+ battery connector melted together

[jasonf150]

What does it mean when my positive batter connector (going from the battery to the controller) is melted together. It's not horrible, but enough to where I can't pull the connection apart. I can see where the plastic of the connector is melted a little.

52.8v, 9.2aH A123
40a 12fet controller (CA limited to 30a)
hs3540 motor

I ride 30 minutes twice a day, avg about 26mph. I stop pretty seldom thoughout that 30 minute ride, so I guess I ride it pretty hard, but I don't think I ride anywhere near as hard as some of the guys running closer to 100+amps. I got the battery from http://www.emissions-free.com and it does have the precharge (red/black/white wires) setup.


On a most-likely unrelated note, my chargers charge to a different voltage. I have one at work and one at home. When I leave my house in the morning, my cycle analyst reads ~55v, but after charging at work, it reads ~58v. Is that a problem? They are both 48v lifepo4 chargers which I also got from emissions-free.

Thanks for any help.
Jason

 

[Russell]

The power connector is either of poor quality, inadequately sized, not properly installed on the wire or not fully seated however without being able to inspect the connection it's impossible to say. FWIW I immediately change the input power and phase connectors to andersons. I also crimp then solder all connections. I don't use high power levels but I have never had a problem with any connections I have made.

LiFePO4 battery chemistry is different from most other lithium chemistries in that they can be charged to a voltage well above fully charged with no harm. This extra voltage often referred to as a "surface charge" will then bleed off over time. If the pack is equipped with a "bleeding" type BMS, and most are, then the shunts on the circuit board will discharge each cell to a predetermined level thus balancing the pack. To do this effectively the charger needs to initially charge the battery above the shunt voltage.

An example:

LiFepo4 cells are effectively fully charged at around 3.45V or 55.2V for a 16 cell pack. The shunts on a BMS may be set to balance at 3.60V (varies) so the charger needs to output at least 3.65V/cell or 58.4V for a 16 cell pack. Once the charge voltage is removed the shunts bleed each cell to 3.60V or 57.6V and the pack will be balanced. When left to sit the remaining surface charge will dissipate and the pack will stabilize around the 55V level.



-R

 

[jasonf150]

Thanks for the reply. I guess I need to look into redoing the connector.

As far as the battery charge level...now tonight it only charged to 51.4v. What's up with that? It's not even nominal voltage. That means it's not charged right? I don't want to get stuck on the way to work tomorrow. The green light on the charger is coming on and the charger shuts off.

 

[dnmun]

the charger turns off when the BMS stops the charging because one cell reaches 3.9V. the pack is not balanced so it has low cells that are not charged and high cells that are full that are now shutting off the BMS charging mosfet.

the connector melted because it is a bad connection with high resistance that gets hot when current flows through it.

 

[jasonf150]

so you're telling me the battery pack needs to be disassembled and the individual cells balanced?

 

[dnmun]

you said the pack stopped charging at 51V because the charger shut off. the charger shut off when the BMS detected HVC on one cell and shut off the charging mosfet. you can do like others and just leave it on the charger. or you can do what you wish. nobody ever measures cell voltages so it is pointless to try to give anybody advice anyway. just pointless.

 

[dogman dan]

Frustrating isn't it.

At a minimum, replace the battery connector. It melted because of a poor contact. The spark you get connecting it tends to cause a poor contact later, so that could be part of the problem. One reason I like anderson powerpoles for the battery to controller connection, is that the design of that connector minimizes the loss of good contact from that spark. The connector tip gets pitted, but the contact is made on the flat of the blade further in, which stays good.

Do leave your battery on the charger overnight more, if you don't already. It takes an astonishingly long time for a bms to balance a battery once it's not balanced well. You can fix it faster, if you listen to advice.

You might want to use the 58v charger overnight, as well. 56-60v is the range for lifepo4 16s chargers. But the thing might balance a bit faster at 58v.

 

[jasonf150]

I think my last comment came across the wrong way. It was meant like, "awe crap, it needs to be taken apart and reassembled", not like, "you're crazy, you think it needs to be taken apart". I do appreciate any advice you guys give me, please don't take that last post the wrong way.

I was under the impression that (at least with my model charger) that when the green light came on and it turned off, at least it sounds like it turns off, that the charger has shut down and stopped doing anything. But it seems like what you're saying is that it's doing something even after that.

Also, I don't leave the charger on the bike over night. Once I notice that green light come on, I turn it off with the switch and unplug it from the bike. I'll take your advice and leave it on longer. BTW - it's a 58.4v 5a charger from cell man. I'm sorry I called it a 48v charger.

This morning, I was still able to ride all the way to work on power. I used most of the aH so it seems as if it was charged *enough*.

Again, I'm pretty new at this, only about 90 cycles in on my first ebike. I do appreciate any advice or suggestions you guys give and I'm sorry if that last post came off the wrong way.

