TTT

I know I'm not the only one doing this but I haven't been able to find a good description of what the safety issues are. I have a portable inverter, 1 kW/1.2 kWh, and I've brought the lights circuit out of the DB, both L & N, and plugged it into the inverter. It works as expected, and I am pleased. I'm sure it's a code violation, uninsurable, etc. What I want to know is whether it also represents a real safety hazard in my specific situation, and how. The lights were not on the earth leakage breaker. They are all old plastic pendant/bayonet fixtures, no fans etc. I will be the only one changing bulbs. I will be reverting this setup when we move. In principle, I can't see why ceiling lights could not be treated like any other self-contained appliance that could be plugged into a portable inverter. Where is the connection or fault condition that I am not seeing that makes them unlike a freestanding appliance and that makes this unsafe?

There is definitely a shutdown-on-transfer problem with (some of?) these units. Can you reproduce it by switching off at the plug? If not I suspect it could be related to dynamics of how the grid ac goes down, e.g. brownout causing overcurrent trip. If so plugging the inverter into one of those fridge surge & undervoltage protector sockets could be a fix. Otherwise maybe send it back to have the SoC issue fixed and with any luck the shutdown bug is also resolved in new firmware.

There are a couple of photos at the MyBB thread: https://mybroadband.co.za/forum/threads/mecer-sol-i-bb-m1l-inverter-trolley.1218665/page-5#post-30603357

Did a full cycle fix this issue for you? I would probably err towards sending it back regardless as it sounds like you might have the buggy firmware, and that fan issue doesn't sound good either. QA on these units is definitely not their selling point...

Good to hear. The "device" seems to be doing the trick. And it's nice to know the product is being actively supported. (I do still wish we could get more info on the serial interface because I'm sure there's a lot of potential to DIY this thing.) I doubt battery life is affected much either way but it also can't do any good to charge faster than needed. At best there will be more heat dissipation.

Ignore the flashing indicator; if you monitor the power into the device you will see it does stop charging when it reaches charge termination voltage. This is probably why it stalls showing < 100% on the display in the first place: the battery reaches "full" and stops taking charge. Now the charge integrator no longer sees current flow, so it just stops incrementing, regardless of the % it counted to. In your laptop, at this point the SoC display would jump to 100% and you'd not notice it skipped a % or 5. Such a simple fix...

Regardless of what various support people say, I am 99% sure there is no hardware damage or defect, just a charge monitoring & display issue in software. The fix may well be no more than to set new calibration coefficients in EEPROM via some hidden menu. (It would be great if an OEM rep would post the procedure here!) The underlying charge (and discharge) cutout is based on voltage regardless of displayed SoC. When it runs at "0%" this does not mean it's over-discharging the batteries. Conversely when it hits "100%" it has not necessarily finished charging. Unfortunately battery voltage isn't displayed anywhere. So you'll have to take my word for it unless you're willing to open your unit up and monitor it yourself.

That does sound worse than having no indicated SoC at all. I would do a full-cycle capacity check now and then regardless of this issue. As a side effect it should reset the SoC display. But it'll also give you confidence there's nothing actually wrong with the battery. I was pleasantly surprised to discover almost exactly the rated 1280 Wh in mine, about 10 hours at 130 W. Should be double for the 2 kW unit. It'll be interesting to see how these batteries age.

How did it go?

I would also love to know. This is as close as I've got -- the fourth image down looks identical to the mainboard in the SOL-I-BB-M1L and I assume the Kool 1 kWh as well. From Foshan Kemapower. But no evidence they actually designed or manufacture the board, nevermind integrate the device. And I can't find any other version of this trolley so perhaps the overall design is for the SA market only.

This is great news. Why not post the update firmware binary and procedure here? You can distribute the work and hassle of actually patching the affected units... there do seem to be a lot of them.

As far as I can tell the display indicating charging and the display indicating < 100% are connected. They are a display issue only. In fact the system is always keeping the battery at its max voltage. Indeed if you plug it into a power meter you can see spikes into hundreds of W draw every hour for a few seconds at a time. Presumably it attempts to top up the already-charged batteries. (This is at 100% with no flashing charging indicator.) It's not going to damage the batteries as long as the internal max voltage is calibrated correctly... but you can only confirm that by opening it up and measuring.