cell_man
100 kW
Sag to 3V isn't so bad. I wouldn't worry about it. In EV applications the TS cells often get pulled a lot lower than that.
One Lifepo4 cell that won't hold surface charge
- Thread starter morph999
- Start date
Russell
1 MW
morph999 said:
Morph from what I can tell from the numbers your battery and the one suspect cell are looking fine. The sag is not bad at all either on the one cell or for the pack overall.
It appears there are folks on this board who want you to fail but hang in there it's all good!
-R
EDIT: BTW for comparison look at this video of a 36V/10Ah LiFePO4 battery under test, the voltage sags to 34.8V with a 19.95A load. http://endless-sphere.com/forums/viewtopic.php?f=4&t=8452&start=75#p131908
Whats up man 
If you want a motor better fitted to your batteries, I still have an extra 9c motor laying around. It will do about 18 mph @ 36v, but can be over-volted to do whatever speed you want. I would trade that and installation labor(the wires will not match your controller) for the 5303.
Also if you want to hang onto the 5303, once I get my lipo pack setup I can let you try it out to see if you want to consider that route(this is what I was telling you about a few months ago before you got the TS cells).
I wouldn't really be overly concerned about it like Cell-man said its fairly normal. But I would keep an eye on it, and using less amps would help the longevity.morph999 said:
It would have to be a trade, I just bought 55v x 15ah of Lipo and my fiance would kill me if I were to buy anything else :lol:
I really wouldn't recommend using that forsen motor, It would be better to go easy on the throttle with the 5303 IMO.
If you want a motor better fitted to your batteries, I still have an extra 9c motor laying around. It will do about 18 mph @ 36v, but can be over-volted to do whatever speed you want. I would trade that and installation labor(the wires will not match your controller) for the 5303.
Also if you want to hang onto the 5303, once I get my lipo pack setup I can let you try it out to see if you want to consider that route(this is what I was telling you about a few months ago before you got the TS cells).
I'm reading on other forums that this is normal, for the cell to bleed down to 3.36v after charging. Why my other cells aren't doing it is a mystery. Don't knock Jack Rickard, he knows his stuff. I think lithium is still a bit of a mystery because one group is saying that surface charge bleeding down to 3.35v is normal and another group is saying that it's not normal. Maybe if I continue running my batteries, we can find out if it's normal or not. If these last more than 500 cycles, then we'll know that it's pretty normal.
I don't ride my bike everyday so if these can just last 4 years, that would be awesome.
I don't ride my bike everyday so if these can just last 4 years, that would be awesome.
AussieJester
1 TW
Morph, is you controller programmable? if so perhaps simply turn down
the settings so you can't pull more from the batteries than they can put out?
Don't cycle analysts have current limiting on them? Would be cheap'er' option
if you have non programmable controller, cheaper than than buying new pakcs if these ones get damaged from pulling to high amps.
KiM
the settings so you can't pull more from the batteries than they can put out?
Don't cycle analysts have current limiting on them? Would be cheap'er' option
if you have non programmable controller, cheaper than than buying new pakcs if these ones get damaged from pulling to high amps.
KiM
AndyH
10 kW
morph999 said:
Maybe, maybe not. Who are the people telling you that bleed is normal? And what is their experience?
If you charge a cell to 3.8V and it stays near 3.8 - you're seeing a surface charge. The A123 cells will do that until damaged. I don't know of any others that will hold that surface charge - so they'll bleed down to some voltage. More damage or more aging appears to bring more bleed.
I've run automated logging of new and slightly used cells. All of these are at 20-25°C room temperature. Cell is charged to the manufacturer's charge point with proper CC/CV, then the cell sits for 24 hours. A Headway 10Ah blue cell with a couple of light cycles on it is at 3.45V at the 24 hour point. Highpower 100Ah brand new stays above 3.6V. Sky Energy 100Ah stays above 3.55V. Thunder Sky 100Ah stays above 3.55. PSI 10Ah will drop bel [edit] below 3.4V to about 3.35V. [/edit]
Please - find the real battery experts here and talk with them. You are connected to some in this thread, you were connected with some in the 'jack thread'. I have nearly three years of bench and road testing - and compared with most of the folks here I'm a n00b. I'm a member of 10 EV groups and forums - and this is the only group I've found where people actually log data, run tests, and want to know 'why'.
