MGM plugin for PellMon

[eeyore]

Have you tried this on the TCnet I controller? I think it could work with minor changes. At least I only had to make a few changes to the logging software that I made for TCnet I, to make it work with TCnet II.

[motoz]

I don't have access to any of them, Woop opened a tunnel for me to get this far. The plugin works by GET:ing <controller ip>/status.xml and <controller ip>/cgi.cmd from the controller. The URL for the xml file and the cgi 'command' can be changed from the conf. The details were worked out by reading the javascript code from the controllers web interface. The xxx_data part in the mgm.conf file connects a 'pellmon item' to a value in the xml file, so if the TCNET-I controller uses a similar structure (data in an xml file) then you should be able to modify the conf file to fit. What did you modify between the TCNET I and II versions of your logger?

[eeyore]

Yes, it should work then. I basically only changed the way my script phrased information from the xml file.
Is Pellmon able to read inputs from something like a hall effect sensor? Currently I use an Arduino to count auger revolutions and thereby determine pellet usage and pellets remaining.

[motoz]

Yes, on a raspberry you can use the RaspberryGPIO plugin to connect anything to the digital inputs of the raspberry, and there is a plugin (pelletcalc) to calculate consumption, either from a switch that counts auger revolutions, or by timing the motor running time directly. See https://github.com/motoz/PellMon/wiki/H ... let-burner

But you don't need to, the MGM plugin calculates pellet consumption from the reported power. And the silolevel plugin gives you remaining pellet, and a nice silo level graph that extends into the future to show a prediction of when you run out of pellet (based on last years consumption if available, otherwise using a sort of normalized consumption factor based on month of year). See https://github.com/motoz/PellMon/wiki

And you can set an email alarm for anything you want, for instance on "days left in the silo less than something".

[eeyore]

ok, I will have a look at it when I get the time. My experience is that the Power and silo level reported by the MGM boiler is way off. My monitor was able to measure pellet usage within 1% (measured over 3000kg).

When the MGM reports 16kW the real power is only 11,6kW based on my pellet monitor and an estimated efficiency of 90% and 4,9 kWh/kg energy density.

[motoz]

Well we didn't get so far as to check how reliable the reported power level was. If it's that far off then maybe the reported power is based on the current fan speed and not the currently fed pellet/time. I suppose it varies fan speed according to oxygen level, so even if they should be the same on average there might be a big difference momentarily. Hopefully someone can check how accurate it is in practice. Pity the controller doesn't report auger running time.

Btw, for those without a raspberry lying around, pellmon runs fine on a pc with just about any modern linux. On ubuntu/debian or any derived distribution it's easiest to install. Ubuntu in virtualbox works fine, you can even run it on WSL in windows 10.

[eeyore]

Ahh I forgot that the newer boilers have oxygen level sensors. The old ones with TCnet I controller does not have this sensor as standard. So perhaps Pellmon will be accurate with TCnet II.

I recall that I messed with some parameters to get my boiler running stable at low output. Perhaps that is causing the large power deviation on my boiler. Anyway I don't really need that much power, I have never had a daily average power usage above 7kW.

[jimler75]

@motoz - it’s working great!

[motoz]

Great! Can you leave it running so we maybe can get an estimate of how accurate the consumption logging based on the reported power is?

[jimler75]

Great! Can you leave it running so we maybe can get an estimate of how accurate the consumption logging based on the reported power is?

Will do that ... no problem. :-)