Ok, I bought a (slightly) used duct tape (Vpower) battery...

[sangesf]

Quick update..

Well, I still have the pack, but my Turnigy watt meter crapped out and I wasn't able to test for a while..
I also didn't run the battery down past 15 miles or so (about 10ah)...

Well yesterday I charged it up an went for as long of a ride, until it cut off...

So at about 20 miles is shuts down..

I charge back up and it seems that there is only one cell group that is not getting high in charge (only 3.35v)... I put it on a small charger (4v 500ma) and it charges for about 5 hours and finally sits at 3.62v no charge load.. So I run the bike to LVC again and I see that cell group is 2.7v
I charge again.... And all the others sit at 3.6x to 3.7c and that last cell group again only shows 3.35.. So I put it on the charger again and I'm waiting for it to get to 3.6 or more as we speak...

I finally got another watt meter so I'll run it again to LVC and see what AH I get out of that cels group...

It seems weird that, that one cell group will discharge relative well, but on charge it doesn't quite reach above 3.40 unless I single charge it...
I'm guessing the BMS doesn't like charging that cell group properly, for some reason...
Any ideas?

 

[amberwolf]

If that cell group has a leaky cell (or more than one), some of the current that would go to charging the group instead just gets thrown away. So it takes longer to charge that group than the others. Normally, you'd have to leave it on the charger a long (maybe long long long) time so the BMS can drain down the other groups below HVC, then the charger kick in again, add a bit more, and cycle doing that until the last group catches up.

But with a *leaky* cell, it's actually actively draining the other cells at a rate that's as fast or faster than the top-off cycling can manage, so that group would in reality never finish charging, and possibly get *less* charged, over the days or whatever time it takes the low-balance-drain-rate BMS to cyclically take the surface charge off the good groups and charge for a few minutes more, sit for hours draining, etc.

That's my guess based on the 2 Volgood and Vpower/CammyCC packs I've fixed so far.


Other possibilities for the problem include that all the other groups are higher resistance than the low group, so that during charge they seem to rise up to a higher voltage while the other one doesn't, so the BMS cuts off. But usually that would result in less charge in those, so that they would hit LVC first. Doesnt' sound like that's what's happening, especially since you've single-group charged that one group to "full", and it still hit LVC first. It's more likely a leaky cell.


I suppose something could be wrong with that BMS channel for it's HVC, but since all cells are charged in series it would also be cutting off early on the entire pack, since it can't control anything other than the *drainage* of charge off of any particular group--it doesn't actively control how much charge goes into each group. I'd say that the BMS channel for that group could be turning on it's drainage resistor early, or never turning it off, but you can test that just by feeling it or measuring it's temperature. During single cell charging, that resistor would start getting pretty warm when you're near full charge but not yet at shunting voltage, if it's turning on when it shouldn't.

 

[sangesf]

I notice that when charging that particular cell group (along with a spare 3.3v 20ah cell using an SLA 6v (7.2v - 2.7a/.7a charger) instead of "single cell charging".., that in the beginning of (2.7a) charging it hits 3.7-3.8v rather quickly, but when removed from charging it drops to 3.3xV immediately... So I leave it on there until it shows 3.9v for that group and then once pulled off it's at about 3.4x then I charge group at the 700ma mode until it hits 3.8x and then I know it will settle at 3.6x off charger and stay that way for a while..

What I personally think is happening with the BMS is (given that group's "actions" during the above charging method) that the BMS sees that cell group hit 3.7x or so during charging (more than likely just before the others) and then cuts that particular channel off until the others hit that same 3.7x mark (again, prolly very soon after) and then that "bad" group is now at 3.3x again (problem is that the entire pack as a whole is prolly at CHARGER max voltage (which makes charger go into "trickle mode") and only charges that group at 200ma which could take forever!

What does that say to you? Oh swammie of the lifepo4 charging... (BMS or Cell group)?
(I would open up the pack myself and try to fix that cell group (IF that's the problem), BUT!! I can't (and don't have) a way to re-solder new cells.. That's why I asked for your help in the other thread.

 

[amberwolf]

The BMS cannot cut any channel off from charging. All cells are in series and charge at the same time.

The BMS can only cut off *all* charge, and then drain cells that are over whatever it's HVC limit is.

 

[sangesf]

So what that says to me, is that, that one cell group could possibly hit 4v or even much higher IF the others are still at 3.3x during charging?!?
That kinda makes a BMS useless for charging? Wait.. That doesn't make sense....
Then how does the BMS cut power when that one particular group hits it's LVC (2.5-2.7v) before the others?

Would make sense that the BMS sees that particular cell group hit it's HVC and then would cut power from the charger...

BUT I know that at HVC (about 45.xV or so) when I pull the charger off I see a pack voltage of anywhere between 43.8 to 44.4.

 

[sangesf]

The BMS cannot cut any channel off from charging. All cells are in series and charge at the same time.