TS cells will sag a LOT under normal use, and will drop more deeply when cold. Here's what a 13 month old TS 60Ah pack looks like at 67°F. The controller can pull 100A max.
First - Let's look at the voltage at rest. This pack is monitored by a PakTrakr with three 8-cell remotes. The remotes are powered from the first four cells on each remote - so 12 cells in this 21S pack have a small load on them 24/7. Here's a view of the bleed-down for this pack. The lower cells have the PakTrakr load; the higher cells do not. The pack was charged to 3.7V per cell.
View attachment chart1.jpg
Here's a quick ride to the post office. Notice in the upper right corner that the LVC limited the current to 95A from a max of 100. This pack is fully charged and cells are hitting the 2.7V LVC with a bit less than 1.6C discharge rate. Also notice that 20 of 21 cells dropped below 3.0V under load -- and this pack is still nearly fully charged.
View attachment chart2.jpg
Good Luck.
Andy
If the cell was damaged, wouldn't it be falling well below the others upon discharge? Seems to be falling right in line with the others right now. I'll put in a 10AH run tomorrow or whenever it's sunny again and we'll see what happens. I was expecting to see like 2.6v or 2.5v upon discharge but it didn't do that.
on the next run, what should I look for? I almost broke my multi-meter today so I don't think I'm doing what I did today. It actually fell and smashed into two pieces but it still works somehow. It sounds to me like no one really knows for sure if my cell is damaged or not. If it is damaged, then I shouldn't be able to get 10AH out of it, right? So, that'll be the first test.
Russell
1 MW
As I said before if the suspect cell registers a voltage close to the others when discharged then it's fine. Go on progressively longer rides and check the cell voltages after each until you are confident everything is copasetic then go back to your normal routine.
The bad part about becoming intimate with your battery is you tend to fret about the little things. I suspect there are plenty of folks riding around with BMS equipped battery packs with one or more weak cells but since they never check the voltages they live in ignorant bliss. They never get the full rated capacity from their batteries however if they aren't monitoring the pack with a wattmeter then once again they are completely unaware.
-R
The bad part about becoming intimate with your battery is you tend to fret about the little things. I suspect there are plenty of folks riding around with BMS equipped battery packs with one or more weak cells but since they never check the voltages they live in ignorant bliss. They never get the full rated capacity from their batteries however if they aren't monitoring the pack with a wattmeter then once again they are completely unaware.
-R
AndyH
10 kW
morph999 said:
Do you see the cells that sag the lowest in my chart? Those were the ones that have been below 1.4V before I installed cell-level LVC. You can also see that the cells come back to similar voltages once the load is removed. So no - I don't think that you'll get any useful information by measuring the static voltage after a short ride.
Ever see the car battery test units garages and some parts stores have? They diagnose a battery not by measuring the static voltge but by watching performance under a load...
Now that you've read these last two 'paragraphs' - please go back and look at the chart and see that the cells that sag the deepest do NOT have the lowest static voltage once the load is removed. Please take a minute or three and let that sink in. You cannot tell the condition of the cell by measuring the static voltage.
Andy
Why would voltage sag kill your cells? There are two ways of thinking about lifepo4: one says that voltage sag can kill your cells no matter how many AH you've used from it and the other says voltage sag doesn't matter and it's how many AH you've used. Don't the experiments show that voltage sag doesn't matter? All the charts that I've seen show that a 3C discharge will kill a thundersky cell much quicker than 2C discharge but you still have to actually use almost 70 - 80 % with a 3C load to actually kill it. It's not going to die after only a few minutes on a 3C load.
Andy, your own charts show that voltage sag don't kill cells, it's the amount of AH used. Look at the cells rebounding after voltage sag. You are essentially saying that voltage sag kills cells so if that were true, we would see the cells that fall below 2.5v just basically be destroyed but that's not what it shows. It shows the cells rebounding.
