JBD 7s-21s BMS acting weird

[42OhmsPA]

I'll be sure to report back if I ever have any issues, like what's being reported, with the 4 I have in parallel (1 200A old style with the half open top, 2 200A new style and 1 300A new style).

My complaints about them are poor power readings at low currents and lack of active balancing.

I'll continue to buy them as I add packs. Maybe I'll be eating my words when I add the 5th, time will tell.

 

[42OhmsPA]

View attachment 259074
View attachment 259075
View attachment 259076
View attachment 259077View attachment 259078

So even with a really close zoom in we can still see that it never reached 3.65 volts... And as such should not have triggered OVP

View attachment 259079
As I'm looking through the logs I can't find anything That would suggest there was ever a battery cell over voltage...

I would add values for pack under voltage and increase your undertemp charging to at least 0°C. It also wouldn't hurt to decrease your single cutoff voltage to account for a heavy load dragging down a low pack.

 

[ericfx1984]

Speaking of Chinglish, I have seen MANY BMS claim that they can't be paralleled, but I have always interpreted that as not paralleled on the same battery as there are folks who take that approach when trying to get higher current out of a single battery, and a single BMS is not sufficient.

Wouldn't the correct solution for this be to add an external contactor?

 

[ericfx1984]

I would add values for pack under voltage and increase your undertemp charging to at least 0°C. It also wouldn't hurt to decrease your single cutoff voltage to account for a heavy load dragging down a low pack.

Thank you Good suggestions

 

[fnnwizard]

The real issue is that the 304Ahr pack had a BMS (contactor) disconnect ( we still are trying to find out exactly why) and it would not reconnect until manually reset, which is correct operation for these. It is not a question of parallel packs.

Their weakness is if 1 or the other disconnects, the disconnected 1 would need to see some current to turn back on ( close relay). If current is being satisfied with the 1 that is still connected, the offline one would not see any current therefor wouild stay off until a manual reset.

This is why it is even more imperative to not have BMS disconnect for these.

As Justgary wrote, they can go offline for a few reasons.

At this time, it would be good to see both 280 and 304 BMS charts. We need to find out what caused the 304 bms to disconnect, this is where we need to look.

OP understands conductance of each parallel energy sources just fine and why the battery will charge/discharge at its own value current-wise but V will be very close between both packs most of the time (of course, unless the contactor is open).

 

[justgary]

I went out to winterize the cabin today and the same BMS was off again. Everything in the app shows that it should be working. It was on Tuesday when we were there. It was off today. Surprisingly, this time it has higher voltage than the one that was still on.

I had my laptop but didn't have the correct cable to talk to the BMS directly.

 

[ericfx1984]

I went out to winterize the cabin today and the same BMS was off again. Everything in the app shows that it should be working. It was on Tuesday when we were there. It was off today. Surprisingly, this time it has higher voltage than the one that was still on.

I had my laptop but didn't have the correct cable to talk to the BMS directly.

That's the thing.. in the app it says that charging and discharge are still on... It's an incorrect reporting for sure

 

[ericfx1984]

The real issue is that the 304Ahr pack had a BMS (contactor) disconnect ( we still are trying to find out exactly why) and it would not reconnect until manually reset, which is correct operation for these. It is not a question of parallel packs.

Their weakness is if 1 or the other disconnects, the disconnected 1 would need to see some current to turn back on ( close relay). If current is being satisfied with the 1 that is still connected, the offline one would not see any current therefor wouild stay off until a manual reset.

This is why it is even more imperative to not have BMS disconnect for these.

As Justgary wrote, they can go offline for a few reasons.

At this time, it would be good to see both 280 and 304 BMS charts. We need to find out what caused the 304 bms to disconnect, this is where we need to look.

OP understands conductance of each parallel energy sources just fine and why the battery will charge/discharge at its own value current-wise but V will be very close between both packs most of the time (of course, unless the contactor is open).

