Troubleshooting

Thanks I'll try that

@Pewh yes, that's the way. Great work, and thanks for posting the solution!

@Mantvydas-Ri Might be. Data line serves as gnd. But don't operate them like that, things might break!

@hek thank you dear hek

@Christoph-Blank said: well - the blog post I used for the wiring missed the detail that the RFM69 does not seem to be 5V tolerant (inputs). Can anyone tell me: a) if it is damaged if the inputs had 5V b) how I can adjust the levels, can I do this with resistors or do I need more? I made the same mistake, while connecting the modules VIN pin to the Arduino's 3.3V rail, I failed to read that the RFM69's pins aren't 5V tolerant. After I figured out how to make another pair of RFM69 modules work together, this time at the correct voltage, I tried the first to if they were fried. They weren't. As soon as I connected them to a 3.3V Arduino they worked like a charm. So it's very likely that yours are fine to. As for the correct wiring, some are using voltage dividers to lower the signal pins voltage down to 3.3V. Others are using logic level converters. Me, I was lucky engough to have a couple of chinese 3.3V Arduino Pro Mini boards ordered a long time ago and never used since. For minimal clutter, I'd suggest using the same setup. The cheapest 3.3V options are the Arduino Pro Mini (my model) but you'll also need some USB to TTL dongles (the ones having a DTR pin) to program them. Or you may use something like: http://www.ebay.com/itm/Pro-Micro-3-3V-8M-5V-16M-atmega328-Replace-ATmega128-Arduino-Compatible-Nano-/172446438226 but notice that this one doesn't use the ATmega328 chip but ATmega32U4 and that will require these lines of code at the top of your sketches: For MySensors 2.0 #define MY_RF69_SPI_CS 8 #define MY_RF69_IRQ_PIN 7 #define MY_RF69_IRQ_NUM 4 For MySensors 2.1.0beta (RSSI report added amongst a lot of other improvements) #define MY_RFM69_SPI_CS 8 #define MY_RFM69_IRQ_PIN 7 #define MY_RFM69_IRQ_NUM 4 Note the slight difference between the two.

@mfalkvidd Tanks for the help i give it atry. I am new to arduino so maby i need more help. I let you now when it is nessesary. Greatz Ton

@jeti great work!

I think the post @AWI was referring to is https://forum.mysensors.org/topic/1957/lcd-clock-and-text-sensor-node-with-new-v_text/11

Thank you, you made my day! Everything works fine now (told you I was a beginner! :))

It seems that adding: #define MY_PARENT_NODE_IS_STATIC is of some help, however I did noticed a large increase in TX time even though i set boolean requestACK = false; So the whole protocol appears to resemble more with TCP rather than UDP. All things considered I believe the remote battery powered nodes should be able to successfully recover from a temporarily gateway failure without increasing their TX time substantially. The TNR state is excellent when it comes to TX time during gateway failure but it misses a mechanism to automatically recover from this state once the gateway is back. ** MY_PARENT_NODE_IS_STATIC** allows for recovery but at the expense if doubling the TX time. So in production you either manually reset all sensor nodes after a gateway failure, or you take immediate action to minimize the impact on the sensors batteries.

[image: IMG_2762.jpg] [image: IMG_2759.jpg]

@mfalkvidd Thanks heaps, enjoy your trip =)

@Boots33 Thanks for advice! I will definitely get some extra non amplified nrf modules and think about having a separate power supply for my radio.

@tekka: Thanks For All Your Efforts, Buddy ! :) @Mickey: Not Yet, Although I Plan On Utilizing [ https://github.com/tzapu/WiFiManager ] & a Similar Example Is Included There - https://github.com/tzapu/WiFiManager/blob/master/examples/AutoConnectWithFSParameters/AutoConnectWithFSParameters.ino + https://github.com/tzapu/WiFiManager/blob/master/examples/OnDemandConfigPortal/OnDemandConfigPortal.ino

@hek said: You might need to declare them extern so the compiler won't complain. This IS the solution. Thank you for this. To be honest i never used (or knew) extern keyword exists. As a Java-dev, i see this against I believe in : ) Thank you! You made me see the light in the tunnel. (I'm still in the hope its not the train : ) To share the actual solution: In the class, where I was using MyGateway *gw before, now I have this (in the .cpp file): extern bool send(MyMessage &message, bool enableAck) ...and i replaced gw->send(..) with send(..) The imports for the MyMessage class had to be changed to core/MyMessage.h. I'm using visual micro + visual studio for this development.