Post by triggerfish on Jan 5, 2019 14:06:39 GMT -8
Like the problems with the AM2315, which had it's driver rewritten to do some more error handling, the SI1145 driver needs some handling also. No write or read to the device is "protected" by a "try/exception" construction. That is what causes zeroing out of the values after an error. The handling on the AM2315 greatly improved that one. Please make this driver error proof also.
Now the AM2315 handling improves, I notice that after every reboot, the SI1145 produces zeros in Blynk (and wunderground). When I manually run the test program and wait for that to fail, the device seems to wake up or so and starts producing light values... The driver definitely needs some work also.
Well, there is, but you may not like it. We spent over a week trying to figure out why the SI1145 will not work reliably on the Raspberry Pi 3B+ under some (unknown) software conditions. We tried everything.
What we finally decided to do is build a board around the TSL2591 light detector and we are testing that prototype while we wait for the boards to come in.