Troubleshooting

Hello, I teste this solution out and it worked :) My W5100 is now working! Use as reference the post https://forum.arduino.cc/index.php?topic=351477.msg2662072#msg2662072

@miro it doesn't make much difference which you use. I use 4.7uF. Some users combine one large and one small capacitor, for example 4.7uF and 470uF. The smaller is better at eavening out quick fluctuations while the larger can handle more powerful fluctuations. The capacitor needs to handle the same voltage as the power source. The ones I have are specified for 50V but as low as 6V should work.

@jacek By the looks of your code you're doing the following. Domoticz sends a command to your Node. You ignore the value and assign 0 as the value. You send the new value - which is allways 0 - to domoticz. So basically you turn off the lamp no matter what the command is what Domoticz is sending to the node. At least that's what I think. But I'm a bit tired have been studying for too many hours today ;-)

i'm using the latest development branch... I believe there's no void presentation? edit: crap, noob error, I missed my presentation loop!

Glad it helped David. I just thought about it again, maybe sleep(0xff, 0x00, 0) does work too. I did not test it but I think it calls the same function anyway. However the method of direct calling sleep(0xff, 0x00, 0xff, 0x00, 0) should save some clocks because one less jmp is required :)

Seems to be fixed. I cleared the EEPROM and reloaded the firmware.

@Reza you should show the output of the Serial Monitor. There may be a variable overflowing...

You have to add a timer, and an if statement. Start the timer after motion is tripped. If time is less than 30 s, ping. Else reset timer and sleep with motion interrupt. I can post some example code later perhaps.

@TheoL :'( thank you my dear friend

Yeah, @TimO Is right. Check the code example here http://www.mysensors.org/build/motion and look for the gw.sleep sentence. I have a similar sensor and this is what I have in my scketch: gw.sleep(INTERRUPT,CHANGE, SLEEP_TIME); The problem with your sentenc is that the sleep command is not overridden by the motion detection. You need to include the interruption as part of the sleep command.

@hek I did (Not that I don't trust you), and you're right. Go get yourself a beer and pat yourself on the back ;) Now if I could just get my ESP8266 to actually not reply to 50% off messages with st=fail. I've only been trying to get this all working for 3 weeks (sigh).

Here are some scientific messurements: http://cc.oulu.fi/~kmikhayl/site-assets/pdfs/2012_NTMS.pdf Table on last page says: SoftSPI consumes about double the electricity of a Hardware SPI on a PIC system. So, desoldering the LED is the thing to do. Thx for the hint.

@robertober This is probably due to a long time (> 1s) spent in setup() due to which the watchdog timer is resetting the chip. Check this page for more info: github-issue-WDT

Great thanks :)

A simple thing you can do in RFM69Transport is to increase the retry count for the messages that are sent. The default value is 2 (implicit), to increase this by changing the following: return _radio.sendWithRetry(to,data,len); to return _radio.sendWithRetry(to,data,len, 5); To have it retry five times. My guess is that this will greatly increase the operation time of your network.

Yeah, the debug trace shows that the node had an ID of 0. How can this happen (since the node has been functioning well for months before)? I flashed the EEPROM and reinstalled the sketch, after which I was able to include it with a valid node ID.

@Cliff-Karlsson No sure on the button force sensors, I have not used them. To test the force sensors under a mattress, I would wire them up to an Arduino with a sketch that simple reads the analog pin and serial prints the result. Place the sensor under the mattress and see what the serial monitor readings are when the bed is unoccupied and occupied. This would give you the unique range for you particular need. Might be tricky if your computer and bed are in different rooms. After you have the range, you could write a sketch with S_BINARY to trigger V_STATUS whenever the reading approached the high end of your tested range. If you cannot reach you bed while connected to serial, you could write a sample mySensors sketch to send the analog reading every few seconds to your controller as V_LEVEL. This video helps explain how to wire up the sensor and select the static resistor for the voltage divider. video

Hi @GertSanders Thanks for the help. Finally had some time yesterday to mess around and started fresh with the .zip you included. Works like a dream now. I had to disable BOD to flash @ 1MHz; but it works flawlessly. Thank you so much again.

Last night I found that I could put my thumb on the sensor's antenna and the sensor would then work flawlessly. After switching radios on both the sensor and gateway without luck, I reduced the power on the sensor. This has seemed to fix the issue for now. By the way, I'm using a separate 3.3v supply, capacitor on the radios - the works.

@codergirl56 Reading your question, I think your starting point is to have a look at the pingpong example (see below for the direct link) which comes with the mysensors library and try to modify it such that it starts sending the RFID data. direct link to pingpong example First you'd need to install the mysensors library of course, see the download page. How to do that is in the "Getting Started" as mentioned by @AWI. Just start reading and you'll be guided to it. good luck Boozz