[DBCox]

Hi everyone,

I am very new (like just started looking into Lithium Iron batteries 2 hours ago) to lithium batteries. I am considering using them for a DIY ATV instead of gel cell batteries.

Finding the cells does not seem to be a problem, nor does it seem to be too difficult to wire them to the voltage I need (~48 VDC).

Battery managment is confusing me though. I am not used to BMS since I am a gel cell user.

Do I need to have an on-board BMS, a charger that will charge each cell individually, or both???? Battery life is critical as I want to make sure my investment will pay off in the long run.

At this point, I am looking for a typical continuous draw of 30-40 amps, but peak current draw could be in the 400 amp range.

Thanks!

David Cox

[frodus]

what batteries?

[Elithion]

> Do I need to have an on-board BMS,

yes.

> ... a charger that will charge each cell individually,

no

> or both?

Just the BMS and a standard bulk charger
(http://liionbms.com/php/charger_options.php).


Mostly, you need to make two choices:

1) Do you want a BMS that protects the battery, plus it can tell you which cells has an issue, what issue, and how big? Or is it OK to have a BMS that just protects the battery? In the first case you want a digital BMS, in the second case you want an analog BMS.

2) Do you prefer one electronic assembly with lots of wires to the cells? Or lots of electronics assemblies mounted on the cells, with just a couple of cables to a master? In the first case you want a non-distributed BMS, in the second case a distributed BMS.

Then go here:
http://liionbms.com/php/bms-selector.php
and enter your 2 choices from above, plus your number of cells in series, and the utility will spit out which BMS are available, and how much it would cost for your system. You can also answer more questions to refine your search: type of cell, where the BMS is made...

I hope that helps.

[onegreenev]

Do I need to have an on-board BMS

NO. With LiFePO4 Cells you don't need a BMS to safely use them or monitor them. Monitoring with a small pack like that can be as easy as just checking the cells with your multimeter. Bottom balance your cells and don't over charge them and you can forgo the confusing tangled mess of BMS systems that prevail. Everyone will tell you that you need one but guess what, You Don't NEED one. You may choose to use one but you don't NEED one. There are plenty of users of these types of cells that DON'T use BMS's at all and I am one of many.

Contact off list if you need more information. It really is not difficult to do this safely with no BMS.

Now if you decide on cells like Li Cobalt then you really should have a BMS as they are a different beast. I don't disregard BMS systems but for the LiFePO4 cell you don't need one. It is good to monitor your cells if you must but not let your BMS control your batteries. I trust my charger more than any BMS for these cells. I trust my controller more than any BMS for these cells. I trust my own checking from time to time and would rather use the extra money for extra parts or batteries.

Pete

[DBCox]

Thanks for the help guys!

To answer the first question, I am considering using 16 CALB 40AH cells in place of 4 MK MU-1 SLD G batteries. Basically, I will get 25-30% more range with the LiFePo4 with half the weight.
However, I have some concerns after reading all of your responses.

First, I will have to have some sort of BMS, whether it be on the charger or on board. Most of you indicate on board is better, and I would agree from a constant monitoring standpoint. I will not have the time or motivation to regularly check the batteries with a multimeter, so I need some sort of automation. That is a lot of cost and complexity. This setup will be used daily, so I do not have time to chase minor issues to keep it operating. I can just see one cell going bad and delaying the whole day to fix it. Also, this setup will be used by people with NO technical training/skill. If it doesn’t work, they will complain or start replacing major components before finding the real problem.

Also, from a cost standpoint, it does not seem feasible. Typically, the gel cells will last 1.5-2 years. A set of gel cells costs $240, so we are looking at $120-$160/year. The LiFePo4 cells will cost $928 + BMS. If I neglect the cost of the BMS by assuming I will never have to repair or replace it, the cells will need to last 6-8 years to break even from a cost standpoint. The improved performance and lower weight could possibly justify a slightly higher costs, but not much.

I think I need to wait a little longer until the price and reliability improves. Please correct me if my assumptions and numbers are incorrect.

-David

[frodus]

Yeah you could go without, but if you're not going to maintain your cells and watch them from time to time as onegreenev (Gottdi) says, you'll have issues with balance over time. If you do have time to bottom (or top) ballance the cells and watch them, you may be ok. But Bare minimum IMHO, you need to monitor and provide an output from your monitor to shut down the charger or controller if it goes too high or too low, respectively.

If you want everything done automatically, then use a small BMS. BMS for 16 cells is fairly easy to find and are usually fairly reasonable.

Some will say you don't, but they're actually performing the work of the BMS. Some will say you do because they're thinking you won't maintain batteries or be able to watch for LVC or HVC. Both are right that batteries need to stay balanced (whether top or bottom). In each case, the batteries are being watched and maintained by someone (without bms) or some thing (with BMS).

I personally use a BMS, but there are a lot of options out there to do everything from just monitoring, to a full balancing system. Even something like minibms can be a resonably priced solution and would be a good choice.

