ESP8266 Community Forum

There is no much 10bits devices on the market.
This SP03 doesn't seems to be one, it is only a 7bits device for which the writer of the document mentioned 0xC4 address in "already-shifted" manner ... it is mentioned few lines later that reading register is at 0xC5, which means the R/W bit is already OR-ed. So, the real address of (0xC4 >> 1) is 0x62 (in an "unshifted" manner).

Hi martinayotte,

Yes, I see this now, thank you.
However, addressing the unit at 0x62 (100 dec) still gets no response
Am I right in assuming that this is with a read bit (0) not a write bit (1) so I should be addressing 0x65 (101 dec) to get a response?

George

I don't work with Lua, but I assume that the R/W should be handled by the library itself, so this means you still need to provide 0x62 for the address.

A common technique used in I2C programming is to send a start I2C command with its address and then check the ACK flag and send an end I2C command. If you do this for addresses 1-127, you end up with a map of what devices are present.

When an I2C command containing just the address is sent, a device that is that address will respond with a positive ACK.

This is how "I2C" sniffers work. I would start with that setup and validate that your I2C device is present and at what address you might find it.