Pack: 16s15p comprised by (240) 2.2ah lithium 18650s from all-battery.
Controller: 35amp 72 volt crystalyte
Problem: The ammeter on my handlebars, verified by my sense of normal torque, says when just starting out the controller is getting/pulling a max 32 amps instead of the usual 35. After riding for 15 minutes or so, when I crank the throttle, amps go back up to 34-35.
By later in the ride, when the cells are down to ~50% capacity at 3.85 volts, max amps drops to 30 under load. During all these times, the voltage sag remains in it's normal range of 4-5 volts on a fresh charge to ~7 volts at 50% charge.
Considerations:
Lately it's been about 40F colder than in the summer when I first put the pack together. Could cold weather cause a dramatic increase in impedance in these lithium cells? If so, why doesn't the voltage sag under load also increase? Can I expect the cells to return to full power with increasing temperature?
About 6 months ago, I ruptured and killed two of these cells by running them too low under full load . The heat killed two dozen more nearby without rupturing them. I replaced these cells with new ones. Could I have also damaged other cells still in my pack, causing the problem I'm only seeing now, 6 months later? If so, what's the easiest way to find/isolate the damaged cells? I have only a DVM to test with. I could pull the 15-cell parallel-wired subpacks one-by-one. If the culprit is in one of these packs, can I expect full amps again when I take it out of the circuit (albeit at less voltage)?
If I continue to ride the bike this way, do I risk causing more damage?
Thank you for your time and help.
Here's pics of one of the two ruptured cells I replaced -- the heat killed 7 others in the top two rows I also replaced. I'm wondering if the heat damaged the four in the bottom row I'm still using such that it's causing a low-current problem for the entire pack.
Here's a pic of the pack charging after I resoldered it 15s15p, replacing the dead cells (this is before I added the 16th subpack):
"Problem: The ammeter on my handlebars, verified by my sense of normal torque, says when just starting out the controller is getting/pulling a max 32 amps instead of the usual 35. After riding for 15 minutes or so, when I crank the throttle, amps go back up to 34-35.
By later in the ride, when the cells are down to ~50% capacity at 3.85 volts, max amps drops to 30 under load. During all these times, the voltage sag remains in it's normal range of 4-5 volts on a fresh charge to ~7 volts at 50% charge"
Sounds like the temperature is reducing the discharge rate. Is it 40F outside or 40F colder than in the summer, if so what was the temp in the summer? An increase in the cell temperature could explain why you get the extra amps after you've been running for 15 minutes, as the temperature should go up on discharge. You might want to try recording the temp of the cells and seeing what effect it has. Your experience sounds similar to many of the people using lithium cobalt Thundersky type cells in cold weather, I don't think its something to worry about in itself, or is that you just miss the extra amps ; ).
"Could cold weather cause a dramatic increase in impedance in these lithium cells? If so, why doesn't the voltage sag under load also increase?"
There's more to the discharge rate of a battery than just the internal resistance and thus the visable voltage sag. Its a kinda complicated and I don't really understand it myself, maybe someone else will be able to help?
"About 6 months ago, I ruptured and killed two of these cells by running them too low under full load . The heat killed two dozen more nearby without rupturing them. I replaced these cells with new ones. Could I have also damaged other cells still in my pack, causing the problem I'm only seeing now, 6 months later? If so, what's the easiest way to find/isolate the damaged cells? I have only a DVM to test with. I could pull the 15-cell parallel-wired subpacks one-by-one. If the culprit is in one of these packs, can I expect full amps again when I take it out of the circuit (albeit at less voltage)?"
You could test the internal resistance subpacks at a constant load by finding the voltage drop (12 volt halogen bulbs are an easy way to get the desired resistance), if one pack has a higher resistance try to isolate the offending cells (don't forget to factor in temperature aswell).
What actually happened when you damaged your packs? Can you remember how the hot the rest of the packs were when when you had the problem. In otherwords did you damage them through sudden overcurrent or through slowly heating them up? or from discharging below the minimum voltage of 3 volts (I doubt this) From the info LG spec sheet it appears the cells have a PTC, perhaps that was what caused the fire. See here:
http://www.electrochem.org/dl/ma/203/pdfs/0189.pdf
You might want to join the Yahoo Thundesky group. There's a lot of people day to day experience of lithium batteries there and I imagine they'd be quite interested in your bike.
Hope this helps.
Nick
Sounds like the temperature is reducing the discharge rate. Is it 40F outside or 40F colder than in the summer, if so what was the temp in the summer?
I do hope it's the temperature. I'm going to move my bike inside from the garage tonight, then see tomorrow if that makes any difference. I should have added that the daytime temp lately has been about 40F -- 40 degrees cooler than a couple months ago when the cells were pumping 35 amps no problem.
What actually happened when you damaged your packs? Can you remember how the hot the rest of the packs were when when you had the problem. In otherwords did you damage them through sudden overcurrent or through slowly heating them up? or from discharging below the minimum
I thought from the spec's I could run the cells down to 3.3-3.5 volts no problem. So one night I was cruising around the hood with the cells around 3.7 volts in their old 19s12p config. I noticed the sudden onset of twice normal voltage sag; the bike subjectively losing it's "pep". So I rode home and opened the back battery box which at that time held all the cells. One string of 12 parallel cells read 2.5 volts! It was just at 3.7 a few minutes before. All the cells in that subpack were too hot to touch, so I disconnected that string and placed them in a pyrex container out on the cement to cool. One cell ruptured and smoked while cooling. After cooling, the other cells of the twelve read 0.0v, so I disposed of all twelve. Same thing happened again about a week later. I figured that under the 1.5C amp draw, the weakling cells in each parallel subpack, already at a low voltage, became overdrained. The subpack may have read 3.7v, but under load, perhaps the weakest cells were much lower. So I bought some more of the same batteries and resoldered the entire thing 16s15p, cutting the load per cell from about 1.5c to about 1c. I also make sure not to run the cells lower than 3.8 volts or so, assuming again that the stragglers under load are perhaps down to 3.7 by then. Later, I found this chart of lipoly resting voltages versus remaining capacity which is very different than the manufacturer's, but much more in line with my experience with these 18650's:
http://www.rcgroups.com/forums/showthread.php?t=209187
"4.20v = 100%
4.03v = 76%
3.86v = 52%
3.83v = 42%
3.79v = 30%
3.70v = 11%
3.6?v = 0%
These are resting voltages - not under load voltages."
I'm going to scope out that Thundersky group. I'll post back here my results of tomorrow's test.
Thanks!
--xyster
moderator
My rides
Hey Nick, thank you for referring me to the thundersky group. I think my questions have been answered. I pulled some quotes that may be of general interest.
My reasoning in regards to my two cell failures is corroborated:
In regards to diminished thundersky current capacity at low temps:
My cells, spec'd for 1.5C continuous, are delivering a max 0.9C at 40F.
After having my membership to the thundersky group "approved" :-) I feel like I'm a spy for V2.0 :-) Shhh....I don't need radioactive polonium showing up in my refrigerator!
--xyster
moderator
My rides
Just got back from a 7 mile test ride after leaving the bike (w/batteries) in the house overnight. At the start, the batteries delivered 32 amps instead yesterday's of 30 amps. After about 5 minutes of hard riding though, max current climbed back to the full 35 amps. So it must be the cold weather! Other stats: beginning cell voltage: 4.06 - 4.09. Ending cell voltage: 3.88 - 3.93. Top speed: 33mph.
Perhaps the cold does cause a bit more voltage droop too. During warm weather top speed with the cells at 4.0 volts was 35mph.
--xyster
moderator
My rides