Lifepo4 never reaching full charge - panel output ample

[littleharbor]

Hmm. I guess they do not have any quality control department. One wonders even if that panel will produce 150watts.

They are otherwise very nicely made panels. They perform as they should.

[Mike90250]

yep, some sort of a problem Get a voltage reading of each cell in the pack, both idle, and while charging.

[solaar]

Ok, all cells were spot on 3.3V settled this morning.

Charging now in ok sun and two are showing 3.34, two are showing 3.35. Presumably that's not enough variation to be significant.

Voltage is very slowly rising, but still looks a lot slower than it should be.

I had another look at the charge controller menu, and this option here is the only one I'm not sure about. Could anything here be a problem?

20170425_103327.jpg

[Sunking]

First LFP voltage is FLAT. Voltage will take a long time to rise, they are LFP batteries. The ride from 3.4 to 3.6 is FAST. Look at Charge Discharge graphs. Stay away from the knees at either end. Note what happens when you reach around 3.4 volts, the voltage shoots up sharply (fast). Also not what happens around 3 volts, the voltage falls off a cliff, and that cliff is extremely dangerous. Fall off that cliff one time, and it is game over.

[solaar]

Does this mean that overall a LFP battery is going to take longer to charge than a LA battery, given the same current input? Or does it catch up in the end on those "knees" presuming you are charging it all the way?

I'm still experimenting with my LFP, and am comparing its performance to an AGM, but it seems ridiculously slow to the point where it's almost doing nothing. It was reasonably fast to get to around 13.2V under charge (from almost flat) then it just stagnates and you're lucky if you see the volt meter climbing better than 0.01V every half an hour.

Speaking of charging all the way, you suggested the GBS cells are slightly different and do require the 14+ voltages to charge properly; do they also need to be be occasionally topped up all the way? The manufacturer seems to think so, but I'm sure it's up for debate. And where would you set the float - 13.6?

[Sunking]

Does this mean that overall a LFP battery is going to take longer to charge than a LA battery, given the same current input? Or does it catch up in the end on those "knees" presuming you are charging it all the way?

I am a little short on time right now, gotta a Tee Time to make, so I will only address these questions for now. Short answer is no they charge faster the lead acid. The answer lies in the graphs. The Horizontal is TIME, and the ones I have show roughly 2 hours. However do not take away it will only take two hours on your batteries because that is not possible. You only have 120 wat panel and did not mention if you are using a PWM or MPPT controller. Assuming you are using a PWM controller, you are only charging at 6 amps, and that 6 amps is only at solar noon, less at all other times. 6 amps on a 60 AH battery is C/10 which is fine for a Pb battery but very slow on a LFP battery. Manufacture recommends a C/2 rate or 30 amps on your battery. You can charge slower (lower amps) without any ill effects, it just takes longer.

Look at the Charge gaph above and note the Orange Dot. That is roughly the 90% SOC point and where they recommend stopping. You do not want to fully charge your batteries. You will double their cycle life if you stop at 90%.

Now I have reread your thread and noticed something. How did you perform the Initial Bulk Balance of the cells? You cannot do that with a BMS or solar. It is just not capable of doing that. To perform the initial Balance requires putting all the cells in Parallel, walk away for a day, then decide if you will either Bottom Balance or Top Balance. Either Way requires a special charger/discharger to do that. You either discharge them to 2.5 volts, or charge them at 3.6 volts until charge current stops. Once Balanced, they stay balanced unless you use a BMS which is a parasitic load.

[Raul]

I have a victron 75/15 and I don't get a screen as you posted . I think you are going the long way round it. How did you set up the parameters of the CC? There is a Bluetooth dongle that gives life data to your smartphone or similar or download mppt prefs. If I remember for LFP you have a pre set file that you can download and update the device . At the moment it looks like is set to charge victron packs at 13.8v.

[Sunking]

Speaking of charging all the way, you suggested the GBS cells are slightly different and do require the 14+ voltages to charge properly; do they also need to be be occasionally topped up all the way? The manufacturer seems to think so, but I'm sure it's up for debate. And where would you set the float - 13.6?

There is nothing special about GBS cells other than they cost more than all the other Chi-Coms and lower quality. You are paying for the fancy covers and 4-point Term Bars. At first I suspected you might be using Thundersky Turds aka Winston Cells which do use a higher charge voltage.

I keep repeating myself because it is not getting through to you. You do not want to fully charge your batteries. You simply set Bulk = Absorb = Float = 13.6 to 13.8 volts. You have to find the sweat spot so when charge current stops the cell voltages rest at roughly 3.4 volts. That will be roughly 80 to 90% SOC.

