BMS with built-in relay?

[slurrr]

By the way, for each BMS I got a 400A fuse with it (the BMSs are the 200A version). I am unsure why it is, there is no place to mount them.
And of course I already planned to have other lower rated fuses anyway.

The fuse should be attached as close to your main positive as posible. It’s 400a cause the bms can handle 500 - 600a bursts according to the data sheets

 

[Swing]

Yes that was what I already had made. It is probably intended as a nice extra service to include a fuse, but it seems a bit random to me without fuse holder and the fact that I might not want it so high. Oh well, will keep it for future projects.

 

[Nica McNeil]

I was wondering the same thing about that 400 amp fuse for a 200 amp BMS. I will be using a 200 amp fuse instead.

 

[Swing]

I am getting a bit annoyed by now, I can't get the 15th and 16th cells to go away.
I have measured the resistance from the start of the balance circuit for the 14th cell up to the 20th cell, which are now all wired together, and they all read the same resistance value.
In the app I can see the 0.003v for 15th cell and 0.28v for 16th cell, which are not present on my 14s system obviously.

Another thing that worries me is that it seems to reset a bunch of settings depending on the amount of cell detection.
Maybe the limits you set stay that way, and are just multiplied by a different s number whej it automatically sets max pack voltage and so on.

In the app I cannot disable the automatic cell detection. Is this perhaps configurable through pc desktop or through another app?

Edit: Ah, I was misreading the whole thing, from the manual I now get that only the top two (red) wires should go to the highest cell positive (equal to battery positive) and the lower remaining ones should not be connected to the highest positive cell but to the second highest.
So by changing that I got it down to 15s, now it is seeing the 14th cell with 0.002v.
I'll keep on tinkering.

 

[slurrr]

I am getting a bit annoyed by now, I can't get the 15th and 16th cells to go away.
I have measured the resistance from the start of the balance circuit for the 14th cell up to the 20th cell, which are now all wired together, and they all read the same resistance value.
In the app I can see the 0.003v for 15th cell and 0.28v for 16th cell, which are not present on my 14s system obviously.

Another thing that worries me is that it seems to reset a bunch of settings depending on the amount of cell detection.
Maybe the limits you set stay that way, and are just multiplied by a different s number whej it automatically sets max pack voltage and so on.

In the app I cannot disable the automatic cell detection. Is this perhaps configurable through pc desktop or through another app?

Edit: Ah, I was misreading the whole thing, from the manual I now get that only the top two (red) wires should go to the highest cell positive (equal to battery positive) and the lower remaining ones should not be connected to the highest positive cell but to the second highest.
So by changing that I got it down to 15s, now it is seeing the 14th cell with 0.002v.
I'll keep on tinkering.

Read above comments. Wire all the leads you aren’t using to your THIRD highest cell, together with the lead for that cell

 

[Swing]

Read above comments. Wire all the leads you aren’t using to your THIRD highest cell, together with the lead for that cell

Yeah I saw it now, and also in the manual.
The example table doesn't seem correct, but later on there is a table about automatic wiring:
14S BC12~BC18 short, All connect the 12th string of positive
pole of the battery pack

Which must mean, just like commented in this thread, that the 13th cell should go on pin 20.

 

[Swing]

Finally wired it correctly and got it running.
The settings are all pretty straightforward as usual.

Just one setting, the hardware overcurrent protection seems strange to me. It is maxed at 100A

"The hardware discharge overcurrent protection is when the front-end detection chip detects that the current is greater than the hardware discharge over-current protection value, the BMS turns off the discharge switch, stops the discharge, and the general hardware over-current value is 3-5 times of the rated current."

3 to 5 times should be 600 to 1000A for the 200A version. But it is limited to 100A which is a problem.
Unless they mean that 100A will go times 3 to 5...

 

[Swing]

But here are some of my questions about this BMS if anyone could answer I would be appreciated:
1- In the user manual is mentioned that using these BMSs in series or parallel are not allowed? Why ? because there is 900V, 500A contactor ! it is not mosfet based bms that limit you on drain sorce voltage limitation! or sharing equal current of multiple parallel mosfets for parallel connection!
I want to know if someone use it to series up to 4 x 48V , 100A pack still is not allowed to series these packs?

Ok I used this BMS parallel, 2 times a 10kWh NMC block. I didn't like it.
It might look like parallel is better with the relays but now I think it is actually worse. Because a purely MOSFET BMS is contentiously busing with monitoring and possibly limiting current. When the relay is open, it just goes.
Which is fine if they open and close at the same time. The design was as such that the corner cases (low and high voltage thresholds as well as temperature) would rarely occur.
However in real life they were not that in sync. That may have had to do with start up scenarios (inrush current to an inverter)
Anyhow, I witnessed way too much out of sync behavior resulting in balancing currents between the two 10kWh, to often.
I switched to a CAN bus system, monitoring all cells and just have one central activator for opening/closing the relays.
This was more expensive of course, but I will never waste time again trying to parallel with these kind of relay based BMS.

Every occurence of Chinese BMSs being used parallel are always mosfets I think, typically in solutions that are meant to be paralleled. (like server rack batteries etc)
So it is defenitely possible. However, not with this BMS and I have found that the cheap price is causing me to burn more hours into it. So from now on, I will only use higher grade CAN bus BMS for paralleling NMC.

