PDA

View Full Version : Poorman's BMS schematic



Stevo
21 September 2015, 1135
The idea here is to use the first 4 Leaf modules in the pack to control 3 CellogM's, via a 3 pole mechanical relay ( http://www.ebay.com/itm/Potter-Brumfield-KUP-14D15-48-48VDC-Coil-3PDT-Relay-KUP14D1548-NOS-/131451440619?hash=item1e9b1dcdeb ) so that the cellog's hv and lv alarm ports will activate optoisolators. The optoisolators have a normally closed, a normally open, and a common (-) output. The plan is to use the normally closed op to end the charge cycle via a ssr ( http://www.ebay.com/itm/221532661317?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT ) when the hv limit is reached, and to use the normally open op to turn on an amber LED panel warning indicator when the lv alarm signals. This is the reason for the mechanical switch wired in parallel with the ssr to control the 3 pole relay. The Cellogs will only power on when the ignition key is turned on, or when I mechanically turn them on to charge the battery pack.

My main questions are:
1) should I put any fuses anywhere?
2) should I put any reverse protection diodes anywhere?
3) Any other suggestions?

Thanks for looking. I hope the pic below is good enough to see.

6744

DRZ400
21 September 2015, 1205
Just charge to 4.1 volts per cell (set your charger max)....and don't discharge below 3.6v per cell (set your controller min). The Leaf cells are so stable they're boring!! I'm going on 100 cycles....they are within .05v per cell. I checked that all my cells were balanced first by top charging to 4.2v (they all topped off within a 1 minute of each other....yawn). Nissan did a nice job of cell matching for us.

Spoonman
22 September 2015, 0157
I'm unsure of the behaviour of the high and low voltage signals from the CellLogs - are they flags which, once set, remain set until reset; or are they simply status indicators?

If the former, then the above may work effectively, if the latter however then, in the HV cutout instance, as soon as the charger is disconnected, the status will also reset, close the relay and permit charging again - and what you'll wind up with is a looping condition.

In the LV case, the dashboard light will be load dependent, which is fine as you can augment your throttle control to take account of it.

I've thought about using the celllog units in the monitoring and charge cutoff capacity myself but not in quite as agricultural a format as the above - you'd be far better off having the celllogs report to an Arduino or similar, and then using this to program in whatever response it is that you want to any given condition.

podolefsky
22 September 2015, 0659
You don't want to power anything from a subset of cells. It will unbalance the pack. Even if it's only a few mA it will add up if you don't have a balancing circuit.

Athlon
22 September 2015, 0839
Standard Cellog also will unbalance an 8S pack because they get the power from the first 6 cell , if you want to user Cellog as BMS use only 6 cell on each instruments.

Cellog are good fro the price and the voltage readig have a good precision. I did mant thousand of km on cellog based system.
Their weakness are:
unbalancing the pack if more than 6 cell are connected
not optoinsulated alarm port


in your case you shuld build a system abe to stay triggered when the firt alarm trigs , otherwise you will end in a loop fore some time ( at the end the loop will end anyway)

Stevo
22 September 2015, 0934
Thanks all...
DRZ400... That's actually my plan A ! :cool:
Spoonman... I too am unsure how the alarm signal works...it it's momentary or constant... a latching type relay might work if it's only a momentary signal
Noah.... I couldn't find any ssr's that can be controlled by 90 v circuit. I thought the Cellog unbalance problem is when they were wired on all of the time, even when the vehicle is parked with key off. The plan was to only power on the cellogs when vehicle was in use or during charge cycle.
Athlon... Thanks, I can add a 4th cellog

podolefsky
22 September 2015, 0947
Noah.... I couldn't find any ssr's that can be controlled by 90 v circuit. I thought the Cellog unbalance problem is when they were wired on all of the time, even when the vehicle is parked with key off. The plan was to only power on the cellogs when vehicle was in use or during charge cycle.

There are two unbalance issues. One is the Cellogs, as Athlon said. It is worse when they are powered on all the time, but it will still affect you if they are on only when running. Not a huge amount, but it will add up. Unbalance is due to self discharge, and a few mA is about the self-discharge rate.

The other issue is completely separate from Cellogs. It is putting a load only on some cells. Those cells will be drawn down more than others. An SSR won't use much, but like I said even a few mA over time will unbalance the pack.

Why not power the SSR from your DC/DC? It's a 12V source that won't unbalance the pack.

DRZ400
22 September 2015, 1035
Thanks all...
DRZ400... That's actually my plan A !

Why not just wire in your balance leads for periodic checking with 1 cellog? I checked mine for the first dozen cycles....now once a week maybe (I do about 5-6 cycles a week).

Jonny Panic
22 September 2015, 1124
I know there's a mod on endless sphere to get the cell logs to draw equally from all cells.

Stevo
22 September 2015, 1213
Why not power the SSR from your DC/DC? It's a 12V source that won't unbalance the pack.

I had thought of this, but not sure if its ok to keep power on to the vicor when the bike is parked off. I suppose I could put a mechanical switch to control the vicor power supply.

podolefsky
22 September 2015, 1549
I had thought of this, but not sure if its ok to keep power on to the vicor when the bike is parked off. I suppose I could put a mechanical switch to control the vicor power supply.
It's fine. That's how I have mine, vicor power the bms which is always on. My key switch is wired to turn off the vicor via the enable pin when it's in park.

Sent from my SAMSUNG-SM-G850A using Tapatalk