Andy, your own charts show that voltage sag don't kill cells, it's the amount of AH used. Look at the cells rebounding after voltage sag. You are essentially saying that voltage sag kills cells so if that were true, we would see the cells that fall below 2.5v just basically be destroyed but that's not what it shows. It shows the cells rebounding.
icecube57
10 MW
Dude none of my TS cells hold their surface charge long and Ive got almost 20 partial cycles 2 full cycles and 500+ miles. I have 20 cells. I balance them to 3.7 v.They all read 3.7v across the board on my Hobby King battey Medics. If i take them off the charger the voltage falls on all cells. Some more than others. The all are between 3.4-3.6v the do stop self discharging at some point. All at different voltages. I wouldnt worry about it. I take way better care of my cells than you do and they dont hold their surface charge. Its the nature of the beast, Even ping packs dont hold their surface charge. Dont blow a wad over this man.
If i slap the surface charge off they all rest at the same cell level + or - 0.01v. They stay that way to a 90-95% discharge then they are all over the place.
If i slap the surface charge off they all rest at the same cell level + or - 0.01v. They stay that way to a 90-95% discharge then they are all over the place.
icecube57
10 MW
morph999 said:
icecube57
10 MW
AussieJester said:
On my setup I have my lvc set at about 58-59v. The cells will sag lower than that to around 56-57v under 50A+ load. When I hit the lvc you feel that quick pulsation of the controller hitting the lvc and killing the power multiple times a second. By me doing this it keeps the voltage high and doesnt allow my pack to sag as low as it could without the restriction... but it still allows me to maintain full throttle with plenty of power in most situations. Almost as if its limiting the current but the rapid oscillation of the controller hitting the lvc is sorta acting like a second form of pwm on the controller.
AussieJester
1 TW
Voltage sag? whats that? If you bought my 'special batteries' morph you wouldn't have voltage sag either ...might
have a big mogfo lipo fire but no voltage sag hehe
KiM
have a big mogfo lipo fire but no voltage sag hehe
KiM
Thanks Ice Cube. I'm not going to worry anymore about it. It's just odd checking cell to cell and they were all 3.59 and then I get to one and it's like 3.38v and I'm like "oh crap !!". I bought these cells to have fun and ride as much as I want so that's what I'm going to continue to do within reason. It's worth $360 to me anyway. Even if they only last a year, it's still cheaper than buying gasoline or almost the same price.
AndyH
10 kW
morph999 said:
NOBODY SAID THAT SAG WOULD KILL CELLS - I sure as hell didn't!*** I made that caps because I think it's important. Here's another one: LITHIUM CELLS CAN BE DAMAGED BY DISCHAGING THEM TOO FAR - as in "I didn't know the cell wasn't fully charged when I started a ride and and one cell fell off the cliff and discharged below 1.6V." My low cells under load have been stressed by over-discharge during my early testing without a BMS or cell-level LVC. (That's why I recommend cell level LVC - not because I'm a fan-boy of someone that says BMSs are good - but because I learned from my testing that the engineers at PSI knew what they were talking about. :wink
THUNDER SKY CELLS WILL NOT HOLD A SURFACE CHARGE right off the charger like A123 cells will. My bench testing of 40Ah, 60Ah, and 100Ah TS cells shows that a NEW cell will bleed down to about 3.65V. My 60Ah cells with 13 months and 1300 miles on them bleed down to the 3.32-3.52 - But keep in mind that the bottom 12 cells are only pulled down to the 3.3V range because they're the ones powering the PakTrakr voltage sensors. They DO NOT bleed down that far unless they have a load on them - or unless they're damaged.
IF you have damaged cells you probably will NOT be able to tell which one by measuring static voltage until it complely dies and is sitting at below .5V. That is one failure mode but it's not the only one.
My pack has KNOWN STRESSED CELLS - and looking at how they perform under a load will help you diagnose your pack. But static voltage isn't useful unless one is looking at end of charge, or off the cliff into deep discharge.
Here's a look (2nd graph in the post) of three cell conditions. One cell is falling off the cliff - notice the deep sag and the much slower recovery? There's a decent indicator of cell at a lower state of charge. Two other cells (our weak cells!) are sagging to a lower voltage than the rest of the pack - but voltage recovery is 'normal'. Then there's the rest of the pack just doing waht they do.
So - you can tell a DEAD cell by looking at the static voltage and noticing that it's somewhere below .4V. You might be able to spot the damaged but not yet dead cells if they stay around .4-1.5V. You might be able to spot the cells that are somewhere below 20% SOC by tracking static voltage. But cells with higher internal resistance (stressed) won't make an appearance on the charts until they're working. Static voltage checks isn't very useful.
