RFM95 breakout for M0 second radio

[jhoward88]

There is a picture of breakouts for the M0 on the main site.  See attached.
It shows secondary radio breakouts that look pretty nifty.

I want to add a second RFM95 radio to an Moteino M0.   Are those breakouts actually available?
I can't see where to order them if they are being sold.

Also, I want to make sure it is possible to add a second radio.  I had perhaps made a wrong assumption that the M0 had 2 spi interfaces, one already used by the onboard radio and another that was available on the Miso/Mosi/Sck pins.  When I tried to add a radio using those pins, the M0 didn't like it and wouldn't power on as usual.


Thanks,
Jerry

[Felix]

The only one available is the SDCard.
The radio breakout is shared here, can take HCW or RFM95/96_LoRa: https://github.com/LowPowerLab/Moteino/tree/master/M0

Second radio possible? Sure, the real challenge will be in code. See my response to your other thread.

[jhoward88]

Hi Felix,  Thanks for your replies.  We do appreciate your help and advice.

Sounds like it might be tricky code-wise to have 2 radios on the same SPI.

We have a pretty specific need to have one radio receiving and one acknowledging.  For various reasons we can't use the reliable datagram-automatic ack method, so we have to roll our own.
Due to our peak times which are pretty random, the receiver needs to be listening as close to 100% of the time as we can achieve, so having that same radio stop long enough to do an ack isn't very acceptable.

We have been testing using dual moteinos, one receiving and one acknowledging, with a software serial connection between them.  That actually works pretty well.
But we were thinking two radios on one MCU would be less complex.  Maybe not from what you say.

Thanks,
Jerry

[Felix]

It's interesting you bring this up.

I'm involved in a project where a dual radio solution is considered for this very reason. A SAMD51 processor helps with speeding things up in general, but the radios are bottlenecks and switching TX/RX and intermediary modes cuts off the receiving time which causes lost packets with higher congestion etc.
I would say this whole thing is at a whole different level than the hobby general purpose radio that handles both TX and RX. It's time consuming enough to try to offer (free) support for folks getting started with my library and boards, since even though those boards are optimized for tx/rx-in-1 and they work well when used properly, questions always arise as to "what am I doing wrong" - a rabbit hole without an end.

Anyway!
When you go dual, you have to worry about more things and make sure the concert goes nice and unison. You need 2 antennas, 2 frequencies, 2 RFM69 radio object instances with separate interrupts - easily handled by a SAMD51. The library needs changes to support this. Unfortunately I don't have a ready example to share.
Also, the radios are wasting some power in the signal switches and other circuitry that is on the RFM69HCW for instance, and are optimized for a TX/RX in 1 module with a single antenna port. So in this case a more optimized design would be to integrate the transceivers on the same PCB with separate input and output paths and individual antenna ports.

I would add that going high bitrate will cut transmission times significantly, if you can afford it (will limit range and/or transmission reliability). Also up to 10mhz SPI clock (supported by RFM69/SX1231h) will help with making SPI traffic more expedient and save some microseconds.

[jhoward88]

You are right that a faster processor helps.  So far the dual moteino is looking pretty good.  Yes we do use two frequencies, one for receiving, one for acks.

I'm looking into using a little STM32 like a blue pill for example, since that has multiple spi interefaces.

I'll let you know what we settle on.

Cheers,
Jerry