how’s NOV2021 code coming along? Now that I’ve got the shunt working (there’s a Q for later on unless I figure it out myself) I’m about to move the whole bank to the boat, so plan is to install the NOV code (which worked better for me on the last month tests).
Wondering if I should get it right now, of if there’s lots that you’re going to do over say the w/e. I can always get the latest and reflash only wont be extremely fast on board with a 10yo laptop…
I’ve been running it for just over three weeks with only a couple issues, namely unexplained reboots. @stuart has been working on tracing down the cause of these as he has also observed a couple of them. But otherwise it has been pretty solid.
thanks Mike, sorry asked a very vague question. I know it works, actually much better for me than the main branch, so using it now for over a month on my tests.
What I meant to ask is if the plan is to replace the main branch soon.
I believe that is the plan still, but it will require sorting out a few lingering issues (such as the reboots) and more extensive testing of extra functionality. But I’d say it is fairly close to merging to master.
Think the nov2021 version is almost complete, so I’d use that one.
Is that the least rebooting version? I finally just moved from the old controller to ESP32 and it just reboots itself in less than 12 hours or so. I tried a separate power source directly to the board and an old option as well, still the same. While old controller had rebooted involuntarily maybe 3 times in 6 months.
Thinking to try to switch relays to “on” by default so that it would not cut the power source while sailing during reboot
Owch. I’d look at a possible memory leak there.
Frankly I’m not convinced that ESPs are reliable enough for professional use. I don’t care whether the thing reboots once per 6 months or once per ten minutes – the only correct number here is zero.
… thinking about designing my own controller … current monitoring with Hall effect sensors, voltage monitoring by the modules, CAN, Modbus RT, and serial TTL to talk to a separate host for visualization, Home Assistant integration and whatnot instead of WLAN … now if only the week had enough days in it …
fwiw, NOV21 branch is more stable for me (running influx and victron integration)
The NOV21 version is much more stable, please try that one.
fwiw, just managed to wire everything up onboard and do an 1h 75A draw test out of a fully charged 304Ah 24V lifepo4 bank.
well impressed, current values closely matching victron shunt (measuring the combo of the old FLA Trojan bank and the lifepo4) system didn’t miss a beat, resulting values in tune to discharge levels.
Well done to all!
influxDB 1.8 data, one hour constant watermaker running plus fridge and lights:
still temp install, got to figure out where I hang the controller and shorten the RS485 cable to the victron…