Stuck building 48v battery

mcristiani said:
Newly built pack with different ebay board is built and running. Thank god! Cruised about 3 amp hrs. CA read 47v when drawing 7amps.

Of course, there has to be some snag though... When connecting a charger, I can't get the charger to kick on and charge. Resting voltage at both plugs is about 49v right now and I just charged a different 48v battery with the charger I am using.

Any thoughts?

Your pack must be out of balance. Check all the cell voltages on the multi-pin connector.
 
It could be one (or more) cell groups is going above 4.3v or whatever the HVC limit is during charging. Sometimes this is hard to find as it might drop back to normal range as soon as the charge stops.
 
I am back again still with no luck:( I really can't believe it.

A set of cells is now dead on this new battery. I spent a lot of time to get these charged up, confirm they held. I installed a new board from battery space and they held for a bit. Now a week later, series 1 is at 1.3v and effectively dead.

I have no choice but to move away from this board now.

I followed fetcher's description for connecting B- | Balancing Harness | B+ but I am going to ask to confirm something more:

With a 13s 2p pack I am connecting the board in the following way:
Series 1 - I connect B- wire from board along with a balance wire to a tab at the negative pole of the series. There is other series connection at the negative terminal here.
Series 2-12 - I connect the series alternating positive and negative.
Series 13 - I connect the two B+ charge and discharge wires along with a balance wire to a tab at the positive pole of the series. There is no other series connection at the positive terminal here.

I am assuming all of this is in order because I don't get any crazy sparks and the batteries run.

At this point, I am really wondering if those boards are LiFePo? I just don't get it....
 
Your connection sequence sounds right.

If the board was made for LiFePO4, it still shouldn't drain the cells that much just sitting there. If the cells don't drain with nothing attached, I'd have to assume bad board. If you have a meter that can measure milliamps, you could stick a charged cell into position 1 with the meter between the cell and one of the connections and measure if there is any drain. Should be less than 10uA, which is near zero.

If you ran the pack and tried to recharge with a LiFePO4 board, it would kick off before the cells got fully charged but they would still get over 50%.
 
Happy Labor Day!

I am communicating with batteryspace about why their specific boards keep failing.

The 48v pack I built with a different board is holding. I was able to get it to charge. It looked like a balance wire had popped off inside.

I am wondering why the individual groups range from 3.95v to 4.08v though? Is this because I was not diligent enough at the build point?

Additionally, could I connect my volt charger across 2, 3, 4, n sets of parallel groups as long as they are all close voltage? For example, I want to charge up parallel groups 4,5 and 6 as they are 3.99 and parallel group 4.10.
 
Could I "draw power out" of a single parallel group to bring it down to the other groups?

I have one lone group at 4.15v. I feel like it is shutting off the charger prematurely.

I was wondering: can I switch the positive and negative lead on my 5v charger and connect to parallel group? Or is that plain stupid. I also have a little setup I bought off ebay that is a board with resistor - I guess - that draws down the battery at a mellow .5amp. Can I use that?
 
You can charge or discharge individual cell groups to balance the voltage. You can also charge/discharge several groups at the same time if you have the right connector. I always do it through the multipin using a spare 6S balance connector with the right pins connected. You can buy pin-strips from any electronics supplier or from Ebay. Break off the length you want and stick it between the multi-pins and 6S connectors; however, i have a BC168 charger that can charge/discharge each channel independently. if you use an iMax type, you need to double up the outer wires of the groups that your charging to act as the power wires:

http://www.ebay.co.uk/itm/100Pcs-2-54Mm-Pin-Header-Single-Row-For-Arduino-Straight-Male-40Pin-Strip-New-S-/112129439886?epid=923995927&hash=item1a1b6f788e:g:klcAAOSwXeJYHd~3
 
Thanks d8veh.

I don't totally follow what you are saying though.

I clicked on the link and have some arduinos, and header pins, for other projects. I will reread and see if I can decode:) At the moment, I am not touching the balancing harness when charging as I am afraid of it - and the boards that keep failing.

I am currently using some alligator clips with a homemade 5v charger from a bluetooth headset. I clip the wire to the soldered balance wire at the positive and negative sides of a parallel group and visually check with a volt meter. That make sense...
 
It has been a bit since I posted here, but I wanted to say thank you to everyone who took the time to continually respond - fetcher and d8veh specifically.

This thread represents a long 3mnths of feeling very frustrated. In the end, I did learn a lot replacing the same sets of cells within multiple batteries.

Today, I have three functioning 48v batteries! It turns out after all of this, that the manufacturer of the boards indeed identified them as defective!! I can't believe it. I sent back three out of the four - one I just threw away pissed off - and, to their credit, they did reimburse me. I still can't believe that was my issue the entire time....

Anyways, for the moment, I don't have anything to put in here. Although, I do think reading it top to bottom would be very helpful for a new entrant into the field of building a battery.

Thanks again!
 
Very welcome. I learn something from these exercises also.

I have seen cases where the manufacturer accidentally loaded the wrong part into the parts placing machine and made a whole batch of boards with the exact same defect.
 
That's why I kept asking you to check them, but you never did the checks.

If you would do the tests people ask you to do, and report the results, you might not have so many problems for so long. :/


We don't ask you to do the checks because we think it's fun, we do it to help you help yourself find the problem.
 
Everything is working well still, but I can't seem to get past 4.07v when on the charger. Any thoughts on what, if anything, I can do?

I opened up the case and all 13 groups are at 4.07.
 
Okay. I am using a few different chargers, but they are all cheap. One is 'fanless' the other is a little grey aluminum one.

Fanless looks like it is not able to be opened...

Could it also be a function of a cheap/non-exact board? Or a used/fatigued cell?
 
What happens when it gets up to 4.07V?

If the charge current suddenly goes to zero, then I would guess a cell is going high enough to trigger the BMS. You'd have to try measuring cell voltages right before it cuts off.

You can also measure what's coming out of the charger with no battery connected.
 
Regarding my last post about things not getting to 4.20v. I think this could have been being triggered by two separate cells within a group popped off their welds?

If that made two groups only a 2P config, would they then get to 4.20 quicker than the rest? I am just guessing.

Separately, if I am trying to use my own 5v plug and wires to charge a single group - different battery that had one group not like the others - do I need to disconnect the pcb balance wires?
 
Unplugged the board and got the cells charged up.

Now, I have something new I have not seen before. Battery charged to 53.4 if I measure B- to B+ or through the charging connector. However, through the discharge connector I am only getting 4.9v? What gives?

Thanks.
 
Boy, nothing but problems...

If the discharge port is turned off, it would usually be from one cell below the cutoff. Make sure all the balance wires have continuity. If none of the cells are actually low, you could try charging through the discharge port briefly to see if you can reset it.
 
Hi Again and Happy Holidays,

I am still working on the third of three batteries. Battery #2 is running great.

For Battery #3, the description of the current situation is below.

Full charge is 53.5v - this is 13s ncr-b. The bike rides for 3-4 miles and then the whole thing shuts off. VLC on controller is 39v. If I plug/unplug the battery I can get another .5miles.

I did see a number of sloppy spot welds. I redid and used a 4.5v bench supply to charage up individual series groups.

For Testing, I have the battery open while connecting it to an ebike on a stand. Without much load it is drawing 0.8a. Two series groups are registering 3.94 with this minimal load. The best series (s8) is still at 4.04v at this point.

Any thoughts? I am wondering if maybe a single cell in a series group is bad?

I will try to run it down to the point where it shuts off and monitor the groups I have identified as suspect and report back.
 
Update: Rode the bike until it died. Letting it cycle on the stand was going nowhere...

We then put it back on the stand and squeezed the brake to get some amps pushing.

I tested individual groups and located one in particular that was sagging heavily. With the voltmeter on this group, we watched it as it went down to 2.56 and then the whole battery shutoff.

Luckily, it is on the outside so I am going to disconnect that group and check the individual cells.

I am thinking one or more of them needs to be replaced?

Mario
 
I am posting a duplicate of the below here in the hopes I can get some input from fetcher and others.

Not totally sure how these forums behave so here is the other thread: https://endless-sphere.com/forums/viewtopic.php?f=2&t=92191

----

I rode battery #3 today into work. I am at <3ah on a Cycle Analyst and already getting down to a lower limit of 42v with full load of 14a. It then stabilizes around 44v while cruising at 8-9a.

Also to re-state, I can't get past 52.7v on a full charge at the moment.

I have lvc on controller at 40v. I was thinking I would ride it home - as far as I can get - to see if I can get the controller based lvc to trigger. If this trips before the bms then I am thinking there isn't a directly flagrant block that is dead i.e. nothing has crashed to 2.5v yet.

In regards to the BMS, I have been able to gather two data points.
#1 - I used a volt meter to find a block that kept crashing. The BMS would trip when the voltage for that block hit 2.5v.
#2 - A friend put the bms on a bench and saw that it cuts off at 33v and then turns back on when voltage goes back up to 40v.

Once I am home, I will open it up and take a look at individual block voltage: at rest, under load, after charging is complete.

Any thoughts/ideas would be appreciated.
 
It pretty much sounds like one cell group is possibly bad. As you have found, it's hard to test each individual cell group when the pack is assembled. Since you have several packs with problems, it may be worth building a test setup that does individual cells. If you can charge and discharge individual cells and get some kind of capacity measurement, you can determine which cell groups are good and bad.

It sounds like you have a charger that can do a single cell. For Li-ion cells, you could charge to 4.2v until the current drops off to assure full charge. Then discharge with something that can measure capacity. If you have a CycleAnalyst, it could be configured to do this. I found several inexpensive chinese solar charge meters that measure amp-hours that work well to test batteries.

From the description, the BMS is doing what it is supposed to. If not, you'd probably have a fire by now.
 
"If you can charge and discharge individual cells and get some kind of capacity measurement, you can determine which cell groups are good and bad."

I have a standard dual single cell charger and I have a little ebay board that discharges at .5a and gives a readout of how many mah consumed.

For blocks/cells inside the pack, I can't use this though right. I would need to physically disassemble a block and then test each cell one at a time. I did this and it said the last cell I repaired was done.

When I just checked the blocks one at a time, some were three blocks pulling down to 2.8v while others were staying at 3.3-3.4v. The 13s3p pack only registered like 3.5ah on the cycle analyst. I was looking for 9.6+ for 3p ncr-b. Again, couldn't get them past like 52.7v when fully charged.

So... Without really knowing, I will charge them up again and record what voltage they top out at. Then do another test.

I am also thinking I need to lower the 14a limit on my controller for a bit.

As always, any help is appreciated.
 
Back
Top