

[cloudscapes]

Think I'll make this as two boards, actually. Stacked, like how Strymon does it. The main board with the controls and all the analog stuff, and a smaller board with all the digital stuff. Should help me get around the 500-pin limit for now.

Wow, Etienne, impressive.

Thanks.

[knutolai]

nice work. whats the deal with the 100n caps from the wiper to ground on the pots connected to the MCU?

[cloudscapes]

nice work. whats the deal with the 100n caps from the wiper to ground on the pots connected to the MCU?

just a bit of denoising/smoothing on the pot ADCs. there's a bit of jitter on them without.

[SISKO]

finished the hardware design!
8 ics. one microcontroller, one ADC, one DAC, one RAM ic, and 4 quad opamps (5 is pictured below, though only using half of 2).

What kind of micro is that?

[cloudscapes]

What kind of micro is that?

That's this one:
http://www.mikroe.com/mini/pic32/
which uses a PIC32MX534F064H

But when the project's all done, I'll be using a PIC32MX575F256H instead. more or less the same, but a bit more of everything.

[~arph]

Oh those Pic's and their serial numbers 

[alanp]

I'm impressed I'd have to sit down with a handful of datasheets and a few thick books to even start to have a bit of an idea as to what's going on!

As for stacked boards, I've started to do some layouts with this style. Really good for getting a lot of stuff into a small space.

[cloudscapes]

I'm impressed I'd have to sit down with a handful of datasheets and a few thick books to even start to have a bit of an idea as to what's going on!

As for stacked boards, I've started to do some layouts with this style. Really good for getting a lot of stuff into a small space.

haha, I guess I can try and clarify the sections of the schematic a bit more. I thought of drawing arrows along the main audio paths.

I've done one stacked design before, but it was for an eurorack module:
http://farm3.staticflickr.com/2935/14127918274_592f003821_h.jpg
I'll have to be careful for the delay, since it's going in a box. I did some measurements, I think it should be fine, as long as I put my electrolytics on the underside of the stacking board where there's room.

[cloudscapes]

work in progress
I think I might be able to do this on a single board after all! everything added up (including still missing parts), and with the board headers removed, I'll come 40 short of the 500 pad limit. and diptrace doesn't count vias as pads, so...

[cloudscapes]

hard layout is hard