Simple BMS to just prevent high cell voltage?

rocwandrer

100 W
Joined
Oct 28, 2011
Messages
286
Location
Northeast USA
I'm looking for an inexpensive solution for ensuring no cell in a 4s1p LiFePo4 pack gets overcharged when using a constant voltage series charger (14v). I don't care about over discharge protection, etc, and specifically do not want a discharge amp limit. Input?
 
you can use the first four channels of a 8S BMS to charge. i never thought i would see someone need a 4S BMS but saw an ad for a Bestechpower 4S yesterday:

http://www.ebay.com/itm/BMS-LiFe-4-Series-12V-/321298742737?pt=US_Rechargeable_Batteries&hash=item4acee5bdd1

or you can go to bestechpower and order any other model you want.
 
I would also build a bms protected pack. Use it for charging, but come straight off the batteries for your controller.
 
For this application (ice starter battery), charge and discharge must be the same terminals. I was thinking zener diodes in reverse parallel (with appropriate resisters to keep things from burning up). Waste a watt or so anytime an individual cell voltage is above 3.4v to keep things aimed at more in balance, something like that? I suppose it isn't that simple and I'll need 4 transistors and high wattage resisters...
 
rocwandrer said:
For this application (ice starter battery), charge and discharge must be the same terminals. I was thinking zener diodes in reverse parallel (with appropriate resisters to keep things from burning up). Waste a watt or so anytime an individual cell voltage is above 3.4v to keep things aimed at more in balance, something like that? I suppose it isn't that simple and I'll need 4 transistors and high wattage resisters...

what is a zener diode in "reverse parallel" and how does that relate to balancing the cell? i thought you said 14V?
 
parabellum said:
He must be a Very Smart Guy to solve this Simple issue with zeners, transistors and resistors. I would be very happy to see this finally happen! :)

The disconnect here, if any, has nothing to do with intelligence and everything to do with training and experience. This is far outside of my area of expertise. I didn't say "I'm going to....!". I didn't say the problem was simple. I said I am looking for a simple way to do this. I asked a question. It sounds like you think it is obvious that it can't be done. Care to explain enough to reveal to me why it can't be done?

Like 95% of others running this type of pack in this application, including the commercially available packs, I will not be running a full BMS.

How about this:

One sub-circuit per cell, each with an led setup with resisters so that it's turn on voltage is achieved when that cell is above 3.4v. When the voltage across the led is too low to turn on, the current flow is in the µa range, right? When the led lights, the current will be in the 20-70ma range. . The charging system is targeting 90% soc and achieving 80% most of the time. The bleed current on any cell above 85% or so is an order of magnitude higher than the cells below 85%. Since the draw is 24/7 and is greater on the cells at a higher soc, an inherently balanced pack will tend to stay balanced. Since the charger is run at least once every 2 weeks, the discharge due to the constant draw is minimal.

I realize most of you will disagree with my intent here (not running a full BMS), but maybe you can direct your ridicule at why that wont even work for what I want to do :D
 
why come here and ask if you don't wanna listen to others advice. if you need a free yes man then you should get married. you never explained what the reverse parallel diode trick does.
 
dnmun said:
why come here and ask if you don't wanna listen to others advice. if you need a free yes man then you should get married. you never explained what the reverse parallel diode trick does.

The important piece of any advice is to understand the why. That is where the value is. I understand why a BMS is important, a good idea, a smart choice. If you can accept that I want to do something that is a bad idea, I would find input into why it might not do what I want it to do very helpful.

I did not answer about the zener diode because I instead decided to give a simpler concept that I think should do the same thing. I did that to avoid getting bogged down in some minor detail I don't get that is related to zener diodes rather than the general concept I am asking about, since there might be a fundamental flaw in my approach making the complexity of using the zener diode voltage regulator type of approach irrelevant.

If someone can educate me as to why it wont work, no need to get into my (most likely flawed) understanding of how it might be better to do it with a zener diode.

1.) Can a simple led with the right turn-on voltage (tuned with resisters) be used to bleed power faster from cells that have a higher soc? Why not?

2.) Can bleeding power from individual LiFePo4 cells faster above 3.4v/cell than below keep a pack that will be charged to 80% soc target in balance enough to significantly decrease the chances of overcharging any individual cell in series charging with no balance taps? The only flaws I can see in this are 1.) too much standby draw and 2.) making sure the standby draw is very well matched so it doesn't create imbalance. Are those surmountable issues? Any others?

Thanks to all who are here trying to help. I can't really pay that back here very often since I'm not very knowledgeable about electronics, but I pay it forward every day in other areas, which is why I am not reserved about asking.
 
999zip999 said:
Rocwander no worries. Welcome to the shark tank. I have seen single cell bms on top of bigger lifepo4 cells.

Anything that achieves some degree of balancing without limiting discharge current or being very costly? Did it look anything like what I described above? I only need 4 cells worth, and with soc being between 60-80% 99% of the time, it doesn't need to be fancy. Other ideas for how to implement something like this?
 
999zip999 said:
They are on top of thundersky or calb type cells. What type or brand of cells you plan on using.


I was thinking Headway 38120HP.
 
Back
Top