Jason

 

[docnjoj]

Hi jasonf150
I have 2 Pings one almost 4 years old and the other about 2.5 years. I leave them on the charger when I am not riding. They stay well balanced. If you dont give them a chance to balance with fairly long charge cycles they can become unbalance. You could charge the low cells with a 3.65 Single cell charger but it is worth a try to leave it on charge for a few days. I seems very safe to me and may help.
otherDoc

 

[dogman dan]

The way the bms works, when the voltage gets high enough, the charger shuts off. At this point, some cells are slightly overcharged and the bms slowly discharges just the highest ones. Eventually the voltage drops, and the charger will turn back on, often for a very very brief time. Not easy to even catch it doing it. So leaving it on overnight is nearly always required after a deep discharge. But if you use less than 80%, you might be able to fully charge without needing to balance.

Nothing wrong at all though, with checking the voltage of all your cell groups one by one from time to time. If one group l is way low, it's good to know which one, and keep an eye on just that group till it balances better.

 

[jasonf150]

Dang! I wasn't aware of that practice. I discharge over 80% almost every time. And I thought I was doing good to remove the charger once I saw the charger turn off. Geeze. Going to start doing as you suggest today!

 

[jasonf150]

OK, I apologize for not doing more testing the other night. Here is what seems to be the case.

First off, the battery IS in fact charging full. When tested at the battery, it measures the normal fully-charged voltage of about 58.4 So the chargers are both fine and doing their job.

The issue is that the CA is reading low, like <52.1v, even though the battery is at 58.4v. And the reason I couldn't ride to work this morning, is because the CA was hitting it's LVC. When I twist the throttle it reads like 35v.

So the question is why is the CA reading the voltage incorrectly? I checked all the connections and the melted positive is the only one that's fishy. Hopefully cellman will be sending me some new connectors shortly. Could that be it? The melted connector is not allowing the full voltage to get to the CA? And maybe it's been a bad connection for a while and that caused the melting in the first place?

 

[dnmun]

what does the CA read when it is not flowing current to the motor? it should be the same as the 58V you measured on the battery, except it drops when you remove it from the charger.

the bad connection most probably did eat up the voltage when it was heating up. and the batteries sag under load too.

does your controller have the on off switch? you should be able to leave everything connected if the controller has a switch, and even solder the connection permanently if the pack lives on the bike.

 

[jasonf150]

yes the controller has a switch with the charger not connected its still reading 51.9v right now. i am going to solder the positive where the bad connector is and *hopefully* that solves it.

 

[dnmun]

it should not read 51V on the CA if the voltmeter reads 58V. you may need to charge the pack longer to get it balanced if it's new.

 

[dogman dan]

Do you still have anything flaky between that battery and the controller? could you just have a bad contact, where the controller plugs into the CA dropping it's voltage?

Are you sure your voltmeter reads right?

 

[Ykick]

Are you sure your voltmeter reads right?

Yes, have you checked the battery inside the volt meter (DVM/DMM)? We've seen that before....

 

[jasonf150]

Both chargers are charging to full capacity just fine, although I will start leaving it on the charger overnight to ensure a good balance. (verified by the voltmeter on the battery itself)

I cut out the bad connector and soldered the wires together. Now the CA reads the same as the voltmeter straight to the battery.

I'll check the battery in the voltmeter. It's just a cheap one I got from who-knows-where and I don't think I've ever changed the battery in it.

The connector going from the controller to the CA *seems* ok. I did unplug it and plug it back in to check. That didn't seem to change anything.

Jason

 

[dogman dan]

Well there you go. I had been more or less assuming you'd fixed the melty connector. Quite a voltage drop there eh?

 

[dnmun]

you were misinformed when someone told you that the cells are fully charged at 3.4V. you have to keep the battery on the charger and charge it up until all the cells have at least 3.65V when measured on the BMS where the sense wires plug in. this has to be verified and you need to list the voltages here and a picture of where you measured. you have to measure where the sense wires plug into the BMS while the pack is charging, and actually measure when it gets to the point where the charger finally is turning on and off to balance the pack.

 

[Evoforce]

Ok, I'm glad you found the problem. I just found this post. Regarding the 80% discharge of your battery, the BMS is supposed to shut off power when you reach this amount of discharge on your battery. It is designed to keep you from over discharging your battery pack. Just found you on the members map. Let's get together sometime. Amberwolf doesn't live too far from you either. Lcrewes is also in the metro area. Connector are soo often the problem. Look up Ham Radio Outlet in North Phoenix. They carry red and black Anderson connectors. They are the only ones in the valley that I have found to actually stock them. otherwise it is online and I order from Powerwerx when I need different colored ones. Commuting to work on your electric...good job!

 

[Russell]

you were misinformed when someone told you that the cells are fully charged at 3.4V. you have to keep the battery on the charger and charge it up until all the cells have at least 3.65V when measured on the BMS where the sense wires plug in. this has to be verified and you need to list the voltages here and a picture of where you measured. you have to measure where the sense wires plug into the BMS while the pack is charging, and actually measure when it gets to the point where the charger finally is turning on and off to balance the pack.

