LoraWan modules for Espruino Pico

Hi guys,

I'm part of a community deploying an open/free LoRaWAN network in Zurich (thethingsnetwork.org) and I'd love to be able to use my Espruinos to connect to it. I got a couple of Microchip RN2483 modules (LoRaWAN transceiver) that provide a serial interface, but I really don't have the skills to wire it up to the Espruino Pico.

Could you please give me a hand on how I should wire it? What I've found is this schematics: https://www.dropbox.com/s/679gjqza5fk5tp­3/LoRa%20QuickStart%20Guide%20(English).­pdf?dl=0 targeted at Arduino.

So, my (likely poor!) understanding of all this is that I should wire as follows:

1. All 13 GND pins on the RN2483 to GND.
   
2. Pin 12 and 34 (VDD) to the 3.3 pin on the Espruino Pico (would it be better to connect one to each 3.3 on the Pico? or both to the same?)
   
3. Pin 6 (UART_TX) to B7 (USART1_RX) on Espruino Pico
   
4. Pin 7 (UART_RX) to B6 (USART1_TX) on Espruino Pico
   
5. Pin 23 (RFH, because I want to use 868Mhz band) to the Antenna? I don't understand the details here in the schematics.
   
   

Once wiring is done, creating a driver for this should be almost trivial. I guess is just a matter of adapting this sample code to Espruino.

Cheers
Gonzalo

Hi @gnz,

I look forward to see projects taking their power both from LoRaWAN and Espruino :-) That is a good thing too, to see such projects in Switzerland.. There is not so many people here to work on IoT .

Your wiring seems OK. For the antenna, you could just solder a wire 8.6[cm] long, which is a 1/4 wavelength antenna. This should work fine. (it is important that the ground pins 22 and 24 of the RN27483 are connected to ground, as they will act as the reference voltage for the antenna.

Once done, the software should be straightforward:

//UART init
var RN2483=Serial1.setup(57600/*baud*/,{tx:B­6,rx:B7});
// RN2483 Reset
RN2483.print("sys reset\r\n");
// your application code...

That'd be great! When it was announced (before the KickStarter) I tried to get in touch and see if I could do anything to add support, but all my emails got ignored so I gave up. Shame, as I'd have bought one of the access points otherwise :)

I'm afraid the link seems broken so I can't see the diagrams, but...

What you're suggesting sounds fine - I would just use the one 3.3v pin on the Espruino though. I'm not sure how much power the module draws, but if it's 150mA or less you should be fine without any external voltage regulator.

Getting a driver for Espruino should be easy. I'd look at using the AT module though - it's generally meant for serial modems, but it handles the whole send data, wait, get result thing really easily (and asynchronously so Espruino doesn't block).

(The module draws a maximum of 38.9 mA when transmitting @ 868MHz: http://ww1.microchip.com/downloads/en/De­viceDoc/50002346A.pdf )

Thanks guys! I just fixed the link (there were parenthesis in the name that messed up the markdown link).

Do you guys have any tips on how to solder these 0.05" pins on the Microchip? I've been trying but is incredibly tricky, especially because the board is not very thick. In comparison, the ones on the Espruino Pico are a walk in the park because the board is thicker and the castellation is deeper.

Is it the only sane approach the use of a PCB on top of which I can solder this? If that's the case, do you know of any that would fit for this chip?

Thanks!

Microchip provides an evaluation board for this module : http://www.microchip.com/DevelopmentTool­s/ProductDetails.aspx?PartNO=dm164138 , unfortunately it is not cheap at all. (60~70$)
if you just want to try the radio on a short distance, let's say, between 2 devices on your desk, you don't need to take a special care about the ground connexions, therefore you can just solder wires between pads.

I don't know if it helps, but @DrAzzy makes these prototype boards that have SOIC pinouts on them as well as space for a Pico - as far as I know, they're 0.05" pitch.

Hi all I too am a member of thethingsnetwork.org have Pico running with Multictech mDot for over 6 months Pico & LoRaWAN are a great combo

Wow, sounds good. Has the mDot got firmware on it that interfaces with thethingsnetwork then?

Would you be prepared to contribute the drivers you wrote for the Pico? I'm sure a load of people here would be interested in that!

@DrAzzy, nice boards! But we would need more SOIC pins, the RN2483 has 47 in total (20 on each side and 7 on the top). Would be cool if you would add those to your tindie store, I pre-order! ;)

My god, that thing looks dreadful to try to solder without a proper breakout board

Do you think there's enough demand that I'd be able to move a 10~12 board order? I've got a lot of designs that aren't exactly setting the world on fire - more like gathering dust (note the two breakout boards I got a batch of 30 - and haven't sold a single one of)

That would have to be a board dedicated to the RN2483 - that's a bizzaro package, and I don't think the same layout could be readily used for other things, particularly with the ones on the end outputting RF (hence layout sensitive).

