You are all destroying your LiFePO4 cells!

bgeery · Nov 7, 2011

OK, so I'm exaggerating a little bit.

But you are going to get fewer life cycles than specified by the manufacture if used under that standard conditions.

A fully 100% charged battery has a resting voltage of 3.33 volts per cell at 25 Celsius.

The only way to insure we never overcharge the battery is to charge at 3.33 volts until it tapers down to 0.00 amps. That's a fully 100% fully charged cell. For those that want to charge to less than 100%, just discontinue that charge before current drops to 0.00 amps.

The problem with this is that it takes longer to charge this way; so the manufacture gives us a formula of at 25 Celsius, apply .5C, until the cell reaches 3.6 volts, then disconnect the charge. This is 100% charged.

I see allot of people charging at 3.5-3.8 volts per cell until 0 amps. This is overcharging the cells to some degree. Yes, they may accept some additional charge above 100%, but it's past the manufacture's definition of the 100% charged point.

Either charge at 3.33 volts until it tapers down to 0.00 amps, or go with the manufacture's formula of .5C charge rate to 3.6 volts and disconnect. Either one is the only known safe and accurate charging procedures to reach 100%.

How's that for a first post?

PS: Other charging formulas could be developed by the manufacture, if they choose to do so. We could also, if we knew the exact conditions they use to determine and define 100% state of charge.

frodus · Nov 7, 2011

..... except you won't cause any current to flow if the battery is at 3.33V and the charger/power supply is at 3.33V. There has to be a voltage difference between the two.... period.

The larger that differential between supply and load (battery) the larger the amps you'll draw from the supply. Thats how you cause current to flow. Most chargers we use are CC/CV and they current limit to XXAmps, and voltage limit to YYVolts.

As long as you don't A) go over the manufacturer max charge current B) Manufacturer max temperature or C) manufacturer max voltage, you're going to be within the limits of the cell. That's how the manufacturer actually rates cell life-cycle, using those specs.

We're not overcharging them at all. It's actually the only way you can charge a battery. You have to pull the battery voltage above it's nominal voltage and cause current to flow into it.

bgeery · Nov 7, 2011

frodus said:
..... except you won't cause any current to flow if the battery is at 3.33V and the charger/power supply is at 3.33V. There has to be a voltage difference between the two.... period.

The larger that differential between supply and load (battery) the larger the amps you'll draw from the supply. Thats how you cause current to flow. Most chargers we use are CC/CV and they current limit to XXAmps, and voltage limit to YYVolts.

Agreed. But if the cell is at 3.33 volts resting voltage, it's already at 100% state of charge, and any further charging is an overcharge. You don't want further current flow.

A battery under 100% SoC will rest at something less than 3.33 volts until it's fully charged. The charge will be slower and slower as it approaches 3.33 volts, but the current will flow, until the cell reaches the true 3.33 resting voltage equilibrium with the charger and is at 100% SoC, and current flow will cease.

Using higher than resting voltage on the cell does nothing but increase the rate you can charge; avoiding the long trickle charge that would occur as the cell approached 3.3 volts and less and less current flowed. That's a valid practical compromise, but it doesn't get a 100% charged cell. It gets a 100.2% or 99.6% charged cell.

frodus said:
As long as you don't A) go over the manufacturer max charge current B) Manufacturer max temperature or C) manufacturer max voltage, you're going to be within the limits of the cell. That's how the manufacturer actually rates cell life-cycle, using those specs.

We're not overcharging them at all. It's actually the only way you can charge a battery. You have to pull the battery voltage above it's nominal voltage and cause current to flow into it.

The manufacture gives us exatly *one* charge curve that they rate the battery at, and if you are not charging in exactly the same conditions, you can't tell your state of charge with true accuracy. You can come close enough for practical use, but you still don't really know exactly.

For example, how many people are charging at 3.5-3.8 volts until the charge tapers to 0.0 amps? I see top balance supporters mentioning this all the time here. Where does that extra power go between the 3.8 volts and the 100% charged resting voltage of 3.33 volts? It's not a whole lot of energy, but it is an overcharge of the battery. My guess is that it's oxidizing some of the plates and diminishing some minor amount of cell capacity.

