Hi BeauJewel. It sounds relevant to me. However I think you should get an electrician to have a look at it. I don't think it's something we could safely advise you on over the net.
I went out to the bush to visit Black Monolith #1 again today. I updated its software and I made it so the customer can (just barely) read the display on the PIP, and push the buttons to change what is displayed.
I made the display readable in a way that preserved the featureless look of the monolith. Over a 75 x 75 mm area in front of the display, I used a cutting disc on a dremel to cut away the _inside_ layer of polycarbonate, and the flutes or ribs. I left the outer layer of polycarbonate intact, except for four 8 mm holes in front of the four buttons. The customer can poke a biro through to push the buttons.
It was a very fiddly job. I'm not sure I'd bother in future.
I also updated the software to eliminate two cases where the State-of-Charge meter would jump to 100% prematurely, and to eliminate some spurious low voltage alarms now that the weather is getting colder.
One case of premature meter reset occurred when the monolith was shutdown and restarted. The coulomb counter in the IMU (the BMS master) is in RAM and is lost when powered down. The CMUs are never powered down since they are on cells, so I made the IMU send the coulomb counter value to CMU-1 every 2 minutes for safe-keeping, then ask for it back again on reset.
The other case of premature jocularity occurred when there was a high charge current from either the genset or the PV array. Coulomb helped me solve that one on the weekend, by reading the code and figuring out where I'd gone wrong.
The cold-weather spurious low-voltage alarms were eliminated by adjusting the estimated internal resistance based on temperature. I looked up the internal resistance versus temp relationship in that great set of slides that Johny found.
http://www.cse.anl.gov/us-china-worksho ... %20BMS.pdf
See page 16.
The cubic equation they give would have been very slow to run with no hardware multiplier. But it turns out it can be modeled quite well as the sum of a fixed component and an Arrhenius component -- a temperature-dependent component that doubles for every 10 K drop in temperature. And at 10 degC, the temp-dependent component is approximately equal to the fixed component.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).