Master Slave BMS Specs Feedback welcome!

Haven't calculated current draw for bms yet (right now it's as a developer kit/collection of small boards).

We made it have a wide VIN so it would work for anything going from 12v (same power source as BMB, to 48v packs)

We do have another board that is a DC-DC from 425V DC to 24V (and/or 12V)
This was made for the aircraft we are building but we weren't sure if there was any demand for them. If there are enough people interested in them we can make a production run of them.

That way you can power the BMS directly from the battery pack and the BMB's.
Is this something that's useful right now they are 600 watts DC-DC (two DC-DC one intermediate to 65v or 70v If I recall correctly) and a low voltage one from HVDC to 24v or 12v.

"I'm a little confused - the specs on the page say "18V to 75V". If it can indeed operate as low as 12V then I'm happier.

Click to expand...

There will be a direct connector for a 12V regulated power and a separate raw 18V to 75V voltage connector.
Preferably a direct connection from the battery.

Regarding the slave boards - can they bridge a service disconnect gap? In other words, if I have a service disconnect in the middle of a group of 6 cells, will there be any problem when it is disconnected? There, of course, would be no pack charge or discharge when the pack is broken so no need to manage the system during disconnect.

Click to expand...

If that happens, you will get an open circuit fault from the balancer/slave board.
If catastrophic failure (IE, short circuit), the balancer will cease to operate because is self-powered from a 6 cell group. This will show up as a missing slave board.

Also, how long can the battery terminal leads be? What kind of wire will you recommend? Do you do anything for de-noising of the voltage sensors?

Click to expand...

Depending on the balancing current, we are shooting from 50mA to 200mA (Depending on the balancing shunt inside the slave board). Anything from 30 to 24 AWG should be sufficient, of course, if you use a longer wire there will be more heat and the balancing current won't be as effective as it should be. I would say no longer than a foot.
As for de-noising the voltage sensor, there is a low pass filter inside the board helping with this aspect. Inside the balancer chip, there is a digital filter for further smoothing

We been working on making a sub 150$ 16 cell BMS.
Was wondering what sort of thermistor mounting setup is more common?

Right now we have the thermistor on a small PCB with a JST connector.
Are there any feedback or preferences for different ways to mount the sensors?

PAXtss said:

Since i am looking at BMS for a 180V 48 taps - each with 4 cells in parallel. I am interested but a little worried about your dissipation amperage due to the 4s cells

Click to expand...

the number of cells in parallel won't have an effect on the heat generated since the system will balance between series. The important part is how many cells you have connected to a single board 6 or 16.

If you have 16 cells, the balancing algorithm to maximize the lifespan of cells won't balance adjacent cells. this will help in not generating an excessive amount of heat. but yes we took it into account on the design amount of watts generated by balancing and board design.

One thing to keep in mind right now we have 140milliamps of balancing per cell. This should be sufficient unless your pack per cell has a lot of kwh/ah.
Since the cells are balanced before soldering in the bus bar and are carefully maintained all the time in balance.

boekel said:

Please go for a dedicated control board, for reliability, etc.

look at competitors, Lithium Balance sells there 8s CMU's for €60 a piece... BMU about €250.

could be hard to compete with that, where could you save on costs?

And please do a -lot- of testing before selling, not good to damage a pack...or start a fire... because of a bug...

Click to expand...

There is a dedicated control board which is the Master BMS board. The slave board picture above only does balancing when commanded by uart or can bus.
Possibly when detected if we add an algorithm to measure voltage rising across cells. But this is not expected on the first version of the software.

Primary reason is that the Shunt/Current sensor needs to communicate to a master board to accurate calculated SOC of your whole pack.

We been running this boards since 2016 for our aircraft conversion battery packs. We also intend to run them thru thermal chambers when we have a final design to make sure everything is running well. We have been designing software and hardware for the past 10 years for the military and server industry at our day jobs.