question

LiFePO4 batt. You can see the voltage in the screen shot. No switch between the batt and shunt power

That would be a wrong assumption. Examine the screen shot above. Thats why I included it. I am using a LIFePO4 battery. The voltage is 12.97. So no low voltage protection at the time the reporting quit. I am running a JBD BMS, so I have tons of cell data. No low/high cell activation. The cells run very even, with an almost unmeasurable differential. Especially at the low charge/discharge rates when basically at idle. And as a side note, the issue did not occur over percids of heavier use. Zero issues over the coarse of a year with another shunt. But in the few weeks the smart shunt has been in the system, it is outsmarting a lot of folks with issues and no good answers

You can discharge a LFP battery to 10v. Thats how you do a capacity check. The voltage of the screen shot has zero to do with the low discharge rate, and all to do with the time of discharge at any rate, but mostly a higher rate, without a charge. I am trying to isolate the issue to find a common denominator that causes it. But so far, I have not. Only that even though I have kept the phone in front of me nearly continuously, it only happens when I'm not looking

I am not purposely withholding anything. I am frustrated with what should be a simple issue with no answers. I tried to post what I thought was pertinent info. But if I left out something, ask me a question, I'll answer.

The system changed like this. My plan is to convert a hodge podge worth of equipment to Victron. I unbolted the $40 shunt that worked with no issues. I bolted in the $130 Victron smart shunt and something is not right. Tends to make me think it is the shunt. But I am asking what else to check.

A bad connection also occurred to me. That would create heat. I have put the IR gun on every single crimp, connection, battery terminal, and buss bar. Not a single one is over ambient, under load. Not even a little bit. No heat in any of the cells either

My JBD is pretty accurate. If I discharge it 40ah. I can connect a hobby charger and put right at 40ah back into the battery at BMS disconnect. So the BMS nearly matches my charger. And when working, both match the smart shunt very closely. So if the BMS is inaccurate, the charger and smart shunt are all the same wrong

But yet, something aint right

Something is wrong. But I do not think it can be LVD. After having several random losses of SOC, I turned off all chargers, and all loads to see what would happen. The system set with idle draws, which are only 4w! (smart shunt and MPPT LEDs and Rasbery Pi Cerbo with a fan). It ran 4ish days with super close monitoring. Then lost SOC overnight when I wasn't watching.

For the past several hours I have been running a 30amp load trying to hit LVD. Aint done it yet. And it is lower voltage, with multiple magnitude more load than the screen shots above