DavidDymaxion · Nov 7, 2011

Great first post. I throw this out as a discussion point, and not as hard science. Jack mentioned this in his videos. (All hail Jackton! All hail Jackton!)

The voltage for a cell is Vcell = V0 + I*R

So suppose you have a cell with 5 milliOhms of resistance, you want to charge at 10 Amps.

Vcell = 3.33V + 10 Amps * 0.005 Ohms = 3.38V

Suppose you are regenning at 100A (or have a super powerful charger):

Vcell = 3.33V + 100 Amps * 0.005 Ohms = 3.83V

Anyway, it's a thought that if the cells are cool enough and the currents are low enough, perhaps this isn't hurting them at all?

bgeery said:
OK, so I'm exaggerating a little bit. But you are going to get fewer life cycles than specified by the manufacture if used under that standard conditions.

A fully 100% charged battery has a resting voltage of 3.33 volts per cell at 25 Celsius.

The only way to insure we never overcharge the battery is to charge at 3.33 volts until it tapers down to 0.00 amps. That's a fully 100% fully charged cell. For those that want to charge to less than 100%, just discontinue that charge before current drops to 0.00 amps.

The problem with this is that it takes longer to charge this way; so the manufacture gives us a formula of at 25 Celsius, apply .5C, until the cell reaches 3.6 volts, then disconnect the charge. This is 100% charged.

I see allot of people charging at 3.5-3.8 volts per cell until 0 amps. This is overcharging the cells to some degree. Yes, they may accept some additional charge above 100%, but it's past the manufacture's definition of the 100% charged point.

Either charge at 3.33 volts until it tapers down to 0.00 amps, or go with the manufacture's formula of .5C charge rate to 3.6 volts and disconnect. Either one is the only known safe and accurate charging procedures to reach 100%.

How's that for a first post?

PS: Other charging formulas could be developed by the manufacture, if they choose to do so. We could also, if we knew the exact conditions they use to determine and define 100% state of charge.

bgeery · Nov 8, 2011

DavidDymaxion said:
Great first post. I throw this out as a discussion point, and not as hard science. Jack mentioned this in his videos. (All hail Jackton! All hail Jackton!)

The voltage for a cell is Vcell = V0 + I*R

So suppose you have a cell with 5 milliOhms of resistance, you want to charge at 10 Amps.

Vcell = 3.33V + 10 Amps * 0.005 Ohms = 3.38V

Suppose you are regenning at 100A (or have a super powerful charger):

Vcell = 3.33V + 100 Amps * 0.005 Ohms = 3.83V

Anyway, it's a thought that if the cells are cool enough and the currents are low enough, perhaps this isn't hurting them at all?[/QUOTE]

Very interesting. So that might be the "formula" for coming up with a recipe for charging in excess of 3.3V and insuring exactly 100% SoC is reached but not exceeded.
In your example then, if I wanted to charge at 100 amps, I'd apply 100 amps, and charge would be complete to moment the cell reached 3.83 volts. Cease charging and you have a 100% SoC cell. Of course, in a string of cells that all have different resistances, I'm not sure how that would interact.

My main issue is with those holding cells at 3.5-3.8 volts and forcing more and more current into the cell, until reaching 0.0 charging current.

Anyway, it's a thought that if the cells are cool enough and the currents are low enough, perhaps this isn't hurting them at all?

In theory, if the current is flowing into the cell after 100% SoC, it has to be going somewhere. In a lead acid battery, the forced energy electrolyze the water in the electrolyte creating hydrogen and oxygen, and I believe also can warp the plates. Not a big deal to add water. Maybe in LiFePo4 it's just converted to heat, but I don't think so. Jack would know.

piotrsko · Nov 8, 2011

my $.02:
It is a CHEMICAL reaction. the current is what is forcing the chemicals to physically revert / change to some sort of different state. When the charge current goes to zero, then theoretically all the chemicals have been changed, all the electron holes in the collector plates are empty again. Not exactly true, but close enough for this argument. As you overcharge you may even apply a slight coating made from the collector plates to the materials in the electrolytic thus altering the characteristics, possibly forever. In one of the battery books I have read, as little as 1 percent chemistry change destroys the cell. heat does this process also.
ymmv

EVfun · Nov 8, 2011

For example, how many people are charging at 3.5-3.8 volts until the charge tapers to 0.0 amps? I see top balance supporters mentioning this all the time here.

