Low cost BMS

[harlequin2]

A big thank you to Mark for adding the files to the wiki. It all looks pretty good!
There is one change I made recently that is not shown on those files and that is the value of R7 on the controller is reduced from 2k7 to 2k2 to give slightly better performance with the received comms. It'll still work with 2k7 but the noise immunity is improved a bit (!) with the lower value.

[sawickm]

Originally Posted by harlequin2

R7 on the controller is reduced from 2k7 to 2k2 to give slightly better performance with the received comms. It'll still work with 2k7 but the noise immunity is improved a bit (!) with the lower value.

I corrected the CellTop Master BOM (2k2) R7.

-Mark

[harlequin2]

Yes, one of the reasons that I didn't add current monitoring is that Freyguy's LCD meter thing does all that and I had already built one of those.
It would be easy to add current monitoring, but working out SOC is not quite so easy - but if I stick with Li cells and consequently ignore Peukert, it would work OK.
I'll have a look into what is involved.

[mshalash]

Hi
Where can I find source code and schematics for the master board
Thanks

[harlequin2]

Originally Posted by mshalash

Hi
Where can I find source code and schematics for the master board
Thanks

They are on the open source web site, but if you send me a private message with your email address I'll send you the latest files and comments.

[dcarlson]

Hi harlequin2!

This is great work on a simple BMS. Also a big thanks to you and sawickm (Mark) for putting this all on the Wiki!

I am woking on a project using 12v AGM 15ah batteries for a little scooter. I would love to use this design and see it added to the Wiki for a 12v lead AGM option. Seems that this application could be scaled up to any AGM / SLA (ah amount) series string amount.

You mentioned in one of your posts that all would need to be done is modifying the software in the PIC. Am I wrong but would you not also need a voltage divider on the input to scale from lets say 20v down to 5v to protect the PIC? The charts I have checked out for charging AGM's their CV is around 14.5 - 14.8 volts while the charging amps taper off.

I would love to be apart of making this happen if I can. You all have done some awesome work!

Doug

[harlequin2]

What I had in mind when making that comment about only needing a s/w mod was connecting each module to a single PbA cell ie 2.2 V. The PIC chips work down to 1.8V.
To connect it to a 12 V battery, you would definitely need to drop the voltage down. Dividing by 3 would do it - it only draws a milliamp -and you would need to change the shunt configuration.

[dcarlson]

Wow! Thanks for the quick reply. Okay, this makes sense in regards to dividing the input voltage and the shunt resistors. I am thinking dividing the input voltage by 4. Most AGM's operating range including charging is from 11v to 15v. So, this gives a range once divide of 2.75 to 3.75. Perfect for the operating range of the pic on your BMS!

The shunt resistors are two in series and then parallel? Looks like 1 amp per series string of resistors as the board has ~2 amps of shunting capability? I might be looking at that wrong....

What program are you using to make the schematics. They are *.gbr files on the wiki. I have some time right now so thought I could work on this.

Doug

[harlequin2]

Just happened to turn the computer on for the first time in several days and there was your message! (Its 10:00 am on Monday here).
The shunt resistors are 4 x 1 watt, 6R8 smds connected in series/parallel to provide a single 6R8 resistor capable of dissipating up to 4 watts. They are connected across the module supply terminals in series with a transistor that is turned on or off by the micro.
You would have to re-arrange the circuit so the voltage divider feeding the micro etc is separated from the shunt. Also, feeding say 15V into those shunt resistors would cause lots of melting! ie 33 watts.
The power dissipation of the shunts is something to be considered carefully and the Li cell circuit dissipates 1.9 watts when shunting at 3.6 V. This is about the max that can be tolerated in that small space. For the same dissipation from a 15V battery, you'd need to change the resistors to 120 ohms and then you'd only bypass .125 A which is not much to try and do balancing with. If you want a shunt current of 1/2 amp, then you need 30 ohms and dissipate 7.5 watts, so need 15 - 20 watt resistors! Or a heat sink. So that's what I mean when I say you need to change the shunting.

I use DesignSpark for schematic and pcb files. You can download it free from their web site.
The C code is all compiled with the CCS C compiler running under MPLAB.
I'll be sending Mark some updated files shortly.

[dcarlson]

Thanks again for the speedy reply! Super glad you turned on your machine! Thanks for all the information it is really appreciated. I think this could be a really neat project. But, I would like to keep with your original idea of a simple BMS with keeping the costs as low as possible.

I am going to work on the voltage divider first, then the shunting. Just an FYI, I am jumping into the deep end. I have done a bit of simple circuit design. But, I love the idea of a challenge!

I was thinking it would be good to be able to shunt 1-2 amps max like your original design. Seems like that would be some serious heat at 15v which like you mentioned would need a complete rework of the shunting approach which most likely would include some sort of heat sink.

I took a look at the wiki, you don't happen to have a copy of the schematic file I could muck with do you? Or, maybe if you don't mind providing an editable file to Mark for the wiki?

Doug