search

MrKevinWeiss / riot-10-promo-board

A promotional board for RIOT's 10 year anniversary
GNU Lesser General Public License v2.1
1 stars 0 forks source link

Component selection #7

Open MrKevinWeiss opened 2 years ago

MrKevinWeiss commented 2 years ago

We are closing in on an idea and pretty soon we will need to select the components. The big things to look at will be RIOT support, and availability.

Seeing as the team is ~10, as soon as we are happy with the initial selection I will order 25 of all high risk components (2 per person and 5 for safety).

I hope we don't need to break this into seperate threads (ie, one issue per major component).

We need:

chrysn commented 2 years ago

Do we consider these for workshops and that like? (Like, if any of us were to do a RIOT workshop, would we go there with a bag of kits and encourage people to take one for the list price to use during and after the workshop?) If so, we'd be aiming at quite a few pieces more, and I'd also take a dozen or so to put into our hackerspace's cigar^Wparts vending machine.

MrKevinWeiss commented 2 years ago

I think that would be for the manufacturing stage, I want 25 just for the prototyping. Once all the bugs are worked out then we can make a bigger order (perhaps with some funding).

Once we have something proven then sure, assuming we have some amount of resonable EMC compliance.

MrKevinWeiss commented 2 years ago

From the MCU selection side that supports LoRa natively, it seems like we have:

Out of the 2 I would tend to go with the samr34 as the stm has a cortex M4 and M0+, and in RIOT we are using the M4.

The samr34 also has some battery backed sram which should allow us to save the number of button presses without worrying about constant writes to eeprom or something. It has a RTC and capacitive touch (though I don't know how low power).

To me it is pretty clear of the 2 we should go with the samr34. The other option to consider is using a seperate lora module. This would make it nicer for the modular design then, we would have a seperate lora block. Though I think it could still be modular, the samr34 block would also consume space for a theoretical loraboard, with the same pinouts needed...

luisan00 commented 2 years ago

From the MCU selection side that supports LoRa natively, it seems like we have:

The only drawback I see, is the BGA style packaging in the case of reworking the MCU, but for price, save space on the PCB and for simplify the design I think the best option would be to use one of both and being pragmatic, because there is not stock for the STM32WL, i think the SAMR34 would be the best bet right now

luisan00 commented 2 years ago

We may need some kind of shield over the RF section(and the MCU if it's with LoRa integrated) to block spurious emissions if we need to pass some EMF requirements, it may not be necessary, but it should be considered for budget and PCB design purposes.

The footprint can be made on the printed circuit board and take a decision about the shield once the EMF measurements are made.

MrKevinWeiss commented 2 years ago

but it should be considered for budget and PCB design purposes.

Sure, a solderable tin can type thing!