Multi Interface (KNX, UART, I²C, SPI, 1wire ...)

[Eugenius]

Yes, separate

[fluppie]

No specific reason, thought not to "hijack" the BME280 thread. Eugenius, what do you like, seperate BME680 thread?

[Sonnengruesser]

fluppie did you intentionally switch the thread? If not please continue your questions in the other thread or create a new one. Also at the other thread I suggested a change which will help you to get the real time.

[Eugenius]

After reprogramming, device will be resrarted. If not, something gone wrong

[fluppie]

Ok cool, did some changes and I have it working
Working code: https://github.com/fluppie/Temperatu...or_BME680_1.0/

00:02:45.280 -> Start reading BME680
00:02:45.657 -> Finished reading BME680
00:02:45.657 -> currentTemp from BME read
00:02:45.657 -> currentTemp: 22.15 C time: 176 ms
00:02:45.657 -> currentHumd from BME read
00:02:45.657 -> currentHumd: 60.66 % time: 0 ms
00:02:45.657 -> currentPress from BME read
00:02:45.657 -> currentPress: 1007.31 hPa time: 0 ms
00:02:45.657 -> currentGas from BME read
00:02:45.657 -> currentGas: 269.60 kOhm time: 0 ms
00:03:15.267 -> Start reading BME680
00:03:15.648 -> Finished reading BME680
00:03:15.648 -> currentTemp from BME read
00:03:15.648 -> currentTemp: 22.16 C time: 176 ms
00:03:15.648 -> currentHumd from BME read
00:03:15.648 -> currentHumd: 60.86 % time: 0 ms
00:03:15.648 -> currentPress from BME read
00:03:15.648 -> currentPress: 1007.33 hPa time: 0 ms
00:03:15.648 -> currentGas from BME read
00:03:15.648 -> currentGas: 273.63 kOhm time: 0 ms
00:03:45.285 -> Start reading BME680
00:03:45.666 -> Finished reading BME680
00:03:45.666 -> currentTemp from BME read
00:03:45.666 -> currentTemp: 22.16 C time: 176 ms
00:03:45.666 -> currentHumd from BME read
00:03:45.666 -> currentHumd: 60.85 % time: 0 ms
00:03:45.666 -> currentPress from BME read
00:03:45.666 -> currentPress: 1007.33 hPa time: 0 ms
00:03:45.666 -> currentGas from BME read
00:03:45.666 -> currentGas: 270.27 kOhm time: 1 ms

Only the sensor read-out is between 377 and 381 ms according to the serial debug console. While Konnekting says 176ms ...
And I think I found a bug when changing the GA's. Had 1/5/3, 1/5/4, 1/5/5 and 1/5/6 to start, changed it to 1/5/11, 1/5/12, 1/5/13 and 1/5/14. Reprogrammed it and it kept posting to 1/5/3 etc. Only when removing it from the KNX bus power it started posting with the new group addresses. Bug or by design?

[Eugenius]

first of all, IDs must be unique:
https://github.com/fluppie/Temperatu...device.xml#L74
In this line 15 and so on...
Same here
https://github.com/fluppie/Temperatu...config.xml#L73
and other IDs


This will not work:
https://github.com/fluppie/Temperatu...ter/loop.h#L98
I hope you have pressure more then 900

values above 900 is a special error handling in HTU21D library (like, if lib retuerns 900, there is something went wrong).
you can not use this values with BME680...

[fluppie]

When "adapting" the HTU21D code to BME280/BME680, this works fine. But when I want to fetch the pressure and gas resistance from the BME680 the whole thing stops getting sensor data. I made a Github repository with the code I have so far.
https://github.com/fluppie/Temperatu...sor_BME680_2.0

If someone wants to have a look at it, that would be much appreciated.

[fluppie]

OK, I had to mod the HTU21D Example, because they don't compile with 4B

kdevice_Temp_RH.h:
const int KonnektingDevice::_numberOfParams = sizeof (_paramSizeList); // BETA4b do not change this code

mi.h:
byte readMemory(int index) { // beta4b
void writeMemory(int index, byte val) { // beta4b
void updateMemory(int index, byte val) { // beta4b

[Eugenius]

300mA is to much. Out-of-the box it will not work...or not stable.
If a peak short enough, you can try to use a buffer capacitor.
30mA continuous is ok.

Pleas don't use beta4 any more.
beta4b works fine. Some examples are still for beta4 and have to be updated. Share your error message.

[fluppie]

Hello Eugen, after a very long time, I got the modules out of the box to make something new with them. I want to use a BME680 instead of the HTU21D and also a Senseair S8 CO2 sensor. Can the onboard regulator provide enough power for the Senseair S8 ( https://rmtplusstoragesenseair.blob....rat/PSP107.pdf )? It's 30 mA continous, 300mA peak (lightbulb inside the CO2 sensor).
I have one of the first boards you made.

To succesfully compile & upload I had to install beta4 and not beta4b of the device library. Is that normal?

[Eugenius]

Nein, zumindest nicht out of the box. Dafür muss man Host Pin irgendwie beschalten, was nicht gemacht ist...

[Sonnengruesser]

Eugenius ich hab mal wieder eine "lustige" Idee: kann das MI auch als USB Host verwendet werden? Grundsätzlich soll das mit dem SAMD21 möglich sein. Zumindest gibt's z.B. Sparkfun boards die das können sollen, der Arduino Zero aber nicht...

Ich würde gerne 2 Smartmeter-interfaces auslesen, aber keine 2 KNX Geräte dafür kaufen. Und die Smartmeter-Interfaces gibts hauptsächlich als USB-Version.

[Sonnengruesser]

Jetzt muss ich mir überlegen ob ich nicht doch noch 1-2 MIs mehr brache, da die Plattform echt gut ist.

Willkommen im Club

[traxanos]

Das Verhalten ist ja auch korrekt, ich hatte nur nicht damit gerechnet. Ansonsten habe ich meine Samsung TV Steuerung fast fertig. Ich kann per KNX...

• An / Ausschalten (bool)
• Eingang wählen (int)
• Laut / Leister (bool)
• Lautstärke in % (int)
• Kanal Hoch/Runter (bool)
• Kanal direkt (int)
• Remote Control per KeyID (int)

Jetzt warte ich noch auf die fehlenden Klemmen und dann wird das Ganze noch in ein Gehäuse verpackt Jetzt muss ich mir überlegen ob ich nicht doch noch 1-2 MIs mehr brache, da die Plattform echt gut ist.

[Eugenius]

Es ist eine Eigenheit des USB Treibers... und zwar nicht von SAMD21 sondern auch von 32u4 und wohl allen MCUs die einen richtigen USB haben (Nano hat keinen USB im MCU)

Es reicht USB ab und an zu stecken. Manche USB Hub kommen damit auch nicht klar...