Tweet
Testing amp hour capacity of each individual cell in a 100ah LifePO4 12v pack.
Graphs linked in the description of the video.
Graphs linked in the description of the video.
-
-
OK very good, I have the exact same equipment I use on Model Planes and EV from time to time.
To answer your first question about the Revolectric Power Supply it is 1320 watts not 1000. 24 volts x 55 amps = 1320 watts maximum. No big deal. In fact very good, as you now own a charger that can charge any battery of today and tomorrow up to 60 volts. Great match-up of power supply to charger. They were made for each other which makes a wonderful marriage.
See a couple of issues with your method, again may be I do not not understand exactly what you said in the video .
Issue 1. Did you fully charge each cell before did the Discharge AH test? I assume yes, if not then...
Issue 2 How did you do a full charge? This is the kicker question because if you used PL8 Default algorithm would have left you slightly undercharged. In the PL8 firmware it is programmed to terminate charge when the current tapers to C/10 of the charge current selected. That can cause you to either undercharge or over charge the battery. No problem you just have to be aware of it and how to work around it.
To fully charge a GBS 100 AH cell requires you to CC/CV charge at 3.65 vpc (default is 3.6) and Terminate when charge currents tapers to 3% of C. Well 3% of C is 3 amps. So the trick is to select a Charge Current where C/10 = 3 amps or 10 x 3 = 30 amps. Set charger to charge to 30 amps @ 3.65 volts. Terminate when current tapers to 3 amps. That is 100% fully charged.
Charge Wattages you quoited do not jive with the parameters given. If you are charging at 30 amps x 3.6 volts is 108 watts. Where were you coming up with 200 watts? To be 200 watts would require 60 amps of charge current at 1S. Your numbers do not jive. I suspect I have the details wrong.
Lastly your Internal Resistance Measurement are complete Bogus or you have terrible batteries. It is True GