BMS heat?

[powersupply]

Can I put a SmartBMS plus battery pack into a quite closed box?

I have this battery box holding 4 SLA batteries to power my scooter.


I might replace those SLA with 13s4p LiFePo.

The battery box is very inaccessible and beneath the foot rest area of the scooter.

I would love to put the SmartBMS into that box with the batteries but don't know about the heat.

Reading some posts, I see that the total heat generated could be below a few Watts.

http://endless-sphere.com/forums/viewtopic.php?f=14&t=49820&start=200

I also see that most battery packs purchased online have their BMS taped into them.

But in my case, all that would be in an enclosure. Which has only two holes, one for the power cables and one for the lock.

Calculations:

Considering 48V/10A charging, the heat generated by the charging shunt (a FET) should be below 1W.

Don't know about the power dissipated by the equalizing resistors, they are SMD 6332, 39 Ohms 1W .

So if we take the worst case scenario that's 12 cells @ 1W, 12W.

Plus the power dissipation of the batteries themselves, no idea how much that is.

What experiences do you have with the heat??

Thanks!

Attached some pictures: one SLA battery plus the battery box(holding 4 SLAs), plus the container (removed from the scooter).

The container plus the battery box form the foot rest area of the scooter.


[/attachment]

 

[dnmun]

never heard of a 6332. if it is 39 ohms and the shunt is at 3.9V then that is 100mA. so I^2xR=.39W

 

[dmwahl]

I don't know the exact power dissipation of that particular BMS, but if it's only a few watts you will barely notice it. Just make sure there's internal room for airflow. Most BMS boards taped to the battery are pretty minimal in terms of balancing, so power dissipation isn't a concern. I run a 16S LiFePO4 pack with my own custom BMS inside an insulated bike bag, balance power is around 0.5W per cell, so 8W total and while the board itself is warm to the touch it's nowhere near dangerous temperatures.

[EDIT] You mentioned the power dissipation in your post, sorry I missed it. So as dnmun pointed out you're below 0.5W per cell.

 

[friendly1uk]

I wouldn't want my discharge resistors sat against a cell. With careful positioning I reckon you could cause yourself problems with localised cell heating. You would almost have to do it on purpose though. With very little thought it should be fine.

 

[dnmun]

the most heat comes from the mosfets during discharge.

 

[powersupply]

6332 refers to the size of the SMD resistor.
So 0.24" by 0.125" or so.

Recalculating , it gives 0.0005W for the FET shunt and less than 12x 0.5W for the bleeding Rs.

How efficient are Li batteries with charging, i.e. how much heat do the batteries emit?

When I charge 12s3p 5Ah cells with 5A (1/3C) I don't notice any heat from them at all.

We are at 6W worst case, a number which looks ok for now.

dnmun, "the most heat comes from the mosfets during discharge" do you mean the bleeding discharge or the actual usage discharge?

Thanks!

 

[dnmun]

if it is a 2512 you should say 2512 since nobody knows what a 6532 is. the mosfets produce heat when the battery current runs through them on discharge.

 

[friendly1uk]

the most heat comes from the mosfets during discharge.

You would think, but I have never felt mine hot. My resistors total something similar to my soldering iron though. Yet look so harmless. Your right though, the obvious bit will get hotter. Don't know what I was thinking. That other voice did say fets, I remember brushing it off, I was so fixated on the unexpected heat from the resistors. The fets must get hotter. They tried to build a heatsink for them with solder lol. The resistors are switched by things so small I'm yet to identify them.

I might try and repeat that performance right away. To the fridge.... lets go

 

[dnmun]

i guess i meant that the risk from heating is that the mosfets go into thermal runaway when they get hot, but the 24S Bestekpower BMS has only 25 ohm/channel and it would be too hot to pick up when i balanced the pack the first time.

 

[dogman dan]

I had problems with a singnalab v1 bms overheating when I smothered it in an insulated rear rack bag in summer. It would start tripping and shutting off.

I solved the problem by just unzipping the bag enough to let a tiny bit of air in.

In a warm climate, I think you should have the bms exposed to the air, and let a small trickle of wind get at it. I like to put a small spacer under the bms, then let the top stay uncovered by anything more than shrink tube.

I've also seen the same bms have it's charging fet get very hot, hot enough to melt shrink wrap.

I've also seen bms come from the factory sealed in a small box. Bet the factory is not in the desert.

 

[powersupply]

Now that I understand that the BMS might produce some critical heat, I am thinking of not including it in the (totally sealed and very inaccessible ) battery box in the belly of the scooter.

The benefits would be:
A) cooler, I could place it beneath the instrument panel (plenty of space) where it can get as hot as it wants.
B) The individual charging LEDs would be visible.
C) no extra connector to the battery box needed for the charger (charger is on board BTW).
D) batteries in the sealed compartement would not be heated up by the BMS.
E) BMS would not shut down due to possible overheating.

There would be a 14 wire, 2m long cable going from the battery (DB25 connector) to said BMS/Celllog.

Except of the added ohms, would there be any other drawbacks of separating the BMS from the batteries??

Thanks for your thoughtful inputs!!

 

[dnmun]

i think you could get away with it in the box if there is air around it. the resistance of the balancing wires is a no op. the problem with long sense wires is the risk of them shorting on something. that is bad, really bad. so try to keep it close.

 

[powersupply]

I feel quite uncomfortable putting the SmartBMS into the sealed box, it would be really hard to get to.

Thought about possible shorts, too.
Was thinking to put 0.33 Ohm (SMD array) in series to the cells (at the DB25).
Must look them up, they have I think .25W or so.

They would be the fuse link (sacrificial fuse).