Author Topic: RFM66W  (Read 3115 times)

ChemE

  • Sr. Member
  • ****
  • Posts: 419
  • Country: us
RFM66W
« on: July 27, 2016, 04:02:30 AM »
Poking around Hope's website I'm wondering why we are not using the RFM66W.  It appears to have the same main performance stats as the RFM69HW (300kb/s and 20dBm) but a significantly lower receive current of 9.3mA vs 16mA for the 69HW.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: RFM66W
« Reply #1 on: July 27, 2016, 08:20:59 AM »
They come up with new radios all the time. RF Chip makers also come up with new chips with lower consumption all the time.
I was not aware of this RFM66. I see the registers are different, the major improvement would be the RX current, from ~16mA to 9mA (with no LNA boost).
The pinout matches that of RFM69HCW/LoRa radios so from a PCB standpoint you don't need a new revision to change from RFM69HCW/LoRa to RFM66.

To answer your question why - I guess it's because it's probably quite new and there are no libraries for it. Here's a chance to write one :)

UPDATE: also the RFM66 has no hardware encryption apparently, so the RFM69 would be superior in this area and you'd have to do this is the MCU if you wanted encryption, it would be much slower for 8bit MCUs and you'd spend lots more clock cycles and power doing that for the 66W, whereas the 69W this can happen in hardware very quickly.
« Last Edit: July 27, 2016, 10:31:13 AM by Felix »

ChemE

  • Sr. Member
  • ****
  • Posts: 419
  • Country: us
Re: RFM66W
« Reply #2 on: July 27, 2016, 09:50:11 AM »
Thanks Felix.  Too bad they didn't give some even lower power TX options.  I myself am still a long way from  writing a library.  I still need to buy an FTDI from you so I can start playing with my two Motes.  Perhaps these radios will make their way onto R5?

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: RFM66W
« Reply #3 on: July 27, 2016, 10:02:29 AM »
Perhaps these radios will make their way onto R5?

I've vested so much in making the RFM69 happen including the library that it would be a complete remake to switch to a new radio. I didn't study the 66W DS too much to see how similar they are, but I noticed both are based on semtech technology so maybe the RFM69 lib could be ported to 66 without spending a ton of time creating a new library. But I've so many things of priority right now that I cannot see how this will become a reality anytime soon. I cannot do everything by myself, too little time for all the requests, I would hope contributions would make things like this more possible.

ChemE

  • Sr. Member
  • ****
  • Posts: 419
  • Country: us
Re: RFM66W
« Reply #4 on: July 27, 2016, 10:20:44 AM »
I don't mind taking something like that on myself Felix.  I guess most of the reason for my question since I am nowhere near as expert as others is "is the lower RX current and the slightly higher TX current low enough to make these attractive or is the difference trivial in practice".  If the answer is no that is significant I'll probably buy a pair and fool around with porting existing code over to see what can be done.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: RFM66W
« Reply #5 on: July 27, 2016, 10:29:30 AM »
I don't see these as a revolution. The packet engine is similar, but I just noticed there is no AES encryption? So .. there we go, that's a major letdown compared to RFM69.

ChemE

  • Sr. Member
  • ****
  • Posts: 419
  • Country: us
Re: RFM66W
« Reply #6 on: July 27, 2016, 11:37:56 AM »
You're right.  A quick spreadsheet looking at these two radios sleeping for 5 minutes, waking every 3s to listen for a 12-byte packet, and waking every 5 minutes to broadcast one 16-byte packet all at 300kb/s showed the 66W saving 0.7% compared to the 69W.  Across my five minute scenario, the 66W saves 214µC listening for a tiny packet every 3 seconds.  Oh well.  Thanks for the sanity check!
« Last Edit: July 27, 2016, 11:44:09 AM by ChemE »