BMS Wiring, Balancing, Charging Issues.

[The Mighty Volt]

Hi All.

This is a small 8s LiFePo pack I made up with some cheap, second A123 18650's

The pack is around 9.9Ah Nominal, I have pulled 17A Continuous out of it during testing. Testing is an Ananda motor in a work-top vise with the motor spinning under no-load and artificial load conditions, alternately.

However something, either with the battery pack, or with the BMS, is just not correct.

Before I installed the BMS, I hooked the battery up, bareback, to a Meanwell, set the voltage for around 28.8v and let it off.

Some of the strings charged up to ridiculous levels, they would have hit 5.0v or more if I had allowed them to. I pulled the plug on it, installed the BMS, and tried again.

Now when I charge, even though it should cut out at 28.8v {3.6v times 8} it goes all the way to 29.7v before the V-Power charger cuts out. The V-Power charger is around 30.9v or so. It seems that the charger, rather than the BMS, is doing the cutting-out.

When I was using the battery, I did note the voltage of all the strings, and I was getting some very very good results, some of the strings had identical voltage but one was lagging behind.

After 5Ah+ had been pulled from the pack, the battery started to flag, thanks to one particular string weakening.

The BMS did not cut out the flow of power to the motor, instead the battery continued to supply power to the motor and the weak string continued to plummet, I let it drop away, and then pulled the plug at around 1.6v or so.

I then performed the charging process, via the BMS and the V-Power charger, some of the strings charged all the way up to 3.8v+ {never going beyond 3.9v} and some of the strings only went as far as 3.4v.

I did notice that some of the resistors of the BMS did get hot during this process, ostensibly burning off excess current. So something is working.

As soon as I ran the motor, the strings all dropped down to 3.4X v, except the weaker strings which stayed around 3.3X volts.

Any suggestions or help would be gratefully received. Here are some photos of my set-up. It is an 8s battery, with 7 balance wires coming from the BMS. Those 7 wires are going to one positive terminal, each. The battery is put together in such a way that the strips alternate and are connected in series.

Any help here would be appreciated, I just want to get this working and actually learn something.

 

[scoot]

Everythiung you have described thus far sounds perfectly normal to me for a newly assembled pack. The cells that are still lingering around 3.4 volts near the end of charge have just not completely filled up yet. If you have access to a single cell charger, you can top them off one by one. Otherwise, it may take quite a while and many charge cycles to wait for the BMS to complete the process of balancing.

 

[The Mighty Volt]

Hi Scoot, many thanks for your reply.

However, why would the BMS fail to cut off the power to the controller when the weak string started to plummet? I thought it would cut out at 2.8v, no?

Also why the charging of certain strings to as high as 3.9v- I thought the BMS would cut out anything above 3.6v and burn it off through the resistors, all the while making sure that the voltage of the entire battery never went beyond 28.8v?

Anyways sorry to burden you with this.

One thing which is encouraging is that the BMS does seem to be burning off some current as the resistors are getting hot.

Something is forcing the charger to cut off, but I think it is the pack voltage equalizing with the charger voltage and tripping the charger, rather than the BMS.

Anyways thanks for your response Scott.

 

[scoot]

Well maybe I should retract the "normal" comment regarding the LVC & HVC. That does sound like like it may not be functioning as desired... especially the LVC part

In regards to the charging, it may be that the higher cells are starting to shunt away some of the excess voltage from the high cells somewhere around the "3.6 volt" level (hence the heat on the resistors), but i suspect they are not quite keeping up with the bulk charge current in (quite normal actually), so they'll rise above that 3.9v level as they bleed off while the lower cells (or parallel strings if you will) are catching up. I only start to worry when they run over 4 volts for prolonged periods on my Headways (LiFePO4 also). Others out here may have different comfort limits in their minds.

As far as the HVC, is concerned, it may be just that the pack voltage limit on the charger is slightly below the limit of the BMS, and if the cell charge cut off limit designed into the BMS is around, say 3.9 volts (I have seen some BMSs/balancers that use this limit), then the pack voltage limit on the BMS may be 31 volts or more. From what i have seen, there seems to be bit variation from one bms to another when it it comes to the precise limits they incorporate.

In any event, I suspect that once you get your pack completely balanced for the first time, then you may see things behaving more like you expect, and you'll get a better idea for the actual charge limit settings of your BMS. As for the LVC, best if let someone else field that question, since I have so very little actual experience in that arena.

 

[The Mighty Volt]

Thanks scoot.

The battery pack as such is actually working fine, it puts out great amps for such a small Ah rating.

The cells are definitely cast-offs, so there is every chance that certain strings are just "better than others"- none of the cells are able to break 4.0v even during constant charging since I fitted the BMS, but there are 2 strings which simply cannot break 3.40v even during constant charging.

I am not worried about that.

What I am worried about is the failure of the BMS to cut out at below 2.8v.

Anyways we shall see in due course.

 

[DAND214]

MV, what brand of BMS are you using?
I have seen 3.85v on most of my signal lab v.2 bms on all cells. Low shut off at 2.0v so 30.9v charger is correct. As for the cells that are at 3.4v need to be single charged to 3.8v so it will stay balanced. I have also noticed the 18650 cells seem to stay at higher voltage longer then fall off like a nicad does.
Charge the pack full then single charge the low ones, they should start charging without assistance after a couple charges.
Like I said cutoff on my packs is 2v per cell so 8 x 2 is 16v.
16s shuts down around 32v as long as the cells are all the same strength. All it takes is one runt and it's done at 48v.
Dan

 

[The Mighty Volt]

Cheers Dan.

I cannot actually tell what brand of BMS it is, it seems to be a generic sort that comes with the duct-tape LiFePo4 batteries.

Here is a better photo of the BMS alone.

Thanks for your advice. I will try to single-charge the weaker string on its own today using a CC-CV power supply, I'll set voltage to around 3.65, charge it to that then let it go onto Constant-Current until it is done.

Thanks again.

 

[dnmun]

be careful handling these packs. if you look at your parallel strip spot welded to the top of the positive terminals, there is a short section on the end that extends beyond the end of the spot welds to the top of that last one in the row.

the yellow plastic sleeve that is shrink wrapped around the top of the cylinder is all that separates the metal strip from the case, which is the negative terminal of the cell.

if that strip cuts through that insulating strip, it will short out the entire row of cells and cause a real significant amount of heat to be released rapidly. this explosion can result in loss of your eyes if you are not wearing glasses.

i have not decided what to do with these tails on my vpower pack i am rebuilding, but i may add a sliver of thicker rigid plastic underneath that strip where it crosses over the edge of the case, and glue it.

 

[The Mighty Volt]

Thanks for the safety hints, dnmun, I just pulled out the battery for a few quick photos, it's just a battery I am using to test motors and refine my battery building techniques.

 

[cell_man]

Hi Trao,

the BMS will have a HVC of about 3.9V, so if none of teh cells reach that, the BMS will not disconnect the charger. Cell balance shunting starts at around 3.7V typically. If you have some cells with a lower SOC than others, they will take a long time to catch up. Typically the charger might be supplying only 100mA when the cells are balancing, so any small differences take a long time to balance out. Single cell charging the low cells is often the quickest solution. I generally use an RC charger to balance packs after I have built them and will leave it running overnight to do it's magic.

The LVC is typically 2.0V but there will be some time delay, so it might not happen instantly. You might just need to give it some more time.

Regrding isolation of the cells from the negative terminal. Most cells have an additional insulator, typically something like a plastic washer to give some additional isolation where the terminal and the case come into close proximity. This is the case on the A123 26650s and other LiFePO4 cylindrical cells I've seen.

 

[The Mighty Volt]

Cheers cell_man, appreciate it.

The photos are not the best {my batteries went dead, how appropriate}, the way the BMS is connected is there are 7 balancing strings, the first of which goes to the first positive of the string immediately "after" the string which the main Positive terminal is taken off.

The 2nd string then has a balance wire, also to the positive terminal, now on the other side of the battery pack, thanks to the alternating +/-/+/- of the pack build.

There is no balance tap on any negative terminal of any string.

I do think the BMS is working to a greater or lesser extent, just not sure how much.

I have taken the advice of people here and I am performing an individual charge on the weak strings, of which there are two, of differing weaknesses.

Encouragingly, right now, I see exactly identical voltages on three or four of the strings.

Here is hoping it all squares up after the charging of the weaker strings.

Volt.

 

[scoot]

As a side note, the only time having "good agreement" of voltages between cells is when they are at 3.45 volts or higher. Having what seems like good agreement in the 3.2 to 3.4 volt range can be deceiving when it comes to comparing states of charge. Once all of the cells show 3.45 v and higher soon after the charge cycle is completed, you'll know they are all filled up, hence the pack is finally top balanced. Interestingly enough, having a wide variation of voltages at levels above 3.45v is also insiginficant in the opposite way... that is to say, having one at 3.5 and another at 3.9 means they are more or less equally filled up.

 

[The Mighty Volt]

Thank you all for all your help.

Just when it seemed I was getting somewhere, I suffered an obscure set-back.

Using a multimeter, the battery puts out 28v or so.

Moving to the BMS, and still using the multimeter, I get around 26.7v or something like this.

However- once I attack my DocWatson too the battery {at "source" wires} and attempt to measure the voltage {at "load" wires} I get a reading of around 1.6v or so.

The DocWatson itself does not display this voltage as it is too low to power the unit- it is the multimeter, attacked at the "Load" terminals, which displays this value.

One thing I can say is that yesterday, I had the battery charging and discharging into a running {no load} motor, all at the same time. Would this have damaged anything?

Sorry for all of this nonsense anyways- if anyone can help I would be more than grateful.

Volt.

 

[The Mighty Volt]

Hi sorry to bother you all again....

ES member "Mike" kindly sent me this BMS a while back.

My problem is that it is a bit different to other BMS I am used to working on, in that it seems to lack a "P+" terminal.

I see: P-, B-, B+, CH-

The only "P+" is for a very tiny wire which actually looks like the positive for the charger.

http://i1238.photobucket.com/albums/ff494/ElectricPonies/DSCN4923.jpg?t=1297352948[IMG]

[img]http://i1238.photobucket.com/albums/ff494/ElectricPonies/DSCN4924.jpg?t=1297352980


Again any help in wiring this sucker would be gratefully received and appreciated. Thanks.

 

[cell_man]

The normal convention IME is B goes to the battery, P goes to the charger and C or CH goes to the charger. The sign next to the letter seems to indicate the polarity to me. If you are not confident just fit a very small fuse in line when you try a connection to avoid frying anything.

IME most BMS do not have a positive from the pack other than the balance wires but I have seen some where there is an additional pack positive which disables the BMS when not connected. I can onlly think that in some cases this could be a way to isolate the output voltage using a low current switch.

Stop being a baby, just go for it. If only you were so careful when it came to putting that plumbers flux all over those A123 cells :lol: :wink:

 

[The Mighty Volt]

The normal convention IME is B goes to the battery, P goes to the charger and C or CH goes to the charger. The sign next to the letter seems to indicate the polarity to me. If you are not confident just fit a very small fuse in line when you try a connection to avoid frying anything.

IME most BMS do not have a positive from the pack other than the balance wires but I have seen some where there is an additional pack positive which disables the BMS when not connected. I can onlly think that in some cases this could be a way to isolate the output voltage using a low current switch.

Stop being a baby, just go for it. If only you were so careful when it came to putting that plumbers flux all over those A123 cells :lol: :wink:

If I had been so careful you wouldn't be eating so well!!

So P and Ch go to charger?

Why does the other BMS have a P+/P- and a B+/B-??

This is the only 12S BMS I have so I don't want to break it.

Cheers.

 

[DAND214]

I have that in a 16s and 12s also it has only one output/input. The Positive battery and the output positive is shared. the neg from battery goes to the corner and other neg is output and charger. there is only one anderson set used for both the charger and also the power to the controller. Not like a Ping that has a charge lead and a power lead.
That ios how I received the battery from Lau Chang 2+ years ago and it still has full ah of 36v20ah+.
Hope that helps.
Dan

 

[The Mighty Volt]

I have that in a 16s and 12s also it has only one output/input. The Positive battery and the output positive is shared. the neg from battery goes to the corner and other neg is output and charger. there is only one anderson set used for both the charger and also the power to the controller. Not like a Ping that has a charge lead and a power lead.
That ios how I received the battery from Lau Chang 2+ years ago and it still has full ah of 36v20ah+.
Hope that helps.
Dan

Hi Dan thanks for your reply.

I might be a bit slow but I just cannot follow quite what is going on here. Just what do you refer to when you say "the corner"

I understand this much: the positive lead to the controller and the positive lead from the battery are sharing the same bottom-left-hand terminal on the BMS.

Does B- accept a cable from the battery negative?

Does P- go to the controller negative?

Does Ch- accept a negative terminal from the Charger? If so, which terminal accepts a positive charge from the charger?

What are the small P- and P+ terminals for {the very very small gauge red and black wires}

Finally, I assume the balance wires all go to the positive terminals of the strips, as usual.

Sorry to be so slow on the uptake.

Cheers.

 

[The Mighty Volt]

Ok I think I might have figured this out.

B+ accepts the positive lead from the battery.

B+ also sends a positive lead to the controller

B- accepts the negative cable from the battery

P- sends a cable to the controller negative.

Ch- accepts a cable from the Charger negative.

Ch- and the B+ to the controller are used to charge.

P- and the B+ to the controller are used to power the controller.

 

[DAND214]

MV, which BMS are you using? The first pic or the other ones? I have the first one in the post and all it has is 4 wires. Battery neg goes to the rt corner the positive in and out lt corner and the power neg is next to the positive wires. No other wiers, no seperate charge cord, it shares the output for charging and power.
Dan

 

[cell_man]

Ok I think I might have figured this out.

B+ accepts the positive lead from the battery.

B+ also sends a positive lead to the controller

B- accepts the negative cable from the battery

P- sends a cable to the controller negative.

Ch- accepts a cable from the Charger negative.

Ch- and the B+ to the controller are used to charge.

P- and the B+ to the controller are used to power the controller.

Yes, that looks correct

 

[cell_man]

But don't blame me if it goes BANG! :?

 

[DAND214]

Here's me battery wired.
View attachment 1.jpg
Is this what you have?
This works as I received it. One connector shared.
Dan

 

[The Mighty Volt]

Actually that looks to be a different BMS to the one I have but I will try it. Thanks for uploading the pic.

I don't understand why they have to introduce third categories like CH- into the equation...keep everything P and B and its easier to follow.

 

[The Mighty Volt]

MV, which BMS are you using? The first pic or the other ones? I have the first one in the post and all it has is 4 wires. Battery neg goes to the rt corner the positive in and out lt corner and the power neg is next to the positive wires. No other wiers, no seperate charge cord, it shares the output for charging and power.
Dan

I am using, currently, the second one.

Sorry for the confusion caused by using two photos.