ESP8266 Community Forum

- Fri Oct 09, 2015 10:46 am #30954 are they found at the address you expect them to be? which one is lost first? in the same test? at what speed? Usually in one cycle some devices fail at higher speeds. So the speed where the first device disapears is to high, repeat the test at the lower speeds. Sometimes devices have a minimum timing and won't show up at the slowest speed. Try the speed(s) inbetween these two limits.

if that doesn't help, determination is the way forward: seperate tests. Test the modules one by one and try to find which upper/lower speed it reliably works. Then combine two of them and try the same. This should help you determine a) the best speed for all modules and b) the culprit that is picky about busstates and timing...

There are documents about i2c bus that suggest to "clear" the bus befor the master starts a new communication by toggling the clk xx times; worth a try.

- Fri Oct 09, 2015 10:50 am #30955 Thanks for the info.

How do I force the i2c bus to use lower speeds in the libraries I'm using?

- Fri Oct 09, 2015 11:08 am #30956 wire.setClock(clock in Hz) should make you happy

- Fri Oct 09, 2015 11:20 am #30957 The devices are showing up on the correct addresses. When I run the scans at 100KHz, all 3 are found every time. I went back to my original libraries and added Wire.setClock(100000); both before and after Wire.begin(); and still no joy, even after power cycling :/ I tried 200000 too, with the same results...

So close and yet so far away