LiFePO4 cells are at 99.5% capacity at ~3.45V. Most manufacturer's do spec 3.65V as the charging voltage however that last 0.2V adds only about 2 Wh on a typical 48V/10Ah (480 Wh) pack. However there is a reason to use a higher charge voltage when a pack is equipped with a BMS, but just how high depends on where the shumts are designed to activate. If the shunts are set to divert energy at 3.60V then of course you need to charge to at least 3.65V/cell. Since there is so little actual extra energy in the cells at this voltage level rather small and cheap components can be used on the BMS.

So while the convention for charging LiFePO4 is 3.65V/cell it really depends on the design of the BMS. I have read a couple of posts on this forum where the person said their BMS was set for a lower balance point and their charger set for a lower voltage. Also if the pack has a more robust BMS with bigger shunts then a higher voltage can be used which will balance the cells quicker.

After the charging voltage is removed and the shunts have done their job and the pack is balanced the voltage will continue to fall over time until what surface charge there is left dissipates. After several days or perhaps a little longer the voltage for a 16 cell pack will be at around 55V (Mid 3.4V/cell range) and this is fully charged.

-R

 

[Russell]

Ok, I'm glad you found the problem. I just found this post. Regarding the 80% discharge of your battery, the BMS is supposed to shut off power when you reach this amount of discharge on your battery. It is designed to keep you from over discharging your battery pack.

Not exactly. An 80% DOD is often recommended to extend pack life but that is not the limit a BMS uses. The BMS will trip if any one cell falls below a preset voltage or if the pack voltage falls below a preset voltage or if the current limit is exceeded. So if the pack is very well balanced you can get more out of it than 80% especially at low discharge currents. For example I drained a BMS equipped LiFePO4 pack to 90-93% DOD a couple of times without tripping the BMS. Conversely if just one cell is weak or badly out of balance the BMS could trip well before an 80% DOD.

-R

 

[dnmun]

again misinformation. 3.4V is not charged. it is not close to being charged. if you charge your battery to only 55V it will not balance. you will cause the pack to become more imbalanced with every cycle until you can force it back into balance by adjusting the voltage even higher than the 59V or so it takes to balance them properly.

for example, with headway cells that are resting at 3.34V after sitting for years self discharging, as soon as i put it on the charger the voltage immediately goes to 3.41V even though the cell is essentially totally discharged. i have to charge it for hours with 5 others on a 10A charger to get it to 3.61V and then even longer until it goes to 3.65V.

whoever wrote that 3.45V is 99.45% charged has never charged up a lifepo4 cell.

 

[Russell]

again misinformation. 3.4V is not charged. it is not close to being charged. if you charge your battery to only 55V it will not balance. you will cause the pack to become more imbalanced with every cycle until you can force it back into balance by adjusting the voltage even higher than the 59V or so it takes to balance them properly.

for example, with headway cells that are resting at 3.34V after sitting for years self discharging, as soon as i put it on the charger the voltage immediately goes to 3.41V even though the cell is essentially totally discharged. i have to charge it for hours with 5 others on a 10A charger to get it to 3.61V and then even longer until it goes to 3.65V.

whoever wrote that 3.45V is 99.45% charged has never charged up a lifepo4 cell.

No, not misinformation at all, I did the experiments and I can assure you that a LiFePO4 cell is essentially full when it reaches 3.45V and the charger current is near zero. The key here, and I think where your confusion is, is that you should not go by the charging voltage. To prove this to yourself you need to set the charger voltage to 3.45V then charge a cell until the charging current approaches 0A. Then set the charger voltage to 3.65V and connect it to the cell again with a wattmeter in-line. Charge to 3.65V and you will see just how little energy the cell accepts between 3.45V and 3.65V (it's tiny, about 0.5%).

With that said once again I'll repeat if you want to effectively balance your pack equipped with a BMS you need to exceed the shunt voltage for your BMS however balancing and fully charged are two different things. For example if your BMS shunts are set to 3.60V then for a 16S pack you want your charging voltage to be 58.4-59.2V (3.65-3.70V/cell) or even a little higher if you have good sized shunt resistors. Once the charging cycle is complete, that is battery voltage essentially equals the charger voltage, and the current is below the charger setpoint (mine is around 30mA) the light on the charger turns green. Remove the charger and the shunts continue to bleed the individual cells. After 15-30 minutes the pack voltage will be 57.6V (assuming perfectly balanced 16S LiFePO4 pack w/3.60V shunts on BMS). Left to rest the pack voltage will continue to fall as the excess "surface charge" dissipates.

Actually without doing the charging experiment I mentioned in the first paragraph most people who regularly check their LiFePO4 pack voltage knows this already; the charger voltage is one thing, say 58.4V, the voltage after the BMS is done balancing is another (say 57.6V) and the resting pack voltage after a week (or after traveling a block or two under power) is another (~55V) but the pack is still 99.5% full.

-R