Lithium battery BMS questions

[IamIan]

Originally Posted by TurnNBurn

The only problems I see, making this idea not very cost effective, is the need for 3 different power supplies for the battery chargers.

Each charger is about $110, so two of them will be $220. Power supplies bring total cost up to $300. I can get a standalone charger for my entire pack for $250-$350.

That is a comparison you will have to consider ... I've known people that have done it either way.

- - - - - - -

Note :
Depending on the specific details ... separate power supplies are not always needed ... that gets back into the input-output isolation issue I mentioned above.

I haven't done it with the iCharge myself... but I suspect the topology is similar enough to the PL8 that it might behave in a similar way.

2 examples of input-output isolation concept using PL8:

#1> I have used 1 big power supply (or one feed in supply battery) ... to run 3 PL8's all at the same time .. as long as the PL8s outputs are electrically isolated from each other ... that's how I did multiple tests running at the same time... for your pack if you had jumpers or something you could use to electrically isolate those outputs you might be able to do something similar.

#2> I have also done the opposite ... 3 electrically isolated separate power supplies (one for each PL8) .. or one feed in supply battery for each ... all 3 PL8 outputs going to one battery (that's how I did over 100A charge-discharge tests).

Note: I had to have one of them (inputs or outputs) electrically isolated from the others ... it didn't mater much which one I isolated ... as long as one side was isolated.

Note: Some other devices have their own internal input-output electrical isolation ... those kinds of devices would not care at all even if there was no electrical isolation even on both input and output ... but that depends on the topology / design of the device.

[TurnNBurn]

I had a question regarding my leaf batteries,

One "module" is a 3.6v pack, but inside of that is 4 "cells" wired in series/parallel (2s2p). Does the entire "module" count as a cell? So when I look at chargers like the imax or turnigy, do I need to be concerned with the overall module, or the 4 cells inside?

I'm pretty sure it's 4 cells I need to be worried about. But I'm tired, sleep deprived, and just excited about getting back to this project in a few months.

I'm about to pick the project up again and I'm considering 3 options:

Cell log and arduino BMS
Arduino BMS, custom
Multiple separate chargers (which will have its own advantages)

[Astro]

Originally Posted by TurnNBurn

I had a question regarding my leaf batteries,
One "module" is a 3.6v pack, but inside of that is 4 "cells" wired in series/parallel (2s2p)...

If each cell is 3.6v then wouldn't a 2s2p "module" be 7.2v?
2s2p gives you twice the voltage and twice the Ah of a single cell.
4 cells wired to give 3.6v output would have all the cells in parallel, a 1s4p arrangement (if the individual cells were 3.6v).
I would say the Leaf modules are 7.2v or so output and if the module doesn't have the mid point of the 2S part of the 2S2P then you will have to treat them as 7.2v cells.
If the mid point between the two sets of paralleled cells is available then you could treat it as two cells in a single module. I.e. you could use a BMS that monitors 3.6v cells.

[TurnNBurn]

Originally Posted by Astro

If each cell is 3.6v then wouldn't a 2s2p "module" be 7.2v?
2s2p gives you twice the voltage and twice the Ah of a single cell.
4 cells wired to give 3.6v output would have all the cells in parallel, a 1s4p arrangement (if the individual cells were 3.6v).
I would say the Leaf modules are 7.2v or so output and if the module doesn't have the mid point of the 2S part of the 2S2P then you will have to treat them as 7.2v cells.
If the mid point between the two sets of paralleled cells is available then you could treat it as two cells in a single module. I.e. you could use a BMS that monitors 3.6v cells.

Grrrrr. I was going off of my horrible memory, since all of my notes are at home. But yes, I looked again and they're 7.2v.

The specs say:

60a/h rated - 7.6V - 500 W/h - max continuous power 1875W or 240A
Peak current 540A
Measured 62Ah from 4.2 to 3V, and 64Ah (4.2-2.5V)
Maximum voltage 4.2V per cell / 8.4V per battery
Average voltage 3.8V
90% of the capacity voltage - 4.2-3.7V

Each module contains 4 prismatic cells (see photos), connected 2S2P
Each prismatic battery has rated capacity 30A/h and 3.8V nominal voltage

[e*clipse]

I have a Leaf battery pack in my shop and a couple modules sitting on this table. This info is accurate.

I noticed one of the suggestions early in this thread was to use an arduino based system. If you're not into building one from scratch, but would like some control over the programming, you might consider the system offered by Linear Technology. It's actually not really a "product" but more of a demo system for companies developing a BMS. I'm going to assume you mean 9 modules, which would be looked at like 18 cells from this system's perspective. ( the 2 parallel cells would be seen as one by the BMS ) Each LT IC is designed to manage 12 cells. It uses a serial or parallel communication bus to a central "Linduino" ( Arduino ) communication microcontroller. The Linduino can then communicate with a USB to just about any PC.

- E*clipse

Originally Posted by TurnNBurn

Grrrrr. I was going off of my horrible memory, since all of my notes are at home. But yes, I looked again and they're 7.2v.

The specs say:

60a/h rated - 7.6V - 500 W/h - max continuous power 1875W or 240A
Peak current 540A
Measured 62Ah from 4.2 to 3V, and 64Ah (4.2-2.5V)
Maximum voltage 4.2V per cell / 8.4V per battery
Average voltage 3.8V
90% of the capacity voltage - 4.2-3.7V

Each module contains 4 prismatic cells (see photos), connected 2S2P
Each prismatic battery has rated capacity 30A/h and 3.8V nominal voltage

[TurnNBurn]

So my option is build a BMS with arduino, which I'm considering for a future Mark II enhancement, or buy 3-4 separate chargers and balancers. I like the 3-4 chargers idea because I can charge the pack faster and balance the batteries.

However, like I was wondering, if I go with the 3 or 4 chargers, so I need a high voltage switch? or is there something I can get that will just automate the process of separating the packs? I don't want to have to flip 4 switches, plug in the bike, and then start the chargers. That's quite a process. My main concern is "what if I forget to flip a switch?". I'd kill my pack or do some damage.

[e*clipse]

I'm building my BMS primarily to watch over the batteries. I have it on pretty good authority from a trusted collegue that there really isn't a lot of balancing involved for these cells. His advice was that the balancing was only needed about 1X or maybe 2X per year - for a daily driver.

The primary objective of the BMS is to watch the SOC of the cells and make sure no cells go above or below the voltage tolerance. In addition, it's good to make sure none of the cells overheat.

The BMS can help you ensure you don't overcharge a cell with a charger designed to handle the voltage of all the cells in series. Also, while you are driving, the BMS will ensure you don't kill a cell by overdischarging it. This can be pretty dynamic because higher current will also pull the cell's voltage down.

My BMS will have authority in charging, and will shut off the charger when a cell reaches a pre-set voltage. This (I hope) will reduce the risk of killing cells by overcharging or causing drama with an overheated battery pack.

I'm deciding what level of authority the BMS should have in discharging. It seems some level of limp-mode would be good at a pre-set voltage.

Anyway, that's just my take on the problem.

Good luck with your build!

[Astro]

Originally Posted by e*clipse

...I'm deciding what level of authority the BMS should have in discharging. It seems some level of limp-mode would be good at a pre-set voltage...

I would second that. Some sort of warning from the BMS is better than just cutting drive power as a reaction to a low cell voltage. Then after a short length of time if not manually overridden move on to more aggressive measures. Limp mode or shutdown.
Having some form of manual override would alow you to continue to drive even after the low cell voltage warning. Leaving it up to you to either feather the throttle and find somewhere to charge asap or just safely pull over and shut down.
The last thing you would want is some sort of situation where you needed a burst of power but the BMS cuts the drive due to the increased voltage sag. Leaving you stranded on the railway crossing.

The BMS can make the decision on when to cut the charging cycle but i wouldn't want to leave it solely to the BMS deal with under voltage events. Under voltage events require a decision based on the current driving situation, something the BMS is largely unaware of.

[TurnNBurn]

I was looking at getting a Polaron EX charger, which has 4 separate chargers built in, and wiring it like this:



Basically, divide the pack into module groups of 3. Then have a switch where I can engage/disengage the chargers and the battery pack. This way I can have isolated chargers and I can also get them balanced.

Is there a speed advantage to charging 3 packs at a time?

[IamIan]

Originally Posted by TurnNBurn

I was looking at getting a Polaron EX charger, which has 4 separate chargers built in, and wiring it like this:

Basically, divide the pack into module groups of 3. Then have a switch where I can engage/disengage the chargers and the battery pack. This way I can have isolated chargers and I can also get them balanced.

Is there a speed advantage to charging 3 packs at a time?

I'd recommend rotating the cells .. instead of diagonal bus bars .. keep the bus bars shorter and more perpendicular to each cell.