When you do that your BMS is worthless. It will never ever turn on. You have what is called a Dumb Distributed BMS using Bypass Balance Boards aka Vampire or Bleeder Boards. When the cell voltage reaches 3.55 to 3.6 volts they turn on and Shunt a portion of the charge current. They do not SHUNT all the charge current, just a fixed small fractional amount. I assume you have the smaller ones that only Shunt 300 mah. So if your Controller is pumping 6 amps, and a cell reaches full capacity at 3.6 volts, it Shunts 300 ma, and still is being overcharged with 5.7 amps. cooking the cell. The Vampire Boards will continue to bleed the cell until it reaches 3.55 volts and shut off hopefully.

So to make the BMS work your charger has to be set to 14.4 volts or higher. If you set it to 13.6 to 13.8 volts, the BMS never turns on and does anything. In a higher end BMS requires a smarter charger, a very simple charger with only one mode CC/CV. I build racing Golf Carts and EV's for grins and giggles. I have a 26S Nissan Leaf battery, and I do not use a BMS, and I Bottom Balance. Most DIY EV guys Bottom Balance and do not use a BMS because they are to expensive for a good one, and can only afford what you have, and that type is the number killer of LFP batteries. It does so by over charging the crap out of the batteries, and when the Vampire Board fails, it fails shorted and completely drains a cell which is instant death.

Now for the guys that do use a BMS use a Smart BMS that integrates with the Charger. The Bypass Boards have higher bypass current and the voltage at which they operate is programmable. They use a lower voltage so as not to fully charge the cells. If and when a cell ever reaches the trip voltage will signal the Charger to cut back charge current to say 3 amps which the Bypass Boards can Shunt 3-amps so that cell does not receive any more charge. When the last cell Bypass Board turns on, signals the charger to terminate charge.

But here is the trick. They never charge to 100%, and it is a rare event they ever perform a Balance Charge. Only when and if needed once a year or less. They do exactly what commercial EV Manufactures do. No commercial EV would ever allow the customer to fully charge or discharge the batteries. That is the only way they can offer warranties. Otherwise they would go bankrupt with warranty claims. Most Commercial EV's operate at 80/20 never allowing you to go above 80% or below 20%. Some use 90/10. Now stop and think about that and why are you insisting to charge to 100%. Rarely do Commercial EV's ever do a Balance Charge, only when needed. They have very smart BMS for the most part does nothing except monitor.

You need to focus on 4 voltages: 3 and 3.45 volts which are cell voltages. Also need to focus on 12 and 13.8 volts which is pack voltage. Never let a cell go below 3 volts or pack voltage go below 12 volts under discharge, and never allow a cell to go above 3.45 volts or 13.8 volts while charging. First thing to do is set the charger to 13.8 volt, and the LVD on your inverter to 12 volts. If you want a second added layer of protection requires a Smart Cell Monitor to signal the LVD if anyone cell goes below 3 volts while discharging, or signal the Charger to shut off any anyone cell goes above 3.45 volts. But on a 4S system is way overkill.

But first anwer this question. HOW DID YOU PERFORM THE INITIAL BULK BALANCE? Does not sound like you ever did that. Shame on you.

[LETitROLL]

Does this mean that overall a LFP battery is going to take longer to charge than a LA battery, given the same current input? Or does it catch up in the end on those "knees" presuming you are charging it all the way?

I'm still experimenting with my LFP, and am comparing its performance to an AGM, but it seems ridiculously slow to the point where it's almost doing nothing.

Battery voltage is a good indicator of state of charge on Lead acid batteries, you are assuming it should be the same on LFP, that is the problem, look at the graph SK provided, 2/3 of the entire width of the graph is almost flat, that is showing you that the voltage is not changing much (just as you are experiencing) , LFP batteries are more voltage stable until they get near the edges and then they rise or fall rapidly otherwise the voltage is suppose to be steady, if you have a LFP 40% charged and you charge it up to 70% it will have more energy stored inside and be able to deliver more AH to a load, but it will not be obvious by a dramatically different voltage, the total energy stored in the battery has increased in AH but you cannot check that with a simple voltage reading, good chance your batteries and system are just fine.

[Sunking]

Well stated LETitROLL.