The BMS can only cut off *all* charge, and then drain cells that are over whatever it's HVC limit is.

I guess what you're saying is that once the HVC of THAT cell group is hit, it shuts off charging?!?

 

[amberwolf]

More or less, yes. There's two levels to the BMS function, at least as far as I know. Not all BMSs work this way, either.
1--Any cell group hitting the point at which it is fully charged would trigger the shunt for that group/channel, so that the low charging current at that point is passed "around" the cell group rather than thru it. This can happen for one or all cell groups without stopping charge.

2--The first cell group to actually reach the HVC would cause the BMS to turn off it's charging FETs (which in the case of the two packs I've worked on so far is also the discharging FETs, AFAICT).

So if that self-discharging group is internally high resistance, so that as it nears full charge it begins to have a higher voltage across it than normal, while charge current is flowing thru it, it not only will turn on it's channel shunt, but also could if it goes high enough trigger the HVC and prevent the rest of the pack from finishing charge on the first go-around.

Once charge current stops flowing thru it, it's voltage will drop, and the BMS could allow charge current thru again, so it would cycle on and off.

How that part works also depends on the charger. A "dumb" charger would just keep outputting the constant voltage of max pack voltage, so the BMS cutting in and out will simply interrupt voltage to the pack or not. But a "smart" charger might detect that there's no current flowing anymore (once the BMS shuts off current to the pack), and turn itself off. Some of the smart chargers also turn back on once reconnected to the pack, but may have a hysteresis in voltage amount, meaning that they must see the load voltage change downward by some amount before it will turn charging back on.

The charger for the Vpower/CammyCC pack I have is like the latter. I think it has to drop a volt or so (maybe less, I'm not totally sure) below the full-charge voltage before it will turn back on again. Or I can unplug it from either the pack or the wall and reconnect it, and it'll turn on again.

 

[sangesf]

All I see at this point is...
When the pack is charging... Once it hits about 45.xV it stops allowing charge to go through.. If I pull the charger off immediately, the pack is close to that 45v. If I let it sit for a couple of hours after charging instead of pulling it off immediately, and then I pull off the charger, the pack reads anywhere from 42.9 to 43.8. If I plug the charger back in after that.. Most of the time, I have to wait a little bit before plugging the charger back in or the BMS won't let the charger start charging again. THEN about 2 min later, charger goes green (it's at 45.xV) and I read the pack voltage is again at about 43.xV. And if I want to charge again, I have to usually wait until pack voltage goes under 43v before it will allow the charger to start charging again...

What I'm confused about, is my other pack when it hits BMS HVC, stops charging and CONTINUALLY watches the charge and if one of the cell groups drops a bit in voltage, it allows the charger (same one I'm using for this pack) start a new charge cycle again.. Back and forth (charger light from red to green) as many times as needed...

It seems this (two-tiered) BMS just stops allowing charge altogether and doesn't start the "re-cycle" over again, unless I pull the load (charger) off for a bit as mentioned above...


Earlier when I was talking about "turning off the channel for that cell group".. I meant the same thing as when you said.. "will turn on it's channel shunt"

 

[999zip999]

My guess is you have two or more bad cells in that parelle group that's bad.

 

[sangesf]

Prolly... I haven't tested full range yet after charging to full.. Haven't had time..
When I've charged the battery (and that cell group singularly), I will test the capable AH of it... If it's close to 15ah.. I'll just leave it... If it's closer to 10ah, I'll probably have to pull that cell group and replace the bad cells..

 

[amberwolf]

You should also do a let-it-sit test: Charge it up, disconnect it from the charger. Wait a day, disconnect it from the BMS balance plug, and measure the cell groups at the balance wires.

Leave it disconnected from charger, bike, and BMS.

Each day, measure it again, noting it down each time.

Do this for at least two weeks, without using or charging the battery in the meantime. If it stays up at the same voltage the whole time, there are probably no significantly leaky cells in the groups.

If any group begins to show voltage dropping, it is probably got at least one leaky cell in that group, and if you don't at least disconnect that cell from the rest, it's going to drain charge from the rest of the cells in that group at what might be a significant rate. If the cell's internal resistance gets low enough I suppose it could even be dangerous to leave connected. (doesnt' matter if it's still in the pack while disconnected, but it most definitely matters if you keep cycling power thru it).

It's also probably going to keep that group from ever really fully charging, when on the regular charger/BMS. You'll wind up having to always manually charge each group that has bad cells in it, and do that not long before you want to use the pack, too (or else it will self-discharge).

Eventually the cell(s) will be bad enough to drain charge in just a few hours, which is what happened to me one day with my Vpower/CammyCC pack, on the last repair I had to do to it. One day I rode to work just fine, but a few hours later the cell group had drained below LVC and I could not ride home. Of course, I had a load of dogfood, too, so it was extra sucky timing.