Also, I don't know where you're getting batteries, but $928 seems high ($57.75 per cell, or 1.45/Ah). Does that include shipping?

[Yabert]

Originally Posted by DBCox

The improved performance and lower weight could possibly justify a slightly higher costs, but not much.

A swap from 32Ah lead to 40Ah lithium will be a huge improvement in performance considering the power capability and the weight of the calb cells. And you can expect almost double your range because the 32Ah lead are just rated 15-20Ah at high rated discharge (1C to 5C for example).

I used the mini bms centralized in my car and it work well.
Simple and easy to assemble.
http://minibms.mybigcommerce.com/pro...ter-Board.html

Hey, don't hesitate to share picture of your DIY ATV.

[dougingraham]

Originally Posted by DBCox

Battery managment is confusing me though. I am not used to BMS since I am a gel cell user.

Do I need to have an on-board BMS, a charger that will charge each cell individually, or both???? Battery life is critical as I want to make sure my investment will pay off in the long run.

At this point, I am looking for a typical continuous draw of 30-40 amps, but peak current draw could be in the 400 amp range.

Thanks!

David Cox

You are in luck since LiFePo4 need about the same level of BMS as Gel cells. Actually I consider them less fragile than Gell cells since leaving them discharged for several months doesn't hurt them much at all. And even a few days discharged with a Gel cell is death. They also have essentially zero self discharge which means that once you balance them they stay balanced.

If you are planning on running the batteries dead on a regular basis then bottom balance the cells and set your charger to cut off at whatever pack voltage shows when the first cell reaches 3.6 volts. That is pretty much all you need for a manual BMS. You can check this occasionally but you will eventually get tired of doing so when nothing ever changes.

If know you will never run the pack to dead then you can top balance and set the charger to stop when the charge reaches 3.6 volts times the number of cells. If you ever do run the pack dead you will most likely ruin a cell in this situation. With bottom balanced packs the vehicle will just stop moving and once you charge it up everything is back to normal.

For the amount of money you would spend on a BMS you can buy quite a few extra cells.

Anyone who makes or sells a BMS will tell you that you need one. People who have bought them and use them will tell you that you should have one. This is just human nature. But it also biases their opinion.

[dougingraham]

Originally Posted by DBCox

Thanks for the help guys!

To answer the first question, I am considering using 16 CALB 40AH cells in place of 4 MK MU-1 SLD G batteries. Basically, I will get 25-30% more range with the LiFePo4 with half the weight.
However, I have some concerns after reading all of your responses.

First, I will have to have some sort of BMS, whether it be on the charger or on board. Most of you indicate on board is better, and I would agree from a constant monitoring standpoint. I will not have the time or motivation to regularly check the batteries with a multimeter, so I need some sort of automation. That is a lot of cost and complexity. This setup will be used daily, so I do not have time to chase minor issues to keep it operating. I can just see one cell going bad and delaying the whole day to fix it. Also, this setup will be used by people with NO technical training/skill. If it doesn’t work, they will complain or start replacing major components before finding the real problem.

Also, from a cost standpoint, it does not seem feasible. Typically, the gel cells will last 1.5-2 years. A set of gel cells costs $240, so we are looking at $120-$160/year. The LiFePo4 cells will cost $928 + BMS. If I neglect the cost of the BMS by assuming I will never have to repair or replace it, the cells will need to last 6-8 years to break even from a cost standpoint. The improved performance and lower weight could possibly justify a slightly higher costs, but not much.

I think I need to wait a little longer until the price and reliability improves. Please correct me if my assumptions and numbers are incorrect.

-David

I think you will see a real range increase of twice what you are seeing with the gel cells. If your gel cells are lasting two years you are not running them down much below half and charging them right away.

All your manual battery management time is done once up front and then you might want to check it once per year. If you treat these the same way as the gel cells you are seeing 2 years of life from you can probably expect 12+ years before they are as bad as the gel cells would be when they are brand new. Unless you abuse them they should last as long as the vehicle.

One of the online resellers has the 40AH Sinopoly or GBS cells listed for $52 each giving a total of $832. I am guessing you can find them for less than that.

[PStechPaul]

I have thought about making a very simple BMS that would just be a tiny device with two or 3 LEDs across each cell, where the LEDs would tell the status of the cell. I would, ideally, use a PIC, but here is an idea using three LEDs, two transistors, and 6 resistors which has one LED change brightness with the voltage of the cell, and the other two would switch (red to green for example) when the cell voltage crossed the threshold of 2.5 volts. I'm just guessing at the threshold for LiPo cells, and that can be changed. It may also be temperature dependent or vary with LED types. But a PIC would be much more accurate, and could flash the LEDs to save power. My simple circuit draws 2.5 mA at 2.5 volts, so that's probably a tiny load on even a 25 A-H cell. Good for 10,000 hours.



If this would be useful maybe I could work on it. It would cost less than a dollar to make. Biggest cost will be packaging and reliable connection to the cells. I really think the manufacturer should put such a device in their cells so you can just look into a little window, like the "green eye" batteries.