To the OP you are not grasping what the graphs are telling you. Start by looking at the discharge curve Blue Line and 2 Orange Dots. Initially at the beginning of the discharge cycle on a 100% SOC battery the voltage is 3.4 volts and in 1 minute drops to 3.27 volts. First thing this tells you is there is not any significant energy above 3.27 volts on a battery on discharge. There was only 2 Ah of a 195 AH cell.

Now follow the Blue Line to the Orange Dot which is 80% DOD or 20% SOC if you prefer. The voltage dropped from 3.27 volts to 3.14 volts some 98 minutes later with a C/2 Discharge. That is only a difference of .13 volts from 99% SOC to 20% SOC. That is 79% of the battery capacity in a span of .13 volts. On 4S only a 0.5 volts.

OK now look at the charge graph Red Line charge current of 100 Amps (C/2), Gray Voltage Line, and the Orange 90% SOC Dot. The voltages are roughly .1 volts higher than you will experience because they are using a C/2 charge current. Initially if you look closely the voltage is 3.3 volts, and in less than 1 minute goes up to 3.41 volts. Again telling you there is very little energy on the knees. 90 minutes later the SOC is 90% and the voltage is up to 3.5 volts a difference of 0.09 volts. If that were a 4S battery it starts at 13.64 volts and at 90% 14 volts. 30 minutes latter we reach 100% SOC, charge current saturated and dropped to 8 amps, and final cell voltage of 3.6 volts. So from discharged to fully discharged is 3.3 volts to 3.6 volts, and .3 volt difference. 90% of the capacity is a span of 0.09 volts in 90 minutes.

So the battery charged fast in two hours. Take away here is stay away from the Knees of the Curves. When you do that there is on only a 0.1 volt difference between discharge and charged. The voltage change will be slow despite the fact if you are charging fast. You are trying to watch paint dry or grass grow.

Here is a fairly accurate model mind experiment to use. to visualize using your battery and C/6 Charge rate. If we took your batteries and assume they are discharge, and put them on a CC/CV AC powered charger set to 3.6 volts and a 6 amp charge current.

1. When we start charging the voltage is 3.3 volts.
2. Ten minutes th evoltage rises 0.1 volts to 3.4 volts.
3. roughly 6 hours later the batteries are 90% charged and the voltage has climbed another 0.05 volts to 3.45 volts.
4. We notice charge current begins to taper and 15 minutes later the battery voltage goes up another 0.15 volts to 3.6 volts, and now the battery is fully charged.

That is what the graph is telling you.

[joerossjr]

WOW... not to hijack the thread, but where do you (The "Sun"s) suggest I learn these fundamentals about LiFePos? I know what BMS means, but not what it does, or how to manage them like LA batts. I could google, but I'd eventually be here asking more questions... so I am hoping to get the best source to learn from you guys.

[Sunking]

WOW... not to hijack the thread, but where do you (The "Sun"s) suggest I learn these fundamentals about LiFePos? I know what BMS means, but not what it does, or how to manage them like LA batts. I could google, but I'd eventually be here asking more questions... so I am hoping to get the best source to learn from you guys.

Been doing this for over 40 years and I still do not know what a BMS as there is absolutely no industry definition. A battery charge can be called a BMS. But I will throw you a bone.

Davide Andrea is a friend of mine and and he has a decent book and excellent website with tons of info. The book is called: Battery Management Systems for large Li Packs. Don't let the name throw you because everything in the book applies to small Li batteries.

His web site has all kinds of stuff. You can find every Integrated Circuit (chips), chargers, BMS, and Lithium battery made. He also has simulators and tutorials.

[inetdog]

... He also has simulators and tutorials.

So you can just destroy the simulators with strange battery management ideas instead of expensive real batteries?

[joerossjr]

Been doing this for over 40 years and I still do not know what a BMS as there is absolutely no industry definition. A battery charge can be called a BMS. But I will throw you a bone.

Davide Andrea is a friend of mine and and he has a decent book and excellent website with tons of info. The book is called: Battery Management Systems for large Li Packs. Don't let the name throw you because everything in the book applies to small Li batteries.

His web site has all kinds of stuff. You can find every Integrated Circuit (chips), chargers, BMS, and Lithium battery made. He also has simulators and tutorials.

AWESOME!! Thank you very much!

[solaar]

I have a victron 75/15 and I don't get a screen as you posted . I think you are going the long way round it. How did you set up the parameters of the CC? There is a Bluetooth dongle that gives life data to your smartphone or similar or download mppt prefs. If I remember for LFP you have a pre set file that you can download and update the device . At the moment it looks like is set to charge victron packs at 13.8v.

I just have the USB connector that goes into the controller and you download the software from victron