Old BMS General Thread

[retepsnikrep]

Finished my 50 cell boards

www.solarvan.co.uk/bms/50Slaves.jpg

Still loads to do of course.

Finish cell mountings and straping, interconnect cells, attach slaves and connect them up, program and test each Slave, connect to Master board and start checking Master functions/charging etc on bench. Charge and discharge cells on bench until satisfied working correctly. Finish Insight prep to receive cells, make mountings etc for cells. Install cells and pray!

[qdos]

Very impressed Peter I'm really looking forward to hearing about this all in a working vehicle.

Would be superb to have something in the mag about all this

[retepsnikrep]

Just a schematic update today to reflect the components I actually used in my build. These are the latest and most upto date versions.

I split the Digital and Analogue Slave into two schematics showing parts for each. Both use the same Slave pcb with a few parts in different positions so it's very flexible. If you want to build Digital slaves grab the Digital schematic or vice versa for the Analogue version.

www.solarvan.co.uk/bms/Master080908.jpg

www.solarvan.co.uk/bms/DigitalSlave080908.jpg

www.solarvan.co.uk/bms/AnalogueSlave080908.jpg

Working on my battery mountings at the moment.

[GregsGarage]

Quick update on my progress. Got 16 slaves built at the weekend, 6 to go. Connected 4 of them and charged with a 12v charger. I am going to recommend a small software change for the Analouge On/Off version. What I would recommend is that after the high opto has turned on ,when the cell has reached AbsMaxCellV, that it be turned back off when the cell gets down to the MaxCellV, instead of simply switching on and off based on the AbsMaxCellV value. This will prevent the charger from cycling on and off at a high rate. Instead the high cell will be drained back down and then the charger can switch back on to repeat the process.

I have found that for my 100ah cells this takes 2 1/2 minutes, so I am going to add a simple delay to my managment module for now. A better solution is to change the slave software. I will look at this at some point when time permits, unless someone beats me to it.

[retepsnikrep]

You mean add some Hysteresis I had been thinking about that to be fair, I'll try to add that to the software tonight.

Ooops!!

Well at least it's doing something even if it is doing it a bit badly

What you really want is the charger to continue but with a lower V
output that does not force the cell over AbsMaxCell V, the bypass load will already be operating at this point anyway so the afflicted cell will be receiveing 250ma/hr less than the others.

We can just turn the charger off and leave the load on, but that will greatly extend charging time for the whole pack won't it? as it waits to burn a bit of power off the high cell before turning back on again. If the imbalance is very small then that may be OK.

Can you turn your charger output down rather than off using the AbsMaxCell V trigger as a cue/control signal? Each time the cell Abs Max V opto turns on the charger is turned down one notch until eventually it's off altogether.

[retepsnikrep]

Anyway I have modified the Analogue Slave On/Off software as Greg has requested above, version 1.04 turns on the AbsMaxCellV opto as normal, but only turns it off again when the cell load also turns off.

Version 1.04 runs at 250khz for low current drain. Greg's my guinea pig a bit as well

In order to modify the software I simply moved the command which turns off the AbsMaxCell V opto and put it into the If/Then branch command that turns off the Load. Saved a few bytes as well and program now 67 bytes out of possible 256

www.solarvan.co.uk/bms/SlaveAnalogueOnO ... _v1_04.txt

and this version 1.05

www.solarvan.co.uk/bms/SlaveAnalogueOnO ... _v1_05.txt

does the same thing, but has the 10x adc oversampling removed for a tenfold speed increase, and runs at 31khz for mimimum power consumption. 53 bytes out of possible 256

Greg please try both and give us your views.

Thanks Peter

[GregsGarage]

Wow, that was quick!

I will try both tonight and report back tomorrow.

Thanks for the software update.

Edit: I will change the MinCellV from 200 (2.0volts) to 250 (2.5Volts).

[dillond666]

I got my boards I am linking to the pics...........remember the little boy who posted the resistors through the back of my monitor?

http://www.dillond1.pwp.blueyonder.co.uk/monster.png
http://www.dillond1.pwp.blueyonder.co.uk/pcb.png

Just waiting to get the last few components delivered then I can get building! I decided in the end to make the network cables myself so that I can achieve a neater finish. To that end I have invested in a crimper and a cat5 electronic tester thingy and lots of plugs and boots etc. The fun and games starts now!

[retepsnikrep]

dillond666

They look nice, good luck with the build! Where did you get them made and how much were they in the end? What's you proper name?

Peter

[retepsnikrep]

Edit: I will change the MinCellV from 200 (2.0volts) to 250 (2.5Volts).

Sorry I forgot about that I have now changed MinCellV in the above versions to 2.5V