Well that's the thing... Did we actually have a contactor disconnect? I don't know according to the app charging and discharging should still be active... And according to the history there was no disconnect

 

[fnnwizard]

Well that's the thing... Did we actually have a contactor disconnect? I don't know according to the app charging and discharging should still be active... And according to the history there was no disconnect

I made that assumption based on your 1st post on delta between 280 pack and 304 pack,, resetting the plug to get 304 working, but it could also have been just soc calc error since we don't have V history of both packs for comparison/confirmation.

The most obvious at the time was ovp on 304 pack. But further data doesn't fully support that yet.

That's why now we need that data from both bms's to determine if anything is actually wrong.

It seems like the 304 pack is working correctly.

But hard to tell since some of the data doesn't line up in my mind. For instance you said charger set to 85A and charging both packs.

304 was taking in 77A so 280 was charging at <10A? While 280 was the lower soc pack?

You may be perfectly correct but to my mind, the timing of events is off somewhere due to timing of posts and incomplete data, so I need to line up that timing with better/more data to better analyze.

 

[justgary]

Well that's the thing... Did we actually have a contactor disconnect? I don't know according to the app charging and discharging should still be active... And according to the history there was no disconnect

I am 100% sure that my issue is that the contactor disconnects for some reason. The app only displays the charge/discharge FET states, but not the contactor. The board has a blue LED that seems to indicate when the contactor is closed, because it is only off when the contactor has opened.

The app does not display an idle time counter either. I'm still looking for a way to disable or bypass the timer. That seems like the most likely cause since the BMS otherwise reports that all is well.

 

[justgary]

The real issue is that the 304Ahr pack had a BMS (contactor) disconnect ( we still are trying to find out exactly why) and it would not reconnect until manually reset, which is correct operation for these. It is not a question of parallel packs.

The manual states that a battery idle for longer than the timeout period (maximum is 65,535 seconds; ~3/4 of a day) will open the contactor to save energy. The contactor has an economizer to hold it open using less current. The discharge FET stays closed so that the BMS can still allow current to flow. When it detects that current is flowing it is supposed to close the contactor again.

In other words, if it is the idle timeout, the BMS should close the contactor upon sensing current. This summer and fall I had a small (~60W) dehumidifier running 24/7 in addition to the ~25W inverter idle current, so the batteries should have been either charging or discharging but never idle. Every time I went there, the one BMS was open again and I had to pull the sense wire plug to reset it. Every time that happened, the app reported complete happiness and acted like nothing was wrong.

I understand how this could happen if the contactor is weak and doesn't hold itself open properly. If that were the case I'm not sure that the blue LED should go off. It should be on if the BMS thinks it is still providing voltage to the coil.

 

[justgary]

And by the way, the manual for my original BMSs (AP20S003S) specifically states:

● When the standing time reaches the set value (page switching time for parameter setting),
the machine will automatically shut down and sleep to reduce standby power consumption.
● Support series connection, but the total number of series after series connection is less
than or equal to 32 series
● Support parallel use, but the total continuous current after parallel connection is less than
or equal to 350A

 

[Quattrohead]

This summer and fall I had a small (~60W) dehumidifier running 24/7 in addition to the ~25W inverter idle current, so the batteries should have been either charging or discharging but never idle. Every time I went there, the one BMS was open again

Not enough current flowing so the contactor opened.

 

[ericfx1984]

Looks like they're both coming up pretty good right now... And the current is dropped down from 85 down to about 55ish split between the two batteries... Just like before the one battery is getting slightly more current than the other. But that is what it is no big deal

The first picture is the 304 amp power battery...

The second picture is the 280 amp hour battery, you can see that the deviation of cells is actually higher on the 280. But still not a big deal





This discrepancy would cause me to believe that this is more of a mechanical issue such as a malfunctioning contactor or bad connection somewhere

 

[justgary]

Not enough current flowing so the contactor opened.

