USB to RS232

• I wanted to start off with some simplicity and then evolve into a full industrial board.
• I used the ATSAMD21G18 this processor is great (in my view) for IIOT applications
• I love the sercom`s and the two internal DAC`s
• I also made it Circuit Python capable (I think)

The other idea I have is to componentize the architecture making easy to change processors or sensor communications  etc. That is why you see the power supplies very separate from the CPU

After proving the concept, I will make it much more appealing to the eye. I like space and lots of test points when proving concepts 

I did evolve it into a concept board already.... To much Coffee! (not produced). Need to prove the basics first.

That get your interest?

Anyone work with this CPU before? I`m having a *** of a time with the boot loader and firmware?

T

• I wanted to start off with some simplicity and then evolve into a full industrial board.
• I used the ATSAMD21G18 this processor is great (in my view) for IIOT applications
• I love the sercom`s and the two internal DAC`s
• I also made it Circuit Python capable (I think)

The other idea I have is to componentize the architecture making easy to change processors or sensor communications  etc. That is why you see the power supplies very separate from the CPU

After proving the concept, I will make it much more appealing to the eye. I like space and lots of test points when proving concepts 

I did evolve it into a concept board already.... To much Coffee! (not produced). Need to prove the basics first.

That get your interest?

Anyone work with this CPU before? I`m having a *** of a time with the boot loader and firmware?

T

Hello Tom,

I haven't used Atmel ARMs for a very long time - I did one project which was successful with a pretty early part (pre- Cortex) but I wasn't thrilled with the support at the time.

I moved over to ST ARM based micros more than 10 years ago and I haven't seen any reason to change.

On a board like this the cost of the processor is trivial - why not use something like an STM32F407 or even an STMH7xx part (400MHz, lots of RAM, 16bit ADC etc etc.)

It looks as if your IO goes direct to the processor - OK for playing but no good for real life in industrial settings.

I have a vague plan for a board for my own use - I'm a bit busy right now but I'll try to set down a few notes and we can see if there is any common ground.

MK

Hi Tom,

Is it for production, or for others to self-assemble? The QFN package may be hard for others.

It definitely needs some isolation, which could be done using (say) opto-isolators or relays. The board I'm working on is constrained in size, so it's only got two inputs and two outputs (using opto and relays in this case), but it's got an option to attach another board. It's described here at high-level, but full details and BoM will be published at some stage. It uses a Pi i.e. Cortex-A, and has the advantages and disadvantages of that, there's a market for Cortex-M controllers as you've noticed too.