I can confirm 2.0.5e works on latest raspbian stretch lite (2018-04-18), clean install.
I don't mind using your image, but looking at dates it seems latest one is from March, I assume corresponding version of Evok API on it is also from March? I don't know how to check version of API on the image itself.
Would be an option to build new image after each release of Evok API?
I've also failed to install latest version (2.0.5) on Neuron. I've also tried with clean version of Raspbian Lite Stretch (latest from April 2018 and one from Dec 2017), same issues as reported.
I've also tried UniPi image (2018-03-09 and 2018-03-16-BETA) which works, but upgrade to 2.0.5 haven't worked out.
BTW, how I can find out which version of Evok API is installed on images, is there some REST endpoint which will return release version or I can look somewhere in the FS?
At the end, what is the recommended way of installing latest version of Evok API on Neuron, since everything I've tried is not working?
I had to open 1 wire hub, there was a short circuit between pins 3 and 4 on port 8. After fixing that, everything is working OK.
Port 1 on hub is marked as IN and I have patch cable there connected to 1-wire on Neuron.
Temperature sensors are from the shop, I think they have only 4 pins wired (5-8), and they are connected to ports 2 and 3 of the hub.
So, 3 ports on the hub used (1-3), first 1-wire to Neuron via patch cable and then 2 temp sensors.
Is this correct?
not sure is this Mervis or HW issue, but I've recently purchased 2 temp sensors and 1-wire 8 port hub from the shop. I can't see temp sensors in Mervis if they are connected via 1-wire 8 port hub, but if I connect them directly to Neuron (m203) 1-wire port they do work (one by one of course).
I have connected Neuron to hub via patch 5e cable (cable is checked, working), hub is shielded, on the hub itself it's stated port 1 is for IN, so patch cable is on port 1, two temp sensors are on ports 2 and 3.
I configure Mervis the same way like for single temp sensor:
When I repeat the same with one temp sensor directly attached on 1-wire on Neuron, everything works OK.
I've also tried to add 2 temp sensors and then to assign them, with the same results, via hub it's not working, directly yes.
Do I need to do something else in Mervis, or any other advice?
Thanks for fast answer, the wall switches are actually push buttons (don't know is that the right term), so when I press it it will power on the relay, next time when I press it it will power off (relays are impulse bi-stable ones). Actually wall switch is sending short 24 V DC signal to control the relay. I have what I believe is called central wiring, so all cables from lights and from switches end in central place, and I have 24 V DC towards wall switches, 220 V AC toward lights.
In the first scenario, if I want to control the relay I can send signal either from wall switch or from Neuron via ROs. When I say from Neuron I mean by software which will be used to control the lighting. DI inputs will be used so Neuron knows the state of relay, so I can present which lights are turned on and which are off, in the application for example. But for the light relay it shouldn't make the difference from where the signal is coming it's just react on it.
I hope it's more clear now, I will try it, wanted to make sure I will not damage DIs of Neuron if I have them constantly (when light is ON) on logical 1 (24 V DC).
I have just got my m203 I want to use for my home automation. First is light controll.
In my home I have all the lights controlled via bistable impulse relays, schrack LQ612048. Wall push buttons are all sending 24v DC signal to trigger relay and light on/off. From each relay (all are 2 NO) I also have 24v DC state circuit which I can use to read state of the relay load curcuit which is 220 AC. I get 24V voltage if the relay/light is on, 0 if not.
Can I wire this relay state to DI of the Neuron, it will be permanent log 1 (24v DC) while light is ON?
I also want to use Neuron ROs to controll the relays the same way like from push buttons, the DIs will be used to read the state, while lights will be controlled by both push buttons or Neuron via schrack relays.
Is this a good approach?
The alternative would be to wire push buttons to Neuron DIs, and use direct switch to trigger Neuron ROs, which will then trigger light relays. But in this case I need to remember the state, or use separate DIs to get light relays state.
Hope you understand :)