@DrAzzy at least 10 I'll buy myself, not sure if there are others interested but it is likely because here in Zurich we ordered about 45 of these little guys in the last couple of weeks, some of the more hardware-skilled guys are trying to connect it as is or build their our breakout boards. I don't have the skills to design my own board, so, I'd gladly buy these ones if you have them for sale. :)

@DrAzzy
I'm also interessted. What about selling a complete shield/board for the pico like the arduino one, but as a shield. See gnz dropbox link. I think this will sell much better.
Maybe in a solder your self version.

I'd buy some as well (only maybe 2 though :)

With the Pico pinout, that'd be awesome but I imagine more people are interested in RN2483 than just those with Picos.

I guess a nice-ish option would be to make the pinout on the header such that you could just push it into a breadboard right next to a Pico, without any jumper wires needed?

Then you could always solder a Pico on the back of it if you wanted...

Hmmm....

I was thinking I'd use the same scheme as the evaluation board for the RF pins (one going to a through hole, the other going to an SMA connector. I don't have the expertise to put the antenna on the board, since you need to do impedance matching for that, and that's a dark art in which I am not skilled.

Originally, I was thinking I'd break out the GPIO lines to rows of through holes, and then put Gnd/Vcc/Tx/Rx on one row of pins, so people could easily connect to them. And maybe a spot to put a 3.3v regulator.

If we were to make it Pico-friendly.... the pico would go where? Soldered to the bottom? In that case, there would have to be a second batch of through holes to make the Pico pins accessible (since only a couple of them would be connected to the Pico), since a microcontroller that can only talk to wireless and has no other hardware connected isn't very interesting. But this extra row of pins might be objectionable for people using it with things other than Espruino...

Thoughts?

Usually, any other device than Espruino, such as comm device, EEProm, etc. are the 'guests' on the board, and Espruino is the 'hosting' pat. Witht restrictions of on-boad antennas on RF/Wifi devices though, the (carrier) board is specified according by the RF/Wifi device constraints. Usually, an onboard antenna - ceramic or printed - has to be at the edge of the carrier board in oder to work as designed, as well as there are restrictions in regard of routings around / underneath the RF/Wifi device.

That's the reason that a boad would first cater to the RF/Wifi device and then have additional space to add the hosted boards, of which one would then be Espruino. Accessibility of Pico's USP board part creates additional placement constraints.

With the givens, a stacking - or bottom and top - does not necesserily work. Inline (with marginal overlap whe top an bottom is used) is probably the most reasonable layout. RF/Wifi devices usually have application information about placement options and constraints.

See the development board linked above - note how it does not have any components connected to the RF pins other than an SMA connector and single through-hole for 433. If doing that is not acceptable, it is beyond my design capabilities.

I would really go for the naked board, no on board antenna or anything. The onboard antenna is particularly trick, because it looks nice and tidy but it give a shit range, which is one of the selling points of LoRaWAN in the first place, so I wouldn't bother much, just leave it open, people can solder the antenna the want, or even a piece of wire if they only want to prototype.

What's your opinion on how/where the pico should be mounted?

I'd go for @Gordon 's suggestion of having a pinout that you can push into a breadboard so that people outside the Espruino community can make use of the breakout as well.

Alright, I'll have a go at it, thanks.

Yes - sorry, I didn't mean to actually leave space on the board for the Pico, but to have pins in the right order. People could solder the Pico right on to those pins if they wanted, but it'd hang off the edge of the board.

Which one?

One concern I have with the second option, which at first looks great for plugging into breadboard is that the pins are an inch apart, typically leaving only one row of holes exposed if you wanted both blocks of pins put into pin header. Which means there's no room to have overhang on the sides to fit labels for the pins.

On the other hand, it may be that the GPIO pins are not typically used, at least with breadboard... So what do you guys think?

In first design, I realize now looking at it, I'd put the pins at the edge of the board, with the labels closer in, not the other way around like it is in the first picture.

(obviously, I haven't routed anything yet, and the spacing of the silk is off- but that's a detail)

Just to chime in; I think the first version could make a lot of sense, but maybe alight headers on 100mil grid? I would add male pins to 0-14 and a header to the other one. I could still plug it in a breadboard and use wires to power/serial.

Did you do the math on those co-planar wave guides?
If I were to consider something like this I would need to have a pi-filter in place for antenna matching. With good calculations you can get away with some standard LINX antenna and a 0ohm, but sooner or later someone will want to test various antennas and then it is very handy to be able to match the impedance of that particular one [you know this] :)