Posts made by rene_2k2

Thanks @korttoma for your feedback ! Looks like the one I use. Did you manage to get this hardware working with MQTT sketch ?

Started some tests to drive W5100 Funduino module SPI_EN pin by a dedicated arduino output as it seems to be the goal of MY_W5100_SPI_EN but after some more failures (a LED on MY_W5100_SPI_EN pin does not blink as I would expect), I realize that I do not see any use of that pin definition in MQTT, only in the EthernetGateway sketch... Am I missing something ?
For those who manage to get their W5100 and RFM69 sharing the same SPI but ( @Snobs , @jpaulin , @korttoma ) what W5100 modules are you using ? Do you know how the Wiznet SPI_EN pin is driven (do you have the module schematic) ?

Hello,
For those like me that have followed the intructions but cannot have their ethernet gateway to work with RFM69, that page may explain the source of the problem : some W5100 hardware modules do not have the logic allowing other SPI device to share the same bus (time to solder...).
Hope that can help others, saving them hours of tests and failures...

What is your hardware configuration ? Looks like the gateway is in a reset loop, no ? Possible power issue ?

Hi, don't know if @Stephan-Noller found a solution, but I had a similar issue with a serial gateway running on a 8MHz arduino : it could not handled serial input at 155200 baud, setting it to 57600 solved my gateway behaviour. What gateway sketch are you using ? What hardware is it running on ?

Just updated the subject so that people with the same issue find it easily.

@etrombly : thank you so much ! Setting baudrate to 57600 on my gateway seems to resolve the issue.

I had just realized there was a problem with gateway not sending packets thanks to the "WITH_LEDS_BLINKING" option but did not realized that my ProMini did not understand serial input at 115200... The tricky part is that it had no problem with serial output at 115200 and it sometimes even parse and process correctly serial 155200bauds input. However setting lower baudrate seems to work always now !

@Francois : Thanks for the answer.
I'll keep investigating then...

Hi,

Sorry for late reply, big storm, falling trees, internet/phone cables on the ground...

@fets The sketches I used on my node is MockMySensors from the examples of MySensors V1.5 (adapted for RFM69), this sketch allows me to test sending and receiving messages. I've also tried others such as TimeAwareSensor, but with the same results...
@tekka I've tried clearing eeprom of node and gateway, but does not seems better.

I now have 2 physical nodes like the ones described by @Francois (Arduino Pro Mini & RFM69W connected by jumper wires, see his post ). The strange thing is that when using LowPowerLab node & gw sketches, it looks like no frame is lost (gw receiving each frame sent by node and all ACK responds OK) [even without caps and antennas when nodes are not far away]. whereas when using MySensors (gw & node or even 2 PingPong), they have difficulties to catch each others data.
When using gw & node, gw seems to get node data but node does not get gw data : for TimeAwakeNode, it often needs to send at least 6 time requests before catching the gw response... This behaviour remains unchanged when swapping hardware. Do you know how I can first determine if this is the node not catching data or the gateway not sending data correctly (in MYSController's debug tab, I see the lines of data which are normally sent by the gateway but can I check that this data is really transmitted) ?

@Francois : you have played with RFM69, don't you have any problems of nodes which do not receive reliably data from the gateway ? (I'm using V1.5)

Any idea ?

Thanks for your answers guys !

@nagelc I think my hardware is OK with DIO0 attached to D2 pin as Node & Gateway sketches from LowPowerLab work (I've just tested once more with ACK enabled and communication is OK in both ways).

@fets You're right : the gateway (at least serial but I think Ethernet too) does not handle ID affectations on its own. However MYSController is a great tool and has an Auto-ID option and also allows you to build whatever message (eg ID response) and to send it immediately or when receiving a message from a node. I see on MYSController's messages tab the ID requests from the node followed by ID responses sent to/by the Gateway (and it loops back to ID-req...).

Hello all,

I'm trying to use MySensors with RFM69W, but I have a problem with the node not having its ID assigned.

I'm using Arduino 1.6.5, MySensors 1.5, SerialGateway on a ProMini and various node sketches (last one tested is MockMySensors) from examples on a Atmega328 on a breadboard. I'm using stable 3.3V sources and have put caps between Vcc and Gnd for both node and gateway.

I have tested sketches (gateway and node) from lowpowerlab sketches with the same hardware and it works (gateway receives data from node, ACK ok).
I also have switched Mysensors gateway and node sketches between my 2 physical "nodes" to check that both can emit and receive data, and the result is the same : I see in MYSController (v0.1.2_282) that the node emits its id-request, then the gateway sends an id-response that never seems to be received OR parsed by the node (nothing shows in the serial monitor of the node).
Using a defined ID in node sketch, that looks better : the gateway can see the node and its children sensors and their values. But if the node requires time, the gateway sends it but the node, once more, does not seem to receive/parse anything.

For those of you using RFM69, did you encounter any problem of data not being received by node ? (When using id=AUTO but eeprom not cleared from previous tests, I did not notice the problem. I had to clear eeprom and use id=AUTO or to use time requests to notice that behaviour).
Any clues on what can go wrong ?

@AWI, @ch3b7
The loop may be caused by a bug in the source code :

 for (int i = 0; i < NoTests ; i++) { 
    [...]
        for (byte j = myTestData[i][3]; i < myTestData[i][4] ; j += myTestData[i][5])
        {
          gw.send(Msg.set(i, true ));

Isn't there a mix between 'i' and 'j' ? Shouldn't it be

 for (int i = 0; i < NoTests ; i++) { 
    [...]
        for (byte j = myTestData[i][3]; j < myTestData[i][4] ; j += myTestData[i][5])
        {
          gw.send(Msg.set(j, true ));

Sorry, cannot test it right now...