Found out that EMC was not an issue, but the LED power supply. I added a 10k resistor in parallel to the 0-10V input which fixed the 0V problem.
Only poor thing now is the bad behaviour of Mervis with the AO extension board. I added a ramp for smooth dimm up and down, but it's actually not smooth but rather stepwise:
Any idea how to fix? Maybe the problem comes from the transformation "to_uint(real)" ?
@naibaF Have you tried connecting just one Sedtronic unica sensor directly to UniPi with proper connection from the previous link?
You can of course connect temperature sensors and other sensors to the same network...
Hi, for the first generation of UniPi there are still a few pieces left of http://unipi.technology/product/4-analog-output-extension/
But we are in the final stage of development of AI/AO extension for the Neuron series which will be available in 3-4months.
Those are the 12 PWM outputs. There is also a schematic how to connect it all togetger below
The IN connectors should be connected to 5-50V DC, the LED is connected as openn drain - https://www.maximintegrated.com/en/glossary/definitions.mvp/term/Open-drain/gpk/1174
This table shows the description of the pin header.
Unfortunately this is not possible. The DIs are connected direclty to GPIOs of RPi and besides that the i2c devices have the same address on both UniPis so they colide...
Get one more RPi and do it on the SW layer.
Unfortunately 1wire is not a way to go. It is too slow and the IC does not support pwm.
If you are a little bit hw person, take a look at the 4ao/12pwm for i2c board.
Btw the Neuron supports RS485 (Modbus) and there are many boards for dimming based on rs485.
I was looking at this screen: http://0.0.0.0/rest/all
and the address for the temp probes shows up, as well as the same address as the example file. BUT: I just noticed an address for the unit.
All working now, Many thanks!