 

[justgary]

Ok I used this BMS parallel, 2 times a 10kWh NMC block. I didn't like it.
It might look like parallel is better with the relays but now I think it is actually worse. Because a purely MOSFET BMS is contentiously busing with monitoring and possibly limiting current. When the relay is open, it just goes.
Which is fine if they open and close at the same time. The design was as such that the corner cases (low and high voltage thresholds as well as temperature) would rarely occur.
However in real life they were not that in sync. That may have had to do with start up scenarios (inrush current to an inverter)
Anyhow, I witnessed way too much out of sync behavior resulting in balancing currents between the two 10kWh, to often.
I switched to a CAN bus system, monitoring all cells and just have one central activator for opening/closing the relays.
This was more expensive of course, but I will never waste time again trying to parallel with these kind of relay based BMS.

Every occurence of Chinese BMSs being used parallel are always mosfets I think, typically in solutions that are meant to be paralleled. (like server rack batteries etc)
So it is defenitely possible. However, not with this BMS and I have found that the cheap price is causing me to burn more hours into it. So from now on, I will only use higher grade CAN bus BMS for paralleling NMC.

Thanks for the report. I am currently building a box for a 16S2P LFP battery that will use two of these BMSs in parallel. I have wired the cells to the BMSs and run a small discharge test with a 100W light bulb as an acceptance test. The bulb drew 0.56A at 52.8VDC (~29.6W), but neither BMS showed any discharge current. I suppose that high current capacity keeps them from having fine resolution.

I do not plan to charge the cells until I have them in a fixture, so I have not put them in parallel. Each battery will have a 300A Class T fuse. I plan to control initial inrush with the 100W light bulb (a great idea that I saw on this forum). My maximum typical usage is about 75A to my inverter, so I'm hoping that the parallel batteries will play well together. Anyway, I'll give a report after I get my ducks in a row.

 

[Swing]

Well, it might work just fine. And if it was my own setup, I eould perhaps tinker along and get it right. But it was for a customer, a drive system. It just needs to work and I didnt have time anymore. Alsp there was 2 meters of wire between them. And perhaps any inrush current alreay triggered them.
I habe 125A ANL fuse on them, and am still using contactors in the same location. Just now controlled via a central system.

 

[justgary]

Idle current measurements in 16S LFP setup: The BMS draws ~45 mA (~2.4W) awake, but with no current on the battery. After about a day with no current on the battery, the BMS will drop the relay (you hear it click) and switch to a draw of about 9 mA (~0.5W). The battery still measures full voltage, so the BMS must also use a FET to provide voltage and sense a load. After a few more days with no load on the battery, the BMS will go to sleep. All the LEDs are dark, the BMS is drawing about 0.9 mA (~50mW), yet the battery still measures full voltage.

Applying a load to the battery (I used a 100W light bulb) causes the BMS to instantly wake up and set the relay. The LEDs are then flashing again. The process will repeat if you leave it without a load again.

Note here that "battery" means the series cells plus the BMS in functional configuration.

 

[Swing]

Well described. Yes there as an extra wire that bypasses the relay, that measures of there is charging applied or power being drawn.
I think it is this behavior that caused me problems when starting again, with them paralleled.

Edit: In the JBD functions menu, there is a "load check" function that you can disable. To influence this behavior.

But you have a much more controlled situations, and more time than I had.

However, economically, wanting to charge my hours to the customer, I think it was a mistakr to go for this BMS. A more expensive centralized BMS costs way less time.

 

[Saxonetter]

Hallo, habe so ein Jbd bms als 16s verdrahtet. An dem Relais liegen nur etwas über 26 Volt an? Es müssten doch über 52 volt Anliegen? Weiß nicht wie ich das Ding starte. Geht das nur wenn Ladestrom fließt? Die bluetoothverbindung klappt auch nicht. Die unterste Aluplatte vom bms Board wird etwas warm

 

[Saxonetter]

Hallo, habe so ein Jbd bms als 16s verdrahtet. An dem Relais liegen nur etwas über 26 Volt an? Es müssten doch über 52 Volt anliegen? Weiß nicht wie ich das Ding starte. Geht das nur wenn Ladestrom fließt? Die Bluetoothverbindung klappt auch nicht. Die unterste Aluplatte vom bms Board wird etwas warm

woran erkennt man überhaupt, dass das Teil gestartet ist, die Spannung an den eigentlichen Batteriepolen liegt z.bzt. bei 52,8 volt

 

[Swing]

@Saxonetter: Better to reply in English.

I think the voltage around 26v is to sense if there is a load or not.
You will need to get Bluetooth app to work, to get further.

 

[justgary]

Hallo, habe so ein Jbd bms als 16s verdrahtet. An dem Relais liegen nur etwas über 26 Volt an? Es müssten doch über 52 volt Anliegen? Weiß nicht wie ich das Ding starte. Geht das nur wenn Ladestrom fließt? Die bluetoothverbindung klappt auch nicht. Die unterste Aluplatte vom bms Board wird etwas warm

Überprüfen Sie Ihre Verkabelung, um sicherzustellen, dass alle Sensordrähte richtig angeschlossen sind.

And I agree that it is much better to reply in English. Google Translate is your pal.