MicroPython Forum (Archive)

LoRaTracker wrote: ↑

Fri Jan 12, 2018 9:24 pm

But they have less flash and the same RAM as the micro:bit ?

Oops, I also noticed after posting. I made an edit. Thanks.

I have been thinking of doing a similar shield board for ESP32, but I have a few concerns as the amount of EMI these boards put out versus the humble ATmega, the EMI can reduce LoRa sensitivity. That is probably why the LoPy has two separate shields. Perhaps the really cheap RAK811 module would be ...

Here is a nice project employing the Microchip RN2483 on the Micro:Bit using MicroPython.

@loboris I have seen those. However, from what I can read, these serial modules only implement point-to-point LoRa; not the full LoRaWAN protocol to connect to for example TheThingsNetwork. This makes a huge difference. Personally, I am looking for LoRaWAN solutions that would easily work with stand...

@LoRaTracker You made my day by pointing out the Microchip RN2483 with serial interface and firmware-based LoRaWAN protocol. Taking that route saves valuable RAM heap stack. Eventually, I was lead to the very similar RAK811 module which is sold on Aliexpress with its populated breakout board for a v...

Here might be what we have been looking for.
Someone made a modified version of Pycom's MicroPython (originally written for the LoPy) adapted for any ESP32 with an SX1272 LoRa module. LoRaWAN reportedly is supported.

@Dave You are right about the noise. That is the IEEE 754 graph above. Anyhow, for me the lesson learned is that round() is not sufficient for string formatting; repr(round()) is required. >>> f = 1011.639952 >>> print('%f' % (round(f,1))) 1011.599898 >>> print('%s' % (repr(round(f,1)))) 1011.6 I wi...

@Roberthh Wow, that is truly impressive!

I only hope Pycom understands and appreciates what you have achieved.

Still, there seems to be something wrong…
OK, but this should have the same internal representation:
However, it has an internal representation that is further off.

Yep, you are right, Dave. The same effect can be reproduced with much larger numbers on CPython. I have also been reading about the IEEE 754 standard. A summary of my findings can be found in the now somewhat redundant bug report . Below graph explains very well the difference in floating point arit...