Overdischarged batteries, no grid connection

[kolek]

Battery discharged to 47.63V (average 2.97V per cell) because inverter didn't stop discharging the battery.
(Still trying to figure out the Deye inverter low voltage disconnect settings, if you can help, I made a post about that here).
Cannot charge from grid because grid not yet connected.
After disconnecting load, according to BMS, just the inverter by itself was pulling 1A (so I switched off "discharging" in the JK-BMS
because I don't want the inverter further discharging the batteries all night.
This shut down the inverter, obviously.

Battery: EVE 280AH 16S 48V
Inverter: Deye 8K

Questions (feel free to only answer one or two of the questions):
1. I assume the inverter discharged its capacitors when it shut down, since the shut down occurred at the JK-BMS, so that means I need to recharge the inverter caps again before I restart the system, or would it be OK to just enable discharging mode again in the JK-BMS?
2. @LydMekk posted here to turn off "Activate Battery" to reduce the charge to a trickle if you have extremely low batteries. Is that needed in this instance?
(with no load, cells have recovered to an average of 3.000V).
3. Next time if this unfortunate event occurs, I assume it's better to manually throw a breaker to switch off the inverter from the battery, rather than toggling the discharge to off in the JK-BMS, so that I'm not needing to recharge the caps in the inverter, plus I guess it's bad for the JK-BMS to be switching off discharging?

 

[kolek]

Replying to my own thread, according to this site a 48V battery at 47.63V would have roughly 10% charge remaining so maybe the batteries don't actually qualify as being overdischarged.

 

[Supervstech]

Correct, LFP cells can safely discharge to 2.5v
Most limit to 3v to baby the cycles and extend the cycle abilities.

 

[hwy17]

I spend a great deal of effort in order to avoid ever having to break a DC connection under load, but if given the choice between opening a breaker or disabling a FET I would use the FET.

 

[hwy17]

If the BMS does not have precharge function/sequence the correct procedure would be to disconnect it from the inverter, enable discharge, precharge, and then reconnect.

 

[kolek]

@hwy17 I have the JK-B2A24S20P, pretty sure that can't do a precharge. I'll follow the steps you spelled out, I assumed I would have to.

I spend a great deal of effort in order to avoid ever having to break a DC connection under load

Bad for the FET / BMS to perform that action, right? I assume it does less potential for damage if the load is small as it was in this case, just 1A.

 

[Skypower]

@hwy17 I have the JK-B2A24S20P, pretty sure that can't do a precharge. I'll follow the steps you spelled out, I assumed I would have to.


Bad for the FET / BMS to perform that action, right? I assume it does less potential for damage if the load is small as it was in this case, just 1A.

Yeah. I’d open a mechanical disconnect, then enable the discharge in the bms, pre charge the inverter, now you can close the disconnect.

 

[hwy17]

Bad for the FET / BMS to perform that action, right? I assume it does less potential for damage if the load is small as it was in this case, just 1A.

I don't have a deep understanding of FETs, but I believe they are generally capable of stopping current without damaging themselves.

But yes current defines the damage. I don't worry about 1A at all, I'll open a circuit with any mechanism and not count it as a loaded open for just an amp. 20-200 amp opens are what I work to avoid.

 

[Luk88]

Questions (feel free to only answer one or two of the questions):
1. I assume the inverter discharged its capacitors when it shut down, since the shut down occurred at the JK-BMS, so that means I need to recharge the inverter caps again before I restart the system, or would it be OK to just enable discharging mode again in the JK-BMS?

I'm not sold on the whole "need to connect via a resistor every single time" idea. Yes, it is a good practice when you're unhooking and rehooking your system, but once it has been installed I let the BMS short circuit protection handle the "precharge".

Granted I only have done it twice each for my 8s280ah and a 16s280ah. When I let the battery down until low voltage disconnect (to test capacity) and I then reenable the "short circuit protection" of the heltec BMS immediately disabled the output for 60s. Then it enabled, and disabled again. On the third try it finally charged fully and worked.

With my 48V it took 4 cycles if I remember correctly.

I haven't seen the recommendation to "always" connect via a resistor in a BMS nor in the inverter manual. So I believe it is it not essential. Am I wrong? (don't just say "yes" please justify it if it is so).

2. @LydMekk posted here to turn off "Activate Battery" to reduce the charge to a trickle if you have extremely low batteries. Is that needed in this instance?
(with no load, cells have recovered to an average of 3.000V).

So Leone has already said there is probably a good 5%~10% charge left. I wouldn't worry leaving it like this for a couple of days.

3. Next time if this unfortunate event occurs, I assume it's better to manually throw a breaker to switch off the inverter from the battery, rather than toggling the discharge to off in the JK-BMS, so that I'm not needing to recharge the caps in the inverter, plus I guess it's bad for the JK-BMS to be switching off discharging?

I think the BMS is a better option.

 

[Skypower]

I'm not sold on the whole "need to connect via a resistor every single time" idea. Yes, it is a good practice when you're unhooking and rehooking your system, but once it has been installed I let the BMS short circuit protection handle the "precharge".

Granted I only have done it twice each for my 8s280ah and a 16s280ah. When I let the battery down until low voltage disconnect (to test capacity) and I then reenable the "short circuit protection" of the heltec BMS immediately disabled the output for 60s. Then it enabled, and disabled again. On the third try it finally charged fully and worked.

With my 48V it took 4 cycles if I remember correctly.

I haven't seen the recommendation to "always" connect via a resistor in a BMS nor in the inverter manual. So I believe it is it not essential. Am I wrong? (don't just say "yes" please justify it if it is so).


So Leone has already said there is probably a good 5%~10% charge left. I wouldn't worry leaving it like this for a couple of days.

I think the BMS is a better option.

Not using pre charge resistor (no pre charge circuit in bms) is very hard on the system. Many people have fried FETs ( shorted closed ) blown fuses, blown capacitors and damaged contacts. The inrush is incredible. The only time we don’t do it is with a charger, but there’s a sequence. My son parks his bass boat in the garage and he turns on the 45amp charger FIRST before connecting the Anderson plug to the 305 ah lifepo4 trolling battery.
You are right, I haven’t yet seen any warnings in any of the stuff I’ve bought, multiple bms’s, chargers, inverters NADA!??

 

[kolek]

Is this a safe and acceptable way to do a precharge?
At this point I don't want to unbolt the large copper wires to my inverter.
So:

1. connect a thin (18AWG) wire from the main battery negative to the inverter negative (bypass the JK-BMS)
2. connect the 2 wires of my 200W light bulb precharger (prefer light bulb to a generic precharge resistor because the light bulb comes on and then goes out as the capacitors reach full charge, giving an indicator that you've done your job) to the main battery positive and the inverter positive

 

[kolek]

When I let the battery down until low voltage disconnect (to test capacity) and I then reenable the "short circuit protection" of the heltec BMS immediately disabled the output for 60s. Then it enabled, and disabled again. On the third try it finally charged fully and worked.

My goodness. May work fine but no way am I gonna try that. Last thing I need is to blow a FET in my BMS or screw up my inverter over a minor thing like this. This is like when Clint Eastwood says "Well kid, do you feel lucky?"
Me: Uh, no. No I don't. I never feel lucky.

 

[houseofancients]

Battery discharged to 47.63V (average 2.97V per cell) because inverter didn't stop discharging the battery.
(Still trying to figure out the Deye inverter low voltage disconnect settings, if you can help, I made a post about that here).
Cannot charge from grid because grid not yet connected.
After disconnecting load, according to BMS, just the inverter by itself was pulling 1A (so I switched off "discharging" in the JK-BMS
because I don't want the inverter further discharging the batteries all night.
This shut down the inverter, obviously.

Battery: EVE 280AH 16S 48V
Inverter: Deye 8K

Questions (feel free to only answer one or two of the questions):
1. I assume the inverter discharged its capacitors when it shut down, since the shut down occurred at the JK-BMS, so that means I need to recharge the inverter caps again before I restart the system, or would it be OK to just enable discharging mode again in the JK-BMS?
2. @LydMekk posted here to turn off "Activate Battery" to reduce the charge to a trickle if you have extremely low batteries. Is that needed in this instance?
(with no load, cells have recovered to an average of 3.000V).
3. Next time if this unfortunate event occurs, I assume it's better to manually throw a breaker to switch off the inverter from the battery, rather than toggling the discharge to off in the JK-BMS, so that I'm not needing to recharge the caps in the inverter, plus I guess it's bad for the JK-BMS to be switching off discharging?

Your batteries aren't over discharged.

Activate battery only does something when using BMS-inverter communications, which you're not using.
I suggest you charge your batteries first, maybe by setting to "charge batteries first" instead of load first
But so far you inverter seems to stop the discharge exactly when it should

 

[Skypower]

Is this a safe and acceptable way to do a precharge?
At this point I don't want to unbolt the large copper wires to my inverter.
So:

1. connect a thin (18AWG) wire from the main battery negative to the inverter negative (bypass the JK-BMS)
2. connect the 2 wires of my 200W light bulb precharger (prefer light bulb to a generic precharge resistor because the light bulb comes on and then goes out as the capacitors reach full charge, giving an indicator that you've done your job) to the main battery positive and the inverter positive

Yes that’ll work but think about getting a disconnect/breaker rated for better than just 48vdc in the future.

 

[kolek]

Yes that’ll work but think about getting a disconnect/breaker rated for better than just 48vdc in the future.

Yeah. I’d open a mechanical disconnect, then enable the discharge in the bms, pre charge the inverter, now you can close the disconnect.

I do have a 200A DC disconnect. That part is not a problem. The problem is where do I connect the precharging wires. The main battery terminals give me an easy place to connect. Everything else is covered by heat-shrink / covers / insulation.

 

[houseofancients]

I do have a 200A DC disconnect. That part is not a problem. The problem is where do I connect the precharging wires. The main battery terminals give me an easy place to connect. Everything else is covered by heat-shrink / covers / insulation.

You need a proper BMS that has a pre charge and inverter commumications build in.

 

[kolek]

You need a proper BMS that has a pre charge and inverter commumications build in.

I could be wrong but I thought there are more supporters for JK-BMS on this forum than any other BMS. Obviously that doesn't necessarily make it better. Refresh my memory, which BMS are you using again? Does your BMS do pre-charge and have inverter communications built in?

 

[Skypower]

My goodness. May work fine but no way am I gonna try that. Last thing I need is to blow a FET in my BMS or screw up my inverter over a minor thing like this. This is like when Clint Eastwood says "Well kid, do you feel lucky?"
Me: Uh, no. No I don't. I never feel lucky.

Get in line about bad luck. I have proof that there is a god, because this much bad luck is statistically impossible. Just haven’t figured out what I did to tick him off.

 

[kolek]

You need a proper BMS that has a pre charge and inverter commumications build in.

Lucky @Luk88 has a BMS with pre charge built in. The way it works is you short the BMS 3 or 4 times to the inverter caps, pausing 60 seconds after each short to wait for the BMS to recover, and voila! Precharged!

 

[Skypower]

I could be wrong but there are more supporters for JK-BMS on this forum than any other BMS. Refresh my memory, which BMS are you using again? Does your BMS do pre-charge and have inverter communications built in?

I think the new “inverter” JK has pre charge and many NEW built batteries now have it. I just don’t need closed loop when a 2 amp balancer can do it. Just one less thing to go wrong. It’s been a year and two months since the first two started and have zero balance issues even with a week or so of bad weather not getting to full charge. My batteries spend very little time in the abusive absorb zone. Gitter done and git outa there.