Sorry, man - too many words. I really want to help you understand but I'm not sure I'm able to get the message across.
[edit] *** But too deep a sag can be an indication of too high a rate of discharge - and that can stress the cells. That's why people here are suggesting you turn the load down a bit to keep the cell load below 2C... [/edit]
liveforphysics
100 TW
It's like talking to a wall Andy. He knows everything all ready. Jack and Morph are the battery experts on this board, and anybody trying to talk some sense on the subject is just trying to make money selling useless BMS's that do more harm than good.
Isn't that right Morph?
Isn't that right Morph?
AndyH
10 kW
liveforphysics said:
Unfortunately, Luke, I suspect you're right. But I'm not skilled enough to recognize that the messge isn't making it across, and isn't likely to even if I talk louder or try to find another analogy. I guess that's why I'm more confortable with equipment than most people.
In all fairness, I got to stress some cells and Morph should have that same right. I'll leave it where it lies and hope that the eventual lessons aren't that expensive.
Andy
Russell
1 MW
AndyH said:
Andy,
First I will agree that it would be best to test a battery/cell under load but most of us do not have sophisticated test equipment to do that. Second, what I have been suggesting is not checking the static cell voltage when charged, rather checking it when the battery has undergone various states of discharge. Because LiFePO4 cells have such a high surface charge checking the cell voltages when the cells are charged doesn't tell you much. With my LiFePO4 battery the cells are over 99% full when their static voltage is 3.40V+ and if one cell reads 3.55V and another reads 3.42V it's meaningless.
There is however merit I believe in checking the static cell voltages when the battery is in a discharged state. At first blush it may appear that little info can be gained from looking at the static voltage of a LiFePO4 pack due to its flat discharge curve however once you collect enough data on a pack the relationship between the resting pack voltage and the state of charge is clearly visible. You do have to remember though that the differences are quite small until the cells go over the edge. For example the range for my 48V/10Ah pack was a high of 52.82V (3.301V/cell) after a 2.747Ah discharge to 51.45V (3.216V/cell) after a 9.291Ah discharge.
After I removed the BMS from my battery pack I checked the individual cell voltages after the first 6 rides to see if there were any weak cells. After the first ride using 57% of the pack capacity 10 cells were resting at 3.29V and 5 were at 3.30V (yes the battery now has 15 cell groups). On the 4th ride after using 77% of rated capacity I noticed one cell which was weaker than the rest, it measured 3.24V while 7 cells were at 3.26V and 7 were at 3.27V. On the next ride using 87% of capacity the spread was slightly wider with the one laggard at 3.20V with 4 cells at 3.23V, 8 at 3.24v and and 2 at 3.25V. If I were to ever discharge the pack to a greater extent I would expect lower resting cell voltages.
Now granted the static voltage tells me nothing about the individual cell voltages under load however I believe as long as the cells are staying in sync at rest after drained that they are well matched in capacity. The lowest pack voltage I've ever recorded from my WU meter while riding corresponded to 2.681V/cell at the end of the ride where I used 93% of the rated capacity, a DOD I've never exceeded. The BMS was connected for this ride and it never tripped though that's not surprising given the 2.0V individual cell trip point (per BMS spec sheet). The lowest pack voltage during a ride recorded since removing the BMS corresponds to 2.817V/cell therefore I'm pretty confident no cell is going too low.
My own feeling is the absolute #1 prerequisite for a good battery is to have each cell as closely matched to others as possible. If this is the case then the cells will discharge and charge together and there will be no chance of damage to individual cells as long as the pack isn't fully drained. Where a BMS is handy is if you don't have well matched cells since the BMS will trip when the lowest capacity cell hits the lower limit. However if the low cell has significantly less capacity than the rest of the cells it would be better to replace it to restore optimal battey capacity.
-R
Well, I don't know if I'm toasting my cells or not but it's definitely a possiblity. I measured the sag today and it was all the way down 34.5v . That's about 2.8v per cell and I had only used 3 AH.
I don't have the money to toy around with this and find out what is what so I'm probably going to buy a cycle analyst and limit my current to about 25 amps instead of the 43 max amps that it's using now. I'm kind of pissed that a 35 amp controller is actually 43 amps. I'd probably be okay if it really was a 35 amp controller. On a 20amp discharge, it only sags to 38v. I just hope I haven't already done a lot of damage to my batteries.
