Strange LiFePO4 behavior

[jdcburg]

I have noticed something very strange happening to my Thundersky LiFePO4s. They are the 20ah cells from Elite Power. I have 2 4-packs for a nominal 24V powering a 350W brushed Unite motor. I’ve had them about 3 months and have gone about 350 miles. The most I’ve discharged them is about 14ah. I originally charged them with a 24V charger and topped them with a single cell charger but lately I have been using a balancing charger. The cells normally stay within a few mV of each other both while they’re in use and after use. I monitor them during use with a Celllog 8M and a Turnigy WattMeter. I normally use them at 0.5 to 1C with a max of 1.4C. In short, totally bland use.

Several times over the past 6 weeks (including 4 times this week) cells 1 and 2 have done a very strange thing. This evening I watched it happen. I was stopped at a light and I decided to pull over and make a phone call. All cells were at 3.32 or 3.33V. Within a 5 second period, cell 1 dropped to 2.8V and cell 2 went up to nearly 4V. I watched them on the Celllog for maybe 10 seconds. Just a touch of throttle brought both right back to 3.32. The rest of the ride home all the cells were very close to each other. After I got home I left the Celllog on for 20 minutes or so and there were no abnormalities. It always happens when the controller is on but there is no load. I’ve tried 2 different Celllogs so it’s not the instrument. Has anyone else seen anything like this? Should I do anything or just not worry about it? - jd

 

[icecube57]

I wouldnt worry about it. Kinda odd tho. But 2.8 under load is about right so maybe it displaying some old ghost data .

I know the cell log only powers off of 6 of the 8 cells too.

Check your balance connectors/ wires. Check them often.

 

[whatever]

assuming the meter readings are correct, there will be some explanation for this behaviour.
My first thought is that one cell is a bit dodgey ( maybe cell 2), you could try checking the internal resistance of the first 2 cells, to see if there is something there.
can you describe more the setup of cells 1 and 2?
are they at the positive or negative end of the battery string? if at the positive end, you always get some odd behaviour of the first couple of cells, since they are taking most of the current when riding or charging ( depending on charger of course)
Is cell 1 or 2, made up of a number of cells in parallel? my guess is theres going to be some internal resistance problem of either cell 1 or cell 2 ( all assuming the volt readings are correct).
A simple check of cell resistances maybe reveal some clues.
The power is on so some small current is flowing ( 65mA or less I guess), it appears that cell 2 is sucking up some current ( low resistance?) it might be taking that current from cell 1 causing its voltage drop, is there any heat in cell 2? I think the cells are trying to tell you something but just what it is I'm not sure, if you can confirm the voltages are correct, i would very much think to further investigate it to save any larger problems later on.
It would be very interesting to find exactly whats going on, its not normal behaviour in my experience, certainly cells at the positive end are more likely to have some voltage variation/lifespan issues than those at the negative end of the string, but this is very odd.
Are the first two cells or first cell being also used to power some other device?

 

[markcycle]

How is your controller case connected to the frame of the bike?
Is the negative of the pack connected to the frame of the bike?
It could be one of the leads on the cellog are high impedance but not open.
The controller could be oscillating at a very low level at zero throttle and through stray capacitance charge pumping the ground level causing weird cellog readings.

 

[jdcburg]

Thanks to all for the replies. I’ll try to answer the questions raised.

To whatever - I haven’t tried to check the resistance of any of the cells. Can I just switch my multimeter to ohms and touch the terminals?

I believe cells 1&2 are at the negative end of the pack. None of the cells are warm to touch, whether in normal use or when they “act out.” The abnormal readings come and go so quickly (< 5 seconds) I have to believe it is some kind of surface charge.

The TS packs are made up of 4 single rectangular cells 6”x1.5”x2.75”. The screw terminals are on the tops. I believe the volt readings are correct. During charging my balancing charger, a Turnigy Accucel 8150, shows essentially the same readings as the Celllog. Also I checked the first Celllog against my digital multimeter. Both Celllogs have shown the abnormal behavior at different times. I don’t believe it is an instrument issue.

The only devices in the system are the Turnigy WattMeter, the Celllog and the motor/controller/throttle.

To markcycle – The battery packs are plastic and my battery case is plywood. The controller is mounted on wood. I don’t think there are stray ground issues. I will check the continuity of each wire going to the Celllog sometime soon.

To dalian – I have not contacted TS yet but I will soon.

I will also check the wire continuity and cell resistance. Then I think my next step will be to switch the packs so the cells in question are in the middle of the pack. That will answer the question of whether it’s the cells or the position in the pack. Any other ideas, suggestions or questions? - jd

 

[texaspyro]

99% sure you have an intermittent cell connection to the balance connector. Note that the voltage on the low cell and high cell average out to be normal. Probably the balance wire/connection between these two cells is funky.

 

[fechter]

99% sure you have an intermittent cell connection to the balance connector. Note that the voltage on the low cell and high cell average out to be normal. Probably the balance wire/connection between these two cells is funky.

I agree. No way a cell is going to be 4v during discharge. If you add the two, the result stays about the same.

 

[GGoodrum]

+1 for the loose connection. Check the wire from the top of the 1st cell. My guess it is the wire end going into the JST-XH balance plug that goes into the CellLog.

-- Gary

 

[jdcburg]

I switched the series order of the 2 packs this weekend so the problem cells (1&2) are now 6&5. I also redid the some of connections, although the resistance was low in all of them. All the battery terminal connections were tight. I went for a short ride Sunday with no problems. Time will tell - jd

 

[GGoodrum]

The problem is not in the connection between cells. I'm sure your packs/cells are fine. It is only with the wire that goes to the CellLog. I've seen this exact same thing. It might even be a pushed pin in the JST-XH plug you are using.

 

[jdcburg]

What I meant to say was I checked all the wires and connections in the Celllog harness, and the terminal connections where the harness attaches to the individual cells of the pack. There didn't seem to be any that had high or variable resistance. I replaced a few crimped buttsplices that looked a little funky and changed a few others to adjust the length and rechecked the resistance. They all seem fine. I am carrying my multimeter in my repair kit now so if it happens again I can check cells and wires on the spot. I agree it probably was in the harness. Hopefully it's fixed now - jd