I have heard of this being used by a few people to top balance a pack in parallel but I have never heard of this being a regular charging routine. It is a drastic way of sharply finding a full charge, similar to those who bottom balance by getting the resting voltage down to 2.5 to 2.8 volts to find a matched empty.

I did my top balance by charging to 3.65 volts until the current tapered down to about 0.5 amps (60 amp hour cells.) It was mostly the time that it took my to get them all the same. That might cause some wear, but it is a single cycle. My charging routine is 12 amps until the voltage reaches 3.50 volts, then hold 3.5 volts for 40 minutes (around 2.5 amps ending current.) The next morning the cells consistently read 3.34 volts each and after 24 hours they consistently read 3.33 volts each.

Here is the Thunder Sky charging documentation, from "Thunder Sky LiFeYPO4 Power Battery Performance Test Instructions":

5.2.4 Charging At 20°C±5°C temperature, the cell is discharged at a current of C3 till voltage of the cell reach 2.8V, and then start to perform constant current charge at a current of C3 till voltage of the cell reach 4.0V.

5.2.4.1 Low temperature charging At -18°C±5°C temperature, the cell is discharged at a current of C3 till voltage of the cell reach 2.2V, and then start to perform constant current charge at a current of C3 till voltage of the cell reach 4.2V.

dougingraham · Nov 8, 2011

bgeery said:
A fully 100% charged battery has a resting voltage of 3.33 volts per cell at 25 Celsius.

The only way to insure we never overcharge the battery is to charge at 3.33 volts until it tapers down to 0.00 amps. That's a fully 100% fully charged cell. For those that want to charge to less than 100%, just discontinue that charge before current drops to 0.00 amps.

I see allot of people charging at 3.5-3.8 volts per cell until 0 amps. This is overcharging the cells to some degree. Yes, they may accept some additional charge above 100%, but it's past the manufacture's definition of the 100% charged point.

Either charge at 3.33 volts until it tapers down to 0.00 amps, or go with the manufacture's formula of .5C charge rate to 3.6 volts and disconnect. Either one is the only known safe and accurate charging procedures to reach 100%.

Where did you get the 3.33 volt number? I have not seen that on any of the manufacturers data sheets. The only reference I have found to a float voltage is 3.45V found on the A123 systems data sheets that comes with the 26650 cells. And since those are plain Jane LiFePo4 it seems like a very reasonable number that could be used on other brands as well. I have floated the A123 cells at 3.45V for several days at that voltage with no ill effects.

bgeery · Nov 8, 2011

@ frodus:
I find it hard to write a reply without nested quotes. sigh.

As soon as you start charging it, the voltage will rapidly increase, and within a very short time it will match the voltage of the charger and the current will drop to zero. You're not overcharging it. This happens almost immediately.

Yes, I said this is a slower method of charging, but is guaranteed to find and stop at exactly 100% SoC. The cell voltage only rises in response to the charge current. Taper the current and the cell voltage will drop again and current flows, this happens until the cell is actually at 100SoC and does not fall.

The manufacturer specifies that the charge voltage is to be no more than 3.6-4.0Volts (depending on your cell). As long as you're at, or below that, you'll never overcharge the cell.

The cell is 100% charged when current stops flowing, but the cell has to be pulled higher than it's nominal voltage, even if it's just a tad. Higher voltages just cause it to charge faster, but if you're only at nominal voltage, you're not going to charge.

Yes, but they don't say hold 4.0 volts all the way to 0.0 amps. And a discharged cell's *static* voltage is not 3.3 volts, thus will accept current regulated at 3.3 volts, at a slower and slower rate, until the cells' *static* voltage reaches 3.3 volts. This is 100 SoC.

Lets make it clear. I'm not saying charging this way is practical for everyday practice, as it may or may not take too long to reach 100% SoC. But, I am saying it's the only way to know you are actually *at* a perfect 100% SoC.

Have you even looked at a charge curve from a manufacturer? It shows voltage and all the one's I've seen are above way 3.3V.
http://liionbms.com/pdf/thundersky/TS-LFP100.pdf
http://liionbms.com/pdf/zhejiangheadway/HW38120LS.pdf
http://liionbms.com/pdf/psi/PC40138F1.pdf
http://liionbms.com/pdf/shandong/200ah.pdfhttp://liionbms.com/pdf/goldpeak/GP18EVLF.pdf
http://liionbms.com/pdf/huanyu/HYP-3.2V-100Ah.pdf

That charge curve is at a given C-rate. The lifecycle is at that c-rate. As long as the cell never goes above the manufacturer's recommended charge voltage, you're going to charge to 100%. Most of the charge curves I've seen are at 1C. True, you could charge at 3.5V, which many do, but 3.5V > 3.33V, so charge flows.

no, you charge a >3.33 so the charge flows faster. But in charging in excess of 3.33 volts, you are forcing the electrons into the cell, instead of alloing them to flow in naturally. That means it's up to you to discontinue the charge a some time befor you have tapered down to 0.0 amps.

Looking at your very first link for the LFP 100 clearly shows that the cell reaches 100% SoC before you taper the current down to zero. It looks to me the current should taper to 4 amps or so and stop. The battery is 100% charged. Continuing to 0.0 amps at *any* voltage in excess of 3.3 is an overcharge of the cell. How much does this damage the cells each cycle? That's a good question that I don't have an answer for.

Thats how I charge, every time. My charger is set to a certain max voltage, and a certain max current. Set it and forget it. It turns off when current goes to 0.

If that certain voltage is >3.3 volts, then you are slightly overcharging the cell each cycle.

It's not power we're looking at... it's energy. What you aren't looking at is that Energy = Wh = the area under the curve.

Lets say you have a 10Ah 3.2V nominal cell. Situation 1: Charge voltage is set to 3.5V. Charge current is 1C, or 10A. Situation 2: Charge voltage is set to 3.7V. Charge current is the same.

The slope of the charge curve is steeper with situation 2, so it gets to 3.7V faster. Then it holds that voltage until current drops to 0. The area under the curve will be XXWh.

The slow of the charge curve is less with situation 1. Not only does it take longer to rise to the 3.5V charge level, it takes longer for the current to drop to zero.... so it takes longer to charge. The area under the curves is essentially the same (some is blown off due to IR of the cell).

I get what you are saying, and you have a point about the resistance of the cell requiring something greater than 3.33 to allow a charge. So lets call it 3.331 volts then-- or whatever small amount is need to overcome the couple of milliohms of resistance in these cells.

I get the point that charging at higher voltages get the charge dome quicker. The point is, these fast charges (and that's what they are as far as the cell is concerned) need to terminate at some point before the current tapers to 0.0 amps. The higher above 3.3 volts you are charging at, the higher point in the current curve you need to terminate the charge. And in any case, with these fast charges, you never will reach 100% charge, and will always overshoot or undershoot by some percentage. The better you picked your termination current point, the closer you will be to 100%, but it will never be exactly right.

One alternative might be to fast charge to say 99% and then slow charge at 3.331 volts to 0.0 Amps. That will get you to 100% SoC much quicker, and insure you don't overshoot and overcharge.

GizmoEV · Nov 10, 2011

bgeery said:
Yes, but they don't say hold 4.0 volts all the way to 0.0 amps. And a discharged cell's *static* voltage is not 3.3 volts, thus will accept current regulated at 3.3 volts, at a slower and slower rate, until the cells' *static* voltage reaches 3.3 volts. This is 100 SoC.

Lets make it clear. I'm not saying charging this way is practical for everyday practice, as it may or may not take too long to reach 100% SoC. But, I am saying it's the only way to know you are actually *at* a perfect 100% SoC.

When I got in my car to leave work today the pack was sitting at 66.6V. I have 20 cells so that is 3.33vpc. According to your statement above my pack was at 100%SOC. So, did I charge at work or not?

If you said I charged at work you would be wrong. My 200Ah pack was sitting at 79%SOC. It hasn't been charged for 4 days and just over 42Ah had been extracted from the pack. This is not just a one time occurrence. It happens every discharge cycle. 3.33V is not 100%SOC. Therefore, it is not a way to determine 100%SOC.

tomofreno · Nov 8, 2011

The voltage for a cell is Vcell = V0 + I*R

So suppose you have a cell with 5 milliOhms of resistance, you want to charge at 10 Amps.

Vcell = 3.33V + 10 Amps * 0.005 Ohms = 3.38V...

I remember Jack talking about this. Seems to me he also demonstrated charging a cell a 100A to higher voltage than the charging spec claiming no ill effects based on this reasoning. Makes sense. Doesn't mean it is correct though.

Some with shunt balancers might hold several cells at 3.5V at 0.5A or so charge current for a half hour to balance. According to the above formula those higher voltage cells should only be at 3.3325V. They are not because they are on the exponential part of the curve where cell resistance is higher. I've often wondered if that does any harm to them. If the voltage drop across the electrolyte is high enough it will break it down, resulting in plating on the electrodes. According to the CMU prof that requires about 4.3 to 4.4V terminal voltage. I think the voltage starts rising exponentially as the intercalation sites start becoming scarce and it gets more difficult for ions to work their way in, requiring larger electric field, E, and force, qE, so higher voltage, to pull them in. But that is just a guess. If correct, I wouldn't think any harm would be done as long as the voltage is not high enough to start breaking down the electrolyte. Though it may stress the electrode material more, which could lead to somewhat lower cycle life.

frodus · Nov 8, 2011

I think I might have misunderstood you when you were saying 3.33V. I thought you were saying that was nominal. My mistake. I have to read in spurts at work sometimes.....

Most people aren't damaging them with voltage.... people damage them by putting a charger on them with too high of a current limit on it. The lower the current for discharge and charge the better you're gonna be. Lowering the charge voltage does that by nature.

So lets say your cell's nominal voltage is 3.2V for instance, setting the charger to it's 100% SOC voltage of 3.33V as you said.... it would take days to charge a cell even from 50% DOD. Really really slow, but safe for the cell.

Now same cell, set to 3.4V, wait for the current to fall below an amp, as shown in some of the graphs. It's just fine. As long as you're not putting too much current into the cell, getting the cell too hot or getting the voltage too high, you're fine.

bgeery · Nov 8, 2011

I think we are mainly in agreement here. Any time we charge above the natural final static voltage of the cell (plus a variable amount to account for IR) we should remove the current before it reaches 0.0 amps. The further above 3.33 volts, the sooner we want to remove the current.

We also agree, charging at 3.3 volts would probably take too long in practical terms, so it's more an academic argument. Interesting none the less.

EVfun · Nov 8, 2011

Looking at the specs you provided some of the manufacturers are calling out quite high termination voltage levels, but finishing current level was less clear.

Several seemed to agree on the idea of 3.65 volts with the ending current of 0.02 to 0.05C (ending amp level 2% to 5% of the amp hour capacity.) I think most of the non-BMS Lithium users are stopping well before that point. I'm stopping at about 3.5 volts and 0.04C. I have wondered how long it would take to charge if I held the cells to around 3.4 volts.

bgeery · Nov 8, 2011

Do you have an user adjustable charging source? I'd be interested to know myself. I know that the lower your finishing voltage, the lower you can adjust your final cutoff current. So maybe 3.4 volts to 0.02 amps?

ElectriCar · Nov 10, 2011

I stop my charge around 3.45V/cell. Using the CC/CV method, my charger ramps up to about 30A initially and steadily increases voltage in order to maintain 30A as the batteries SOC increases until the voltage reaches approximately 3.45VPC. At that point it maintains the voltage level allowing the current to taper until it reaches 0.0A.

I have 200AH cells so I'm well under the .5C charge current. This method as frodus said is very common, efficient, doesn't over current anything and charges the pack quickly so you're in business to drive again quickly. In a pack with AH as high as mine, I fail to see how charging to 3.33VPC would be of any benefit to me whatsoever. It would take forever to charge as well. With a 5kw charger it already takes about 6-7 hours if I'm nearly empty to recharge.

I just checked mine which is charging now and with 25Ah to go or about 7/8 charged, I'm at 3.418VPC. If I lowered it to 3.33 or so I'd be forever charging with a pack my size.

I see NO BENEFIT to charge to 3.5VPC or higher. There just isn't any capacity up there plus you're risking a cell spiking and going into over voltage if you're not running a BMS which I don't. So if there's no capacity up there why bother to go there and in the case of a BMS, waste energy draining it off the high cells?

After draining some of the charge ALL MY CELLS are at the same voltage to within 1/1000V. It's only when you get close to completely charged that they start to diverge in voltage, and that's where you're getting onto thin ice.

Just checked it again FWIW, ALL 50 CELLS charging at full throttle are reading 3.430V, every single one with 9.5Ah to finish. Soon they will start to diverge. I'm going to try and find that voltage. I may adjust my Max V downward after a few more voltage checks, depending on at what VPC they start to diverge and how far from the finish line the pack is.

Edit - Checked again: 3.466VPC (average) Cell voltages noticeably diverged. 0.5Ah remaining and charger is cutting back quickly, now @7.5A

ev4me · Nov 10, 2011

ElectriCar said:
I see NO BENEFIT to charge to 3.5VPC or higher. There just isn't any capacity up there plus you're risking a cell spiking and going into over voltage if you're not running a BMS which I don't. So if there's no capacity up there why bother to go there and in the case of a BMS, waste energy draining it off the high cells?

The idea is that you charge at a higher voltage, but stop the charge somewhere before reaching zero current.

For example, imagine these two charging configurations:
-Constant current at 30 A until voltage reaches 3.33 V, then constant voltage until current reaches 0.0 A
-Constant current at 30 A until voltage reaches 3.65 V, then constant voltage until current reaches 0.5 C

Both situations should push approximately the same amount of charge (amp-hours) through the battery, but the second configuration will do it faster. After removal from the charger the cell voltage will be around 3.33 V in either case (not 3.65 V as you might think in the second case, since some current was still flowing at that voltage).

The disadvantage here is that the cell voltage is held at 3.65 V for a few hours during the charge. I have not seen any sort of consensus on how this affects the battery or how high one can safely go. If temporary high voltages were okay, one could theoretically charge at constant current all the way to 100% SOC. Unfortunately, anything above 4.3 V or so decidedly destroys the cell. But what sort of effect does a voltage between 3.33 V and 4.0 V have? We do 3.65 V, but I've seen others who do anything from 3.45 V to 4.0 V.

In conclusion, there is a difference between "charging to 3.65 V" and "charging at 3.65 V," so long as current is not zero.

EVfun · Nov 10, 2011

I decided to pour over some of the data in the garage. I'm adding 7.2 amp hours (20% of the total capacity) between hitting 3.44 vpc (110 pack volts) and the end of charge at 3.50 volts. That is 12 amps for 14 minutes, from 110 volts to 112 volts, and then tapering down to 3.6 amps over the next 40 minutes. This is a pretty good end of charge for my Thunder Sky pack because 24 hours later the cells are resting at 3.33 volts.

If I held 110 volts (3.44 vpc) I could get to the same point. I just have no good idea how long it would take.

Guest · Nov 10, 2011

The voltage was active discharge voltage and that is what you want to see. The resting voltage after discharging can be deceiving. It will rise back up after sitting but the cell is still empty. It has voltage but not capacity. The capacity has been used up for that charge/discharge cycle. If you let your cell sit for hours then go out and begin discharging again it will drop to that old discharge voltage very very fast. I have tested that. When its empty its empty and its best seen during active discharging.

Pete

ElectriCar · Nov 10, 2011

David didn't you mean 0.001V? That is 1/1000 of a volt.

That's what my Fluke meter measures down to.

GizmoEV · Nov 10, 2011

I meant ten-thousandths. I went back and fixed it. He used an Agilent(sp?) bench top meter. Unless I'm confusing that test with the one where he was comparing 2.5 year old cell with a new one.

spdas · Nov 10, 2011

My sweet spot for charging is at 3.477V per cell. After I top balance, it allows for a "rogue" cell to go up to 3.55 and still be safe and a lagging cell to be as low as 3.40 which fully charged and into the top anyway where there is not much amps stored. When I charged at 3.49V each cell, I saw considerably more variance in the "no amperage" zone at the top. I will be lowering my charging voltage to 3.455, but it will take longer to charge at this rate.

Francis

ElectriCar · Nov 10, 2011

spdas said:
My sweet spot for charging is at 3.477V per cell. After I top balance, it allows for a "rogue" cell to go up to 3.55 and still be safe and a lagging cell to be as low as 3.40 which fully charged and into the top anyway where there is not much amps stored. When I charged at 3.49V each cell, I saw considerably more variance in the "no amperage" zone at the top. I will be lowering my charging voltage to 3.455, but it will take longer to charge at this rate.

Francis

I doubt you'll notice any difference unless you don't charge at the normal CC/CV mode.

EVfun · Nov 11, 2011

I played with the charge cycle tonight. I charged to 3.44 vpc (normally I charge to 3.50 volts) and held that until the current dropped to 1.2 amps (0.02C for my 60 amp hour cells.) Then I turned the charger voltage back up to 3.50 vpc. The current initially surged up to my 12 amp setting but as quick as I could get to the shunt (opposite side of the car from the charger) the current was already down to 2.5 amps and within minutes it was back down to 1.2 amps. As near as I can tell, I put in about 1/4 amp hour.

I bled the surface charge off with the headlights and then readjusted the charger down to 110.1 volts at 200 milliamps. I watched it stay at that voltage as the current dropped to 100 milliamps over perhaps another 15 minutes. I will try this lower charging voltage out for a while. If I'm only missing a small fraction of one amp hour and only adding about half an hour to the end of charge I don't see any reason to charge higher.

tomofreno · Nov 11, 2011

If I'm only missing a small fraction of one amp hour and only adding about half an hour to the end of charge I don't see any reason to charge higher.

I've done similar tests, just watching my Ah counter during charging, and concluded the same, there isn't much charge above that point. Others have done similar tests with similar results, so I think that is well-established now.

ev4me · Nov 11, 2011

Has anyone tested the effect of charging voltage on charging speed? That's where I think the tradeoff may be. Setting a higher charging voltage allows you to stay in CC mode longer.

JRP3 · Nov 11, 2011

bgeery said:
How's that for a first post?

Not that good

ElectriCar · Nov 12, 2011

FWIW, I adjusted my charger down tonight from about 3.46VPC to 3.44 and found the charge only about .5Ah less. Taking it back to 3.45 added about 0.2Ah to the pack. This VPC voltage was the average based on the voltage reading on a 50 cell pack.

sickness1 · Nov 14, 2011

If you look at this datasheet (I pick this one for example purpose):
http://liionbms.com/pdf/thundersky/TS-LFP100.pdf

You will see that the batteries should be charged above 3.33 Volts, in fact the manufacturer tells clearly that, to charge the batteries is mandatory to charge them (in the case of this datasheet) to 4.25V. The charge has 2 phases in this case:
-1phase, current limited charge:
In the graph you can see that charge is current limited at aprox 50Amps and is maintained till the battery reaches 4.2V (the maximum voltage recommended by manufacturer)
-2phase, voltage limited charge:
As you should not go through the voltage of 4.2, when this voltage is reached you limit voltage to this point and wait till the current drops till practically 0Amps.
Doing this charge you are not overcharging the battery nor breaking it, all the manufacturers recommend this kind of recharging (you can see for example yuasa batteries datasheet, to see another kind of battery charging method).
And if you look the discharging of the battery, the beginning voltage is not 3.3V, is 4.2V. The thing is that this kind of batteries have their nominal voltage in 3.2 - 3.3V but that doesn’t mean that they are not able to work at 4.25 or 2.7V. If this would be real then you wouldn't be able to discharge batteries below 3.2 because you would be overdischarging it. So your operative range for the battery would be between 3.2 and 3.3V (the 90% of the time batteries will be between those values at a 30ºC temperate) but you would be loosing between 5 and 10% of battery capacity.

JRP3 · Nov 14, 2011

That's an out of date data sheet, you do not need to charge to 4.25, or 4.20, and I think TS now recommends 3.8 as max charge. I'm pretty sure some people damage their cells by thinking they had to charge to 4.25 each time, when in fact that will shorten their life by increasing the speed with which the electrolyte breaks down.

sickness1 · Nov 14, 2011

As I told it was a datasheet with example purpose, another thing is that thunder-sky changed the maximum voltage of the battery because they noticed that it wasn't correct.

Normally there will be no problem charging at the maximum current that our chargers can afford, if we have a 18Kwh pack and we charge at 3Kwh we will need 6hours to charge, so we will be charging at 1/6C, battery isn't going to suffer at this current, is worst when we are driving and even worst when we let the battery capacity drop below the 10% of its nominal capacity. If you look at the battery life cycles you will notice that thundersky indicates 3000 cycles with 80% DOD and 5000 cycles with 70% DOD. I don't think many people will reach 3000 cycles since there are a lot of factors that will influence in this cycles, factors like temperature, current, DOD...

Jan · Nov 14, 2011

Charge as slow and low as you can afford, this will prolong their life.

But I don't think there are any commercial charger that give you both of these choises. A amp throttle is the maximum you can get. But a %SOC throttle is imho a must too.

JRP3 · Nov 14, 2011

You can limit SOC with your finishing voltage and charge current to some degree. I don't think there is any point in going below 80% SOC anyway. 3000 cycles on a 50 mile pack is 150,000 miles. At 10K miles per year that's 15 years. I hope to have better cells available long before that.

Jan · Nov 14, 2011

JRP3 said:
You can limit SOC with your finishing voltage and charge current to some degree. I don't think there is any point in going below 80% SOC anyway. 3000 cycles on a 50 mile pack is 150,000 miles. At 10K miles per year that's 15 years. I hope to have better cells available long before that.

Yes, you're probably right. But charging 80% SOC normaly and once in a while 100% is not an easy option with most commercial chargers.

JRP3 · Nov 15, 2011

The Winston battery page is showing 4.0 max, http://en.winston-battery.com/index.php/products/power-battery Sinopoly is showing 3.8
http://www.sinopolybattery.com/html/products_batteries.php
Additionally we know that keeping the cells at lower voltage prolongs their life, regardless of what manufacturers show for maximum voltage. You don't need to fully charge LiFePO4 unless you need the full capacity.

bgeery · Jan 24, 2012

In my original post I was under the false assumption that 3.33V resting was a full cell. I've since learned the real value is 3.400V. This also clearly makes sense, as the manufactures states the float voltage as 3.4V. They know regardless of how long you let it trickle, current will cease and hold at 100% charged.

This still makes me caution those that are holding >3.5V all the way to 0 current. What really frightens me is those that are top balancing by connecting all their cells together in parallel and then charging to 3.6V or even higher, and holding that voltage "for days". I will bet a set of new batteries they are harming the cells to some unknown degree and plating the cells by this practice. This demonstrates another advantage of bottom balancing, where connecting all their cells together in parallel and then holding the voltage at some low level "for days" does no possible harm to the cells. I do understand that some folks can't adjust the CV setpoint of their charger so a bottom balance isn't practical, but please minimize your time above 3.4V if you want to maximize cell life.

dtbaker · Jan 26, 2012

in the average charge cycle I doubt the cells are above 3.4 for more than a few minutes... they climb to 3.65vpc rapidly, switch to CV, finish, and start dropping all within minutes from what I've seen.

The original initial top balance you hope to catch within an hour or two, not weeks. The best guess when starting initial charge is that the cells are about 1/2 capacity, so just figure aproximate charge time based on your power supply and total 'missing' amp-hours.

no worries.
d

You are all destroying your LiFePO4 cells!

Top Contributors this Month

Recommended Communities