What would cause battery to lose voltage without discharging?

[fluffball]

When I connect my battery bank to a new AIO, the battery bank loses its voltage. I can charge it back up within seconds so it hasn’t been discharged, just doesn’t show a measurable voltage across its terminals.

I can watch the battery monitor as the voltage goes from 55 to zero in about 30 seconds.

The bank consists of 4 SOK 12v206ah in series. I balanced them and they have passed my tests for capacity (10.5kwh) and can power a 1kW load with just two in series (I tried several combos of just two with a backup 24v inverter).

Could the AIO be shorting the battery bank and the battery bank BMSs are protecting against massive current flow causing this behavior?

If the BMSs behave as such then the AIO is likely faulty, not the battery bank.

 

[fluffball]

I hadn’t thought of this before, but I should test for continuity across the battery input terminals on the AIO. I’ll update thread after I check.

(Edit: there’s no on off so of course there’s continuity—I should check the resistance.)

 

[42OhmsPA]

Did you precharge the inverter capacitors?

 

[fluffball]

Thanks for the reply!

I did not. I followed the shut down/start up sequence that signature solar support recommends for the AIO (EG4 3000EHV-48)

> This is the system reboot order:
To power Down:
1. Turn off Inverter
2. Turn off AC INPUT
3. Turn off solar panels
4. Turn off batteries
5. Turn inverter back on for a minute or so to drain its capacitors. Then turn it back off.

To power on:
1. Turn on the batteries
2. Turn on the solar panels
3. Turn on AC INPUT
4. Turn on inverter

 

[rhino]

Without pre-charging the capacitors you are likely tripping the BMS when the AIO turns on. Some of the server rack batteries have pre-charging built-in.

 

[niktak11]

If it takes 30 seconds after turning it on then it probably isn't short circuit protection. Is it charging the battery as soon as you turn it on? If so then it might be high voltage disconnect. Do you have a way of monitoring the individual cell voltages?

 

[fluffball]

If it takes 30 seconds after turning it on then it probably isn't short circuit protection. Is it charging the battery as soon as you turn it on? If so then it might be high voltage disconnect. Do you have a way of monitoring the individual cell voltages?

That’s an interesting idea but the AIO has no inputs when I connect the battery bank and I think the capacitors are discharged. Seems unlikely that power is going into the batteries.

Sadly I don’t have the capability to monitor individual cells aside from opening up the batteries after they’re drained. If I did check each cell in each battery are you thinking I’d just look for one cell with a different voltage?

 

[Bud Martin]

The capacitor banks are connected to the battery terminals of the AIO, so even if the power switch of the AIO is in the off position the Batteries will still see the capacitor banks.
I would perform precharge (AIO is off), if you do not have resistor, I.E. 25 Ohms 50 ~ 100W resistor you can use 120V 60W incandescent lamp instead.
I think the reason you see the Voltage drops down to 0V in 30 seconds or so is due to the fact that the batteries did try to charge up the capacitor bank then the batteries goes into shutdown so the capacitor bank just discharge due to no more Voltage source present.

 

[TacomaJoe]

My guess is that one of your batteries is shutting down (open circuit) and you are seeing the capacitors discharge rather than the battery discharge.

 

[fluffball]

Thanks y’all!

I did a precharge with a 60w incandescent lamp and it worked like a charm. Pretty neat to see it light up a bit as the capacitors charged.

Conclusion: if you’ve got a EG4 3000EHV-48 with a battery bank of SOK 12v206p, you’re gonna need to precharge or have a backup SCC to avoid BMS shenanigans.

 

[Bud Martin]

Glad to hear that you solved the problem.
The cold resistance of the 120Vac 60W lamp is about 25 Ohms, the hot resistance is about 250 Ohms.