LiFePO4 Battery only has power when on charge?

[andyholloway]

Ok, so first up hopefully this is in the right section, mods please move if necessary.
I have a chinese 1000W (48V) hub motor with a 48V20AH pack (from ebay) and a WattsUp Meter. Never had a problem with it, then stopped using it for about 2-3 months, and just a few days ago tried to charge it up and it seems to charge fine, Watts up shows 59.5V when charged, and i can then turn it on and twist the throttle and the wheel will spin, but as soon as i turn off the charger the voltage drops to 0 over about 2 seconds (as if i had just disconnected the battery).
I can do up a quick schematic to show my wirnig in MS paint if that will help someone understand what might be wrong. Its just weird to me that it used to work fine and now doesn't.
Any help would be greatly appreciated!

 

[Russell]

It sounds like the BMS is tripping, perhaps because of a low cell. I suggest you check the voltage of each cell of the battery which can be done on the BMS or on the multi-pin connector from the battery to the BMS.

-R

 

[andyholloway]

thanks for the quick reply Russell. did a quick search and am not where on the bms i would test to find the cell voltages. a reference page or photo would be great. can i break something if i put the multimeter pins on the wrong spots? if i am at no risk of breaking something i will just have a poke around and hopefully work it out.

 

[dogman dan]

The bms wires, a bunch of skinny ones, usually enter a plug where a multimeter can probe. Or if the bms is not shrink wrapped, you can probe where they solder to the board. One lead on the end one, and the other will test each one down the line. You have to do some simple math along the way. If you've found the right end of the plug, the first two will read + 3. xx volts, then leaving the one probe in the first wire, move the second probe down one wire. It should now read about double the first one, so 6.xx or 7 volts. And so on.

Unfortunately, the sellers don't mention to you that leaving a pack laying around all winter without recharging it can kill a cell group. The bms runs all the time on 3.5v, hooked to one of the cell groups. After a few months the bms, which is intended to save the pack, draws down that group of cells till they die. Now you most likely have a pack with one group of cells that are gonners. If it's a round cell pack, it's very difficult to repair.

 

[dnmun]

you need to take pictures and show us what you are measuring. no way we can guess blindly.

 

[Russell]

When checking directly off the ribbon cable connector you need to be very careful not to accidentally short the meter leads together which is the only way you can mess up when taking a voltage reading. I use my sharpest set of leads and keep them angled away from each other.

Without knowing exactly what BMS you have I can't tell you exactly how to measure for the first or last cell but after that it's a simple process of measuring adjacent pins then shifting over one to take another reading and so forth. Measuring the first and last cells may be different depending how your battery is wired to your BMS; with mine there are 15 wires in 2 ribbon cables, one 11 pins the other 4 pins.





For me to check cell #1 I therefore have to put the negative probe on the big black lead from the battery and the red lead on pin 1 of the 11-pin connector





To check cell #2 I shift that positive red probe over to pin 2 and move the black negative probe to pin 1 on the 11-pin connector then I keep shifting the probes over one to get the next cell reading until I get to the end when I have the black probe on the last pin of the 4-pin connector and the positive lead on the red wire coming from the battery.

If you have a single 16-pin connector like the one below then as you can see from the wiring diagram you can measure cell #16 with the positive meter lead on the first pin (labeled 16) and the negative lead on the adjacent pin. When you get to the last cell, the one some forget, you would have the positive lead on the last pin (pin#1) of the connector and the negative lead on the BMS (B-) pad (the black battery lead).








Hope this helps.

Russ

 

[andyholloway]

thanks again to Russell and to dogman for your help. Hopefully tomorrow i will get a chance to try this.

 

[andyholloway]

Better late than never, i finally got the battery back out of the back of the cupboard to give this a go... Yes this battery has been sitting around another 5 or so months AFTER I left it sitting around too long, I never learn!
I have been a bit scared to do this because of the risk of damaging it further, I finally realised that it doesn't work now so i can't really make it any worse!


My BMS is exactly the same as the first one Russell posted. So I followed his instructions exactly and for the first cell I got 2.7V.
Then I used dogman's method of leaving the black lead from the multimeter on the big black wire on the battery and moved the red lead down the line, and the results i got (including the first one I already tested):
2.7, 5.4, 7.9, 10.5, 13.0, 15.5, 18.1, 20.6, 23.3, 26.0, 28.7, 31.6, 34.5, 37.5, 40.6, and then finally with the red lead in the big red wire: 43.8

so from this it seems all groups of cells are at 2.5-3.2V?

Does this suggest there is not one particular group of cells that has failed?

Thanks again to everyone who has been helping.

Ok, now I was just about to press 'Submit' and i decided to reread my first post, I had forgotten that I was able to make it work with the charger plugged in. So, just now I took the battery back to the bike and hooked it all up:

With no charger plugged in, the WattsUp shows around 44V! It hadn't worked with no charger last time I tried, so I twisted the throttle and it goes

So I plugged the charger in to charge it up and the charger shows green LED (finished charging), WattsUp shows 59-60V.

Next, I unplugged the charger and WattsUp shows voltage dropping over about 2 secs down to 0V and WattsUp turns off as it has no power supply now.

Then, I unplugged and replugged the battery and again WattsUp shows ~44V and it works again. I have put it back on charge even though it doesn't seem to be charging, will check back in a couple of hours...

Why did it drop to 0V when I unplug the charger, then if I simply replug the battery its back at 44V?

 

[andyholloway]

Ok, I got impatient and had another fiddle, expecting to be able to take it for a quick test ride.

Now this is just weird...
It IS now charging, WattsUp shows about 51V now and slowly rising, good sign right? Wrong!
I unplugged the charger and voltage dropped to 0V as expected, so as before, I unplugged and replugged the battery. This time nothing. Now it wont go at all, although, it does still go when the charger is plugged in.
So I started it charging again... Hopefully someone has an idea what is going on :-S

EDIT:
Update: it has finished charging now and reads ~60V and i unplugged the charger and it stayed at that voltage, so i eagerly turned it around, rolled it down the drive and gave it some throttle and it instantly cut out.
I thought it must be a loose connection. I checked every connection was plugged and still no luck.
It's getting late now, so I will check the connections with a multimeter tomorrow.
Thanks again, any thoughts or ideas greatly appreciated!

 

[dogman dan]

Try charging it again, at least for a few miniuites, then lift the wheel off the ground and try the motor. If it works ok with no load, then you know the motor and controller is not likely the problem. But if it trips the bms as soon as you ride with weight on the wheel, then something is up with the battery.

Most likely you have a bad cell. Actually this is a string of small cells wired paralell inside the pack. Then the paralell strings are series connected to make 60v. What's happening is one or more strings of cells has some dead batteries in it. When you charge, one or two live ones in that string give you 3.5v and all looks normal. But when you apply a load, instead of ten small batteries taking the load, only two or three are left. This is much more than they can take, so the voltage on that string drops immediately to below the voltage the bms is set for, and the bms trips. Most likely the bad string of cells is the one that powers the BMS, and over a few months of storage, the bms has drained that string of cells enough to ruin some of them. This is a real common thing, and it's almost criminal that the vendors don't warn you about storing for long periods without unplugging the bms.

Fixing it is nearly impossible, the spot welded round cell packs are a real bear to replace a cell group in. They aren't made in any kind of modular fashion that would make replacing a string of cells easy. You probobally have a good bms, but it's doing what it's supposed to do, when one string can't take a load.

When taking the voltage, you can move both leads down the line, and get the voltage of each cell without doing the math. I think you do have some good cells in every group, but not enough to handle a load. Bummer, you are most likely looking at buying a whole new battery.

 

[dnmun]

no way to know anything until we know where you are placing the probes. doesn't seem like there is anything wrong with the pack, imo.

 

[dogman dan]

Yeah, I'm guessing for sure. But the symptoms do match for one cell group having no capacity left.

 

[999zip999]

Dogman,Very helpful. I though a battery could sit for 5mos.

 

[dogman dan]

It can with nothing connected. But a bms draws power, usually just a tiny trickle, but from just one group of cells. A few months later, it's dead. There is no LVC on the part of the bms that draws the power to run the bms.

That's why now you have to plug in the bms when you get a new battery. Some that got delayed in shipping have arrived with a dead cell. So now they ship em with the bms disconnected.

 

[andyholloway]

Thanks dogman, it all seems to make sense to me now.

I preferred testing using the maths method so i didn't have to have the multimeter leads so close that I could risk shorting any cells.

I tested the whole battery voltage again this morning and it is down to around 44V again. I put it on charge again, charged it up to about 58V and it works with the wheel off the ground. Then I tried it while sitting on it and, as you predicted, it cut out instantly.

I also can't get it to work at all without the charger plugged in. And even straight off the charge the voltage is again down to 44V.

I will open up the battery to at least see what i am up against to find the dead cell(s) and replace them. I've got the time these days to patiently do it, and I don't have the money for a new pack, so I guess I've got to try.

 

[heroic]

Measure the voltages with the charger unplugged and the Watts Up showing the fault.