If someone with money would like to do an experiment, like take a cell that's rated at 2C and deliberately use a 4C discharge and take it down to 1.5V under discharge and then see if it's damaged. Then we can find out for sure whether voltage sag kills cells or not. I'm talking about 4C discharge that only uses about 2 or 3AH because we need to find out whether it's voltage sag and not the amount of AH used. And if it's not damaged, do it maybe 20 or 30 times to see if damages after repeatedly doing this. According to Jack Rickard, this should not damage a cell. I have to admit after thinking about it, It probably does damage these cells to do that. The question is: are people damaging their packs by using too much of it at too high of a C rating or are they damaging them from just too high C discharge alone.
According to Rickard, he says something like this:
A 36v20AH pack rated at 2C...can be used at 3C but it the knee of the discharge curve will be much shorter like at about 15AH.
A 36v20AH pack rated at 1C can be used at 3C but it'd have a knee at probably 50 % of the capacity listed so the knee would be at about 10AH.
That's what he says. I have no idea if it's true or not.
Andy, I can't read your charts so I have no idea what you are trying to show. I just don't want to ruin my investment.
I don't have the money to toy around with this and find out what is what so I'm probably going to buy a cycle analyst and limit my current to about 25 amps instead of the 43 max amps that it's using now. I'm kind of pissed that a 35 amp controller is actually 43 amps. I'd probably be okay if it really was a 35 amp controller. On a 20amp discharge, it only sags to 38v. I just hope I haven't already done a lot of damage to my batteries.
If someone with money would like to do an experiment, like take a cell that's rated at 2C and deliberately use a 4C discharge and take it down to 1.5V under discharge and then see if it's damaged. Then we can find out for sure whether voltage sag kills cells or not. I'm talking about 4C discharge that only uses about 2 or 3AH because we need to find out whether it's voltage sag and not the amount of AH used. And if it's not damaged, do it maybe 20 or 30 times to see if damages after repeatedly doing this. According to Jack Rickard, this should not damage a cell. I have to admit after thinking about it, It probably does damage these cells to do that. The question is: are people damaging their packs by using too much of it at too high of a C rating or are they damaging them from just too high C discharge alone.
According to Rickard, he says something like this:
A 36v20AH pack rated at 2C...can be used at 3C but it the knee of the discharge curve will be much shorter like at about 15AH.
A 36v20AH pack rated at 1C can be used at 3C but it'd have a knee at probably 50 % of the capacity listed so the knee would be at about 10AH.
That's what he says. I have no idea if it's true or not.
Andy, I can't read your charts so I have no idea what you are trying to show. I just don't want to ruin my investment.
icecube57
10 MW
Im going to say again these cells sag. As someone mentioned before they sag alot especially in bigger EV applications. A 55A stall load on my motor gives me 57v. This was with a pack fresh off the charger at 74v. My resting pack voltage after that was 66.9v. Under load that equates to 2.85v pers cell. All my cells perform the same under load + or -.02v. I left my Battery Medics on the packs under load to view their voltage. Ive seen my pack voltage sag down into the low 50s fresh off the charger. These arent high current cells. But they can take short bursts you just have to deal with the sag that comes with it.. I have my controller setup to overshoot the current for 10 seconds before throttling back down to the default controller value. My controller peaks at over 70A and its a 45A. So i wouldnt get pissy about that. A CA isnt going to help much. By the time you controller peaks.. the CA will take a second or so try throttle back to limit the amps so it will be too late for the CA to correct anything. A peak is a peak. It lasts for a short second not continuous. These cells can peak up to 10C for 5-10 seconds. You dont want to know that the voltage sag is at that level. It furkin retarded trust me. Ive cranked a car through on of my 5s my pack. Its 18.5 off the charger 16.6 resting and while the truck was cranking the voltage was in the high 11s.
I don't think the cycle analyst will do what you want morph. When I used mine to limit current, all it did was cut throttle when the current exceeded the set level, then turned the throttle back on when it was below, so it was like a stutter, and I stopped using it.
icecube57
10 MW
nomad85 said:
It does that when you use the e-brake instead of the throttle line
to control amps.
