SHAC available memory reduces over time

And also see the latest commit at https://github.com/autoSteve/acMqtt. I've added automated HA value templates for sensor decimal places.

And for memory, use keywords like "MQTT, sensor, pn=Memory, dec=2, sa=Automation Controller, unit=%, " to get the value displayed in HA as a percentage.

I like this, @philthedill. Especially AC load. Permanent fixture in my HA now! Thanks for the idea! I love being able to see history over time.

 

thanks - that worked perfectly! I now have CBus voltage, SHAC memory % in use, SHAC processor load working in my HASS system dashboard

 

whoops - I might have got ahead of myself there. I've noted that the voltage unit parameter is not updating in line with the SHAC display. In fact, it appears to update only when the MQTT send/receive is restarted.

 

So here's my AC load over time, @philthedill. 15 min avg. I can't really explain the peaks. I'd be interested to hear if you see similar.

 

Interesting. I see a rock steady constant 29.7 V here, with polling every minute, so can't comment yet. It's been 29.7 for days, so I'm not sure how to replicate this.

When I get really bored I'll stare at the unit parameter for a while to see if it's actually changing... I'm tipping I won't get that bored. And I'm loathe to start adding/removing devices to fluctuate the voltage on the bus, as that would involve opening stuff up that I don't want to open, and adding/removing devices.

If your value flaps about then consider enabling logging in MQTT send receive, and possibly MQTT Final, and see what you see when the value changes. If you want a 'selective'/'surgical' logging option to just target unit parameters then I could rig one up.

(Edit: I've enabled object logging on voltage... Toolkit gives a different and sometimes subtly fluctuating voltage.)

 

Found it. Well, that didn't take much staring... Just at the code... I'm on it, but you get to test it, given my voltage is quite stable.

 

https://github.com/autoSteve/acMqtt the MQTT send receive script. This relies on version 3 of MQTT Final, @philthedill, so if you haven't already then update that script too.

 

looking good thank you. My HASS system page looks like this in the CBUs section:


the voltage is moving +/- 0.05v so far and is effectively stable which means that I can now add an over/under voltage alarm as an early warning that something might be astray. By way of interesting graphs - have a look at my HASS core temperature spike at 3am when my system does its backup!

 

I am seeing something similar:


the peaks are pretty regular and 15 minutes apart - I have nothing (that I can think of) that operates with that period

 

Thanks. Looks high for a 15 minute average. Must be one minute. I opted to not go that route because the resident only fires every minute, and the per min. average shifts about quite a bit.

 

good pickup - 15m is theone to track. now looking more consistent (last 30 mins only on this chart)