Hi guys, anybody here playing with embedded MCUs and DSPs and signal processing?
Enjoy the benefits of a premium membership!
Upgrade Account
-
TalkBass has been independent since 1998. Add your voice.
Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
Join free Log in
Want zero display ads or expanded classifieds tools? Compare plans.
The Passinwind Open Source Preamp
- Thread starter Passinwind
- Start date
Working towards it. Just got the analog front and back ends debugged.
Maybe this should be in it's own thread?
- Dec 3, 2003
- 23,049
- 45,913
- Disclosures
- Designer/Toaster Tech @ Passinwind Electronics
@Jeff Siddall
I also have an ongoing project going using the Wavefront AL3201, which is now more commonly available as the Coolaudio V1000. I’ve been looking at the SpinFX VF-1 and the Daisy Seed chips as well, probably via the Pedal PCB development boards. I’m way too backed up right now to explore those particular rabbit holes at the moment though.
Yeah I did a bass with an embedded Daisy on the Pedal PCB Terrarium board.
Nice. My situation is kind of weird because my regular job is C developer and I work with hardware on daily basis so I have ton of ideas how to proceed but I would like to use PCB, DSP all in one pack and give my projekt some "user friendly Ux" and I have zero practical experience with Bom choices etc. Daisy looks interesting and my goto for a concept and study. So for now I have to find some development board for Daisy.
- Dec 3, 2003
- 23,049
- 45,913
- Disclosures
- Designer/Toaster Tech @ Passinwind Electronics
Here's what I did with the Daisy Seed: Winter Buildoff 2022 - Outlaw Reborn
Continuing to work on this: SBO 2022 -- Härkä -- it's not a bass but it's pretty bassy (N) using the Daisy Patch-SM
I'd probably recommend the Daisy-Pod as a development platform, as it's pretty easy to add an OLED display.
Glad to help. BTW my day-job is a C++ developer, and I'm doing this as a hobby? 
- Dec 3, 2003
- 23,049
- 45,913
- Disclosures
- Designer/Toaster Tech @ Passinwind Electronics
I am not sure I understand. Is Question mark typo?Glad to help. BTW my day-job is a C++ developer, and I'm doing this as a hobby?
I'm working on an onboard MCU controlled preamp, but with a completely analog audio path (i.e.: no DSP). Perhaps surprisingly to people who know my rate of progress on things, I recently finished soldering all the components onto the prototype board, and made some changes to the next version of the board. Next step: try it out!
No, more of a comment that many would think it strange that I develop software as a hobby when that's what I do all day.
I personally program this interesting stuff after work to relax and have more enthusiasm for all that boring math and physics applications. 
I had an idea of using MCU (MSP430) to control 2 passwind preamps (per pickup) with just 3 knobs (freq, reso, pan). Idea has quickly digressed in procrastinating over using 18650 batteries and analog switches vs digipots. One of the main design goals was to fully power down mcu if controls remain unchanged. Perhaps some time next year
- Dec 3, 2003
- 23,049
- 45,913
- Disclosures
- Designer/Toaster Tech @ Passinwind Electronics
I'm enjoying this tangent, personally. But if it gets enough traction to need its own thread that'd be fine too.
This is very cool in a way. I find very rewarding when I apply some techniques or even entire disciplines learned so far, in my personal live. Your example is inspiring to me.
I had an idea of using MCU (MSP430) to control 2 passwind preamps (per pickup) with just 3 knobs (freq, reso, pan). Idea has quickly digressed in procrastinating over using 18650 batteries and analog switches vs digipots. One of the main design goals was to fully power down mcu if controls remain unchanged. Perhaps some time next year
I like your idea, mostly because I had the same idea myself!
I have built a my own preamp using switches and digipots and can confirm it would be possible to do MCU controlled Passinwind preamps. The digitpots would be the best bet for infinite control, but selection on the 5k pots is somewhat limited.
Power management is actually the thing that took the most time on my design. Shutting down most MCUs is possible (called deep sleep) but that state can generally only be woken from with an interrupt so you need to figure out how to do that. An alternative that I also deploy is clock speed management. When not doing anything useful I dropped the clock speed way down to minimize power draw. Depending on the MCU there can be a LOT of things to control (ex: multiple clock generators, ADCs, flash, timers, USB, etc.) which is why it can take some serious effort.
Hopefully that didn't come across as discouragement, just wanted to be realistic about the amount of effort required to do this kind of thing "onboard".
- Dec 3, 2003
- 23,049
- 45,913
- Disclosures
- Designer/Toaster Tech @ Passinwind Electronics
I like your idea, mostly because I had the same idea myself!
I have built a my own preamp using switches and digipots and can confirm it would be possible to do MCU controlled Passinwind preamps. The digitpots would be the best bet for infinite control, but selection on the 5k pots is somewhat limited.
The new LPF design uses 20K (res) and 50K (freq) by default, but 100K is also fine for the frequency sweep. And a two or three way switch for the resonance function would probably work fine for many people, Alembic certainly thought/thinks so.
Then I'll contribute.I'm enjoying this tangent, personally. But if it gets enough traction to need its own thread that'd be fine too.
I design and program custom digital logic by day. My analog skills are pretty basic, but Douglas Self's "Small Signal Audio Design" book gave me the missing bits.
One thing led to another and I wrote out the draft schematics and got all the parts for a hybrid analog/digital "pedal" of sorts:
Guitar/bass feeds into a 1M op amp buffer (OPA141) with 2nd order HPF @ 31hz and an RFI filter, plus clipping to prevent exceeding the next stage input specs.
That will feed into a TLV320AIC3120 codec chip which does all the dirty work of ADC filtering and pre-amplification and samples at 192kHz, then applies a compressor and noise gate, then 5 stages of fully parametric EQ, then outputs the digital audio.
That digital audio will go to a Pi Pico which applies a waveshaper to create pretty much any kind of distortion you want, maybe also octave up/down, then that goes back to the codec.
The codec then applies 4 more stages of fully parametric EQ, one final compressor, does the DAC conversion and filtering, then drives a line-level output.
The final output buffer (also OPA 141) filters sub/super sonics, and has a selectable gain to bump the output to a max of 1.4Vrms to drive power amps.
It sounds ridiculously ambitious when I write it out, but really the codec does all the stuff I'm no good at, and the rest is just a bunch of software table lookups and filter coefficient calculations.
The interface is the weird part: about four dozen pots and switches, multiplexed and read by the Pi Pico, so it looks/feels like analog gear. I don't want to mess with screen menus or UI software. It's also cost effective, surprisingly.
Last edited:
