Daisy board 480MHz Cortex-M7, 64MB SDRAM, 8MB flash, CODEC

I just noticed this Electro-Smith music board design started on kickstarter around Feb 2020, looking like an Arduino special purpose MKR board, this little PCB the same length as but more interesting than the Pi Pico, is fairly potent, it’s 64Mb of RAM stands out.

STM32, 480MHz ARM Cortex-M7 MCU, High fidelity AKM stereo audio codec with up to 24-bit 192kHz, 64MB of SDRAM, and 8MB of flash memory, software DSP Lib

Listed on the sales page for $27.95

They offer a line of expensive accessories the Daisy Seed board plugs into.

Must have spent some of their kickstarter seed money on their trip to NAMM 2020 (video)

The site mentions a 7 day ship delay due to covid, while there is some supply issue discussion on their forum, the AKM AK4556 Codec factory in Japan had a fire. ( (Digi-key has no stock of that part) The codec is a consumer level chip with good S/N numbers.

Wiki . . . Github Code

In a brief announcement in TheVerge there was mention of the Kickstarter track record:
“According to a study run by Kickstarter in 2015, roughly 1 in 10 “successful” products that reach their funding goals fail to actually deliver rewards.”

. - - - - - - - - - - - - - - - - - - -

STM32 Linux Port for Cortex M7 boards possibilities for the no MMU chip?
Some years ago uClinux was a sort of Linux that could run on such a chip (the uClinux site seems to be no more, oh well…)
Porting mainline kernel to ARM Cortex-M7 microcontroller (Project stalled? at 80%)

I think running an OS + kernel like Linux on a microcontroller is missing the point. Layers of abstraction provide tangible benefit on large complex systems but the overhead is excessive for a microcontroller. I understand the thrill of the challenge but doubt it has significant benefit.

The board looks interesting due to its audio i/o but as you point out, there is no guarantee this will survive. Also I am wary of the audio performance with so much HF activity so close to the audio. It would be good to see some data on audio performance.

Their price point for their basic board loaded with goodies reminds me of the $9 crowdsourced C.H.I.;P. computer that made quite an impact at the time but didn’t last.
It would be helpful if there were a preexisting library of functions that could be dropped in, like LV2s, but realistically, I assumed code has to be optimized for the M7 capabilities, just have to hope they have prolific coders.
Those boxes with the board on display may have been a helpful marketing gimmick at first, but having the sensitive bits hanging out in the wind can’t help.

Well, product itself looks nice and at an afordable price. But STM32 programming knowledge is a must if you want to do something more then the given demo code … So I’m not exactly the right candidate for this.

But what is really cool is the fact they share the schematics of all their expensives expansions boards here

I’ve done most of my STM32 programming in C or C++.

They do have an Arduino IDE capability, which makes setting up and doing C/C++ coding about as easy as possible. Here’s the DaisyDuino code library, there are code examples for each of their add-ons, and 60 simple processing modules in the Seed example code.
Some have said Arduino gives 80% of the capabilities of a professional tool chain, without the setup complexity.

Searching some of the examples module names turned up Csound, a massive library of C code music generation functions. Their reference doc’s Orchestra Opcodes and Operators lists some 1300+ functions. Some of these may have been adapted to Daisy examples. (I didn’t see obvious examples utilizing the M7 DSP functionality.)

While off on this tangent, I listened to some of the example music generated with Csound code. (Not part of the Daisy examples, but shows what piecing together C code can do)