And yet only one of them thinks that is the case, and always the same one.

 

[42OhmsPA]

Looks like they're both coming up pretty good right now... And the current is dropped down from 85 down to about 55ish split between the two batteries... Just like before the one battery is getting slightly more current than the other. But that is what it is no big deal

The first picture is the 304 amp power battery...

The second picture is the 280 amp hour battery, you can see that the deviation of cells is actually higher on the 280. But still not a big deal

View attachment 259937

View attachment 259936

This discrepancy would cause me to believe that this is more of a mechanical issue such as a malfunctioning contactor or bad connection somewhere

That doesn't look to bad. Keep a close eye on cell 16 in the 280Ah pack. Don't be afraid to clamp a 12V filament bulb to it's terminals for awhile.

 

[Suijkerbuijk]

Looks like they're both coming up pretty good right now... And the current is dropped down from 85 down to about 55ish split between the two batteries... Just like before the one battery is getting slightly more current than the other. But that is what it is no big deal

The first picture is the 304 amp power battery...

The second picture is the 280 amp hour battery, you can see that the deviation of cells is actually higher on the 280. But still not a big deal

View attachment 259937

View attachment 259936

This discrepancy would cause me to believe that this is more of a mechanical issue such as a malfunctioning contactor or bad connection somewhere

Looks more on a running cell on nr16.
3.520 volts .
By 3.520 volts charging is relais and charge again.
So that battery is not balance.

 

[ericfx1984]

Looks more on a running cell on nr16.
3.520 volts .
By 3.520 volts charging is relais and charge again.
So that battery is not balance.

It's not that bad... But I do think something needs to be done to keep it under control

 

[fnnwizard]

The manual states that a battery idle for longer than the timeout period (maximum is 65,535 seconds; ~3/4 of a day) will open the contactor to save energy. The contactor has an economizer to hold it open using less current. The discharge FET stays closed so that the BMS can still allow current to flow. When it detects that current is flowing it is supposed to close the contactor again.

In other words, if it is the idle timeout, the BMS should close the contactor upon sensing current. This summer and fall I had a small (~60W) dehumidifier running 24/7 in addition to the ~25W inverter idle current, so the batteries should have been either charging or discharging but never idle. Every time I went there, the one BMS was open again and I had to pull the sense wire plug to reset it. Every time that happened, the app reported complete happiness and acted like nothing was wrong.

I understand how this could happen if the contactor is weak and doesn't hold itself open properly. If that were the case I'm not sure that the blue LED should go off. It should be on if the BMS thinks it is still providing voltage to the coil.

Yes, the current is sensed through the precharged circuit. The manual only states "high current" to re-close that relay. Has anyone tested to see at what point the actual Amperage is for actuation? Unfortunately I didn't for this one, but most of the 100A+ BMS have ~500mA as the threshold.

It's been a while since I tested it, but I thought idle was closer to 15W (290 ish mA@51.2V).
In sleep mode, I couldn't measure any current. How many packs do/did you have when your's went into sleep mode?

 

[justgary]

It's not that bad... But I do think something needs to be done to keep it under control

Check the connections at that cell and its neighbors.

 

[ericfx1984]

That doesn't look to bad. Keep a close eye on cell 16 in the 280Ah pack. Don't be afraid to clamp a 12V filament bulb to it's terminals for awhile.

Would you pull it down to a specific voltage?

 

[42OhmsPA]

Would you pull it down to a specific voltage?

I'd bring it down in line with the rest of them.

 

[ericfx1984]

When I'm doing this should I charge it first to 100% SOC and then disconnect it so that there's no load?

 

[Quattrohead]

And yet only one of them thinks that is the case, and always the same one.

Calibration

 

[42OhmsPA]

When I'm doing this should I charge it first to 100% SOC and then disconnect it so that there's no load?

You can.
I just reduced current and voltage when I had a cell running so the BMS didn't go into protect.