BMS Wiring for LiFePO4

[redpoint5]

It's not a dead thread. I'm just now getting around to putting a LiFePO4 in the glovebox of the Acura.

The lithium ion battery I tried had too much initial voltage after being charged to full (16v) and was frying the LED balancing on my Ultracap bank, which in turn caused them to go out of balance and over-voltage. Apparently it didn't like going low to the 13v of the alternator, so the battery swelled and went crazy out of balance (probably a fire hazard).

Back to the LiFePO4. I always make the mistake of running the battery too low while I tinker with things or measure loads of various things, so I harmed it to some degree, though balance charging seems to have largely recovered it (slightly puffed, then went back to normal, and voltages are in balance).

I ended up buying a $10 balance circuit from Amazon that forgoes any of the other BMS functions. Of course, whatever connector it has is slightly smaller than the balance connector on the battery, so I've got to build an adaptor. In theory it harvests over voltage (threshold 0.1v and feeds it into the undervoltage cell, and then goes into low power sleep.

Next I need to see if this circuit can handle ultracaps. From some of the reviews it sounds like it might fry itself. I'd rather connect something like this to a cap bank than build a diode/LED circuit for every cap.

Amazon

[Stubby79]

I have a few of those waiting to be hooked up, though I did plug one in to see that it worked. Which it did.

Yes, they will shuffle around the "extra" voltage until things are even...within 0.05V or something.

I don't see why an ultra cap would fry it. I had it hooked up to lithium cells - high output, low internal resistance - and those things can kick pretty hard. Not ultra cap hard, but still hard. Not like, say, lead-acid, where there's a certain level of...resistance (not electrical, more like chemical reaction time) that things like simple chargers take advantage of to prevent too much current being drawn.

They're inductive...which leads me to assume they're like little transformers, which allow the output to be isolated from the input, otherwise things start frying when you're trying to pass power from one cell that's already connected to another cell. I'm making a lot of assumptions here, but using induction to isolate the output so it can be sent wherever safely only makes sense, to me.

Presumably each induction circuit on it acts like a boost converter, like you can buy ones for cheap on ebay and such ranging from tiny ones - for use as usb 5V output from a lithium cell - to big 900W beasts (that I only have several of and use often). So it's sucking power from one cell, isolates and boosts it, and sends it to another. (No boost and it probably wouldn't be able to put out any appreciable current with only 0.05V difference)

...and since it is limited to 1 amp or so, I would assume there is either current limiting built in, or else the "transformer" simply can not exceed a certain amount (like a cheap transformer based charger can only output so much before the voltage falls right off.)

I'm saying that I don't see why it wouldn't be safe to use with ultra caps. Great idea, by the way.

If you're really concerned about not frying a $10 item, you could put low-resistance resistors on each lead. You'll lose efficiency, but it will save it from ever drawing too much instantaneously from frying itself. Would only be inefficient while balancing and not piss away power otherwise.

If I'm doing my math right, the battery I tested mine on would have had a total resistance along any one of the parallel strings of ~0.3 milliohms...which is stupid low...if these things can work with that, I doubt capacitors with their stupid low resistance would bother them. Unless there is some weird interaction between using a capacitor with an inductor that I'm not considering. The source isn't AC so I don't think there's any there to be concerned about.

This is starting to hurt my head. Just hook it up. It'll be fine. No, honest, it will! Maybe.

[serialk11r]

Oh right you have ultracaps on your Acura. The way I imagine a simple passive balancing circuit for ultracaps is to have a "soft" cutoff shunt at 100% of the intended target threshold, allowing them to go out of balance by a bit under heavy current but eventually come down to the correct voltage. Since each cap can go well beyond what the alternator can charge them to at equilibrium and isn't damaged by low voltage either, it's safe to have it go out of balance temporarily.

For lithium ion batteries, the relationship between state of charge and voltage is not as simple, you want the battery pack to normally close to fully charged, and overcharging them is a death sentence to the cell, so I think the simple thing to do would be a high-current hard shunt at say 3.65V on a LFP cell. If all 4 are at 3.65V, you're at 14.6V, so there's no current flowing from the alternator and you're good.

[redpoint5]

Here's my final-ish project.

4.2Ah LiFePO4 battery in the glovebox, connected with a XT-90 plug. The positive lead goes through the firewall to the positive connector, and the negative lead terminates on the bolt in the glove box. There is a 2nd XT-60 connector in parallel, mostly because I decided to upgrade the wiring and connectors. I can connect a 2nd battery in parallel, or leave one plugged in while the other one is charging. There is a balance board connected to the battery. My 6s 350F capacitors are in series and have balance LEDs built into the board. I've had to replace 3 of the LEDs recently.

It's no virtual alternator delete. I'll have to tackle that differently, probably using a voltage regulator. I'll probably never try it considering I don't have much time, hardly ever drive the Acura, and plan to get an EV or PHEV soon. If I had sized my wiring and battery properly, I probably could have eliminated the ultracaps and started straight from the glovebox battery.

[serialk11r]

Looks pretty good to me, nice budget setup. With a 4.2Ah battery though, those wires seem more than adequate given the ultracap bank is contributing the vast majority of starting power, and the real limit is charging amps not discharge amps.

[redpoint5]

2x 10 gauge wire to the glovebox. I bet 2 batteries alone in the glovebox would start the car.

That other fused wire in the engine bay is a charging lead to a 12v trickle charger that is capable of charging LiFePO4 also.

[redpoint5]

Shoot, 2nd time I tried to look up more info on the BMS in the OP, and got my own post as a top hit.

My brother in law gave me a motorcycle (same one as I already have, CBR600), but it had a bad battery. Initially I just used a 6s Ultracap from my other bike, as it had done pretty good before, reliably starting the bike. On one ride though, I accidentally killed the bike and went to restart, but didn't have enough juice.

That got me thinking again that I'd just like to run a straight LiFePO4 battery. I bought 4x 7.2 Ah LiFePo4 cells from Amazon. They are size 32700.


Amazon

I'm going to see if I can get this BMS to work with it, but first I need a spot welder so I can assemble the pack. I've got this on order;


Amazon

It's my birthday this weekend, so nobody can tell me to stop playing with electronics.

[redpoint5]

I discharged and charged all cells, and they all appear to be close to 6,500 mAh. I've spot welded a 4s battery together and just need to figure out how I'm going to attach my 10 gauge leads and balance wires.

Before doing this, I connected the battery in a very jenky way to my motorcycle and started it as proof of concept. No sense building it any further if it wouldn't start the bike.

I'm so impressed with the batteries so far that I placed an Aliexpress order for 12 more, along with a couple balance boards that are more sensitive than the 0.1v ones I've got.

My plan is to replace both motorcycle batteries with these 6.5 Ah LiFePO4, and the 4.2 Ah battery that's in the Acura's glove box.

These battery boxes are super cheap, but alas, they are too tall to fit in the motorcycle.


I'll post more photos and details as I build these things out. Been a busy month so far.

As a tangent, Tesla is now offering some Chinese made Tesla's with the LiFePO4 battery for the Model 3 SR+ Apparently they can fit these less energy dense batteries in the vehicle so long as they aren't a long range model.