Parallel diy 48v lfp batteries

[D. Abineri]

What is the general max voltage difference between two battery packs before connecting them in parallel so as to avoid large sparks and worse?

 

[kommando]

I aim for 0.010V diff and had no issues on connection, now at three in parallel. I have the connected batteries under the unconnected battery V in the morning and allow the PV to charge the connected batteries and as the V converges I unisolate the unconnected battery when the V is close.

 

[D. Abineri]

I aim for 0.010V diff and had no issues on connection, now at three in parallel. I have the connected batteries under the unconnected battery V in the morning and allow the PV to charge the connected batteries and as the V converges I unisolate the unconnected battery when the V is close.

Does your software allow for connecting and disconnecting your batteries? I am using JKBMS's on an EG46kv system. Thanks for your assistance.

 

[sunshine_eggo]

I feel like a wildman...

0.2V for me.

Assume 0.2mΩ/cell, 32 cells (2 batteries) = 6.4mΩ total series resistance - ignores EVERYTHING but cell resistance, i.e., bus bars and cables have no resistance (they do).

0.2V / .0064 = 31.25A max theoretical.

As an additional safeguard, the initial resistance will be VERY high prior to fully tightening the connection, so you could even modulate the current flow based on connection tightness.

Whatever the current flow, it will rapidly taper as the two battery voltages converge.

 

[D. Abineri]

I feel like a wildman...

0.2V for me.

Assume 0.2mΩ/cell, 32 cells (2 batteries) = 6.4mΩ total series resistance - ignores EVERYTHING but cell resistance, i.e., bus bars and cables have no resistance (they do).

0.2V / .0064 = 31.25A max theoretical.

As an additional safeguard, the initial resistance will be VERY high prior to fully tightening the connection, so you could even modulate the current flow based on connection tightness.

Whatever the current flow, it will rapidly taper as the two battery voltages converge.

Thank you! On a different topic, what is the benefit to connecting two jkbms,s with their comm ports (can?)? May I just run the bms's on each pack without onnecting them to each other?

 

[sunshine_eggo]

Thank you! On a different topic, what is the benefit to connecting two jkbms,s with their comm ports (can?)?

No idea.

May I just run the bms's on each pack without onnecting them to each other?

If you're asking if you can parallel two batteries with their own BMS at the main (+) and (-) terminals, and the BMSs have no idea the other is there, that's fine.

 

[kommando]

Does your software allow for connecting and disconnecting your batteries?

I have the JK and could use the app to turn on charge/discharge but I also have a manual isolator switch which I can operate much quicker.

 

[ScrodusGargleRotten]

May I just run the bms's on each pack without onnecting them to each other

For sure! 85% of people do that without a problem.

 

[D. Abineri]

I have the JK and could use the app to turn on charge/discharge but I also have a manual isolator switch which I can operate much quicker.

Thanks! So each jk manages the pack it is connected to. Do you have their comm ports connected to each other? Is there a benefit to such a connection?

 

[D. Abineri]

Thanks! So each jk manages the pack it is connected to. Do you have their comm ports connected to each other? Is there a benefit to such a connection?

Thanks! So each jk manages the pack it is connected to. Do you have their comm ports connected to each other? Is there a benefit to such a connection?

 

[ScrodusGargleRotten]

Is there a benefit to such a connection?

Only benefit is when you have the jk bms comms connected to your inverter.
This is called closed loop and has its own good and bad with it.

 

[D. Abineri]

Only benefit is when you have the jk bms comms connected to your inverter.
This is called closed loop and has its own good and bad with it.

Thanks for this.

 

[ScrodusGargleRotten]

Thanks for this.

I have given up on closed loop. It brought more headaches that it solved.



EG4 customer service sucks... just letting you know

 

[D. Abineri]

Good to know, thanks.

 

[ScrodusGargleRotten]

Good to know, thanks.

Only reason I brought that up is because they will tell you it's not their problem if you have a jk bms.

No matter what your problem is, it's because you didn't buy their batteries.

 

[LakeHouse]

On a different topic, what is the benefit to connecting two jkbms,s with their comm ports (can?)?

Unless they're the inverter model, none. I'm not sure they would even talk to each other at all.

I just run the bms's on each pack without onnecting them to each other?

Yes, nothing wrong with this.

 

[doing it different]

Only benefit is when you have the jk bms comms connected to your inverter.
This is called closed loop and has its own good and bad with it.

What's the good? Seems overly complicated....
The opposite of what you want offgrid.

 

[LakeHouse]

I feel like a wildman...

0.2V for me.

Assume 0.2mΩ/cell, 32 cells (2 batteries) = 6.4mΩ total series resistance - ignores EVERYTHING but cell resistance, i.e., bus bars and cables have no resistance (they do).

0.2V / .0064 = 31.25A max theoretical.

As an additional safeguard, the initial resistance will be VERY high prior to fully tightening the connection, so you could even modulate the current flow based on connection tightness.

Whatever the current flow, it will rapidly taper as the two battery voltages converge.

I'd aim for something like this too, but in reality if the batteries are not massively mismatched in capacity (say, a 50Ah connected to a 280Ah), anything in the flat part of the curve is probably going to be fine.

 

[Kornbread]

If you can set everything on the aio based on voltages, IMO, no benefit to going closed loop. When connecting different packs with different voltages together, you can limit the amount of charge/discharge in the jk bms app. As sunshine eggo, I like to get fairly close in voltages then set the bms to charge/discharge around 30amps. If the voltage spread is too far apart, the jk bms will attempt to regulate the amperage, if it cannot regulate the amperage, it will stop, then restart later. It eventually gets close enough to balance voltage between the packs. When connecting packs with different voltages, I've never gone over 30amp charge/discharge, although, if a 30amp setting gives issues, I don't know why one couldn't go higher. Not 200amp high, but maybe something like 50 or so?

With the jk bms, especially the inverter model, it might be better to stay in open loop.

 

[upnorthandpersonal]

With LFP it doesn't matter too much. I've paralleled packs that had a delta of 1V and more and nothing interesting happened.

 

[sunshine_eggo]

If you can set everything on the aio based on voltages, IMO, no benefit to going closed loop. When connecting different packs with different voltages together, you can limit the amount of charge/discharge in the jk bms app. As sunshine eggo, I like to get fairly close in voltages then set the bms to charge/discharge around 30amps. If the voltage spread is too far apart, the jk bms will attempt to regulate the amperage,

Please elaborate on this. I am not aware of any of the common BMSs that will actually REGULATE current. The typical JK, JBD, DALY, etc., do not regulate current in any way. If it's above the limit, open the relevant FETs. and allow 0A.

 

[ScrodusGargleRotten]

Please elaborate on this. I am not aware of any of the common BMSs that will actually REGULATE current. The typical JK, JBD, DALY, etc., do not regulate current in any way. If it's above the limit, open the relevant FETs. and allow 0A.

Closed loop, the bms will tell the inverter.

 

[kommando]

Thanks! So each jk manages the pack it is connected to. Do you have their comm ports connected to each other? Is there a benefit to such a connection?

I run closed loop with battery 1 as the master and it sends average figures to the Sunny Island inverters, but the Sunny Islands ignores any 100% SOC so charges based on V, Chinese AIO,s stop charging at 100% SOC message. Later JK firmware stops 100% SOC being sent until the RCV is reached.

 

[sunshine_eggo]

Closed loop, the bms will tell the inverter.

Right, but claim was made in the implicit context of open loop:

If you can set everything on the aio based on voltages, IMO, no benefit to going closed loop. When connecting different packs with different voltages together, you can limit the amount of charge/discharge in the jk bms app. As sunshine eggo, I like to get fairly close in voltages then set the bms to charge/discharge around 30amps. If the voltage spread is too far apart, the jk bms will attempt to regulate the amperage,

Furthermore, open or closed loop doesn't come into play AT ALL when connecting batteries in parallel to limit current.

So again, I return to questioning if any common JK, JBD or DALY BMS has the ability to regulate current as stated. It is my understanding that this is a very rare feature and the BMS is simply an on/off switch when it comes to the current exceeding spec.

 

[Kornbread]

Right, but claim was made in the implicit context of open loop:



Furthermore, open or closed loop doesn't come into play AT ALL when connecting batteries in parallel to limit current.

So again, I return to questioning if any common JK, JBD or DALY BMS has the ability to regulate current as stated. It is my understanding that this is a very rare feature and the BMS is simply an on/off switch when it comes to the current exceeding spec.

Although I have not verified it with an amp probe, in the app, I've watched the current stop and maintain at 30amps when connecting a battery pack, that was a little off voltage wise, to the main pack. Set the bms to charge at 30amps, and it did. If all the bms can do is wide open or nothing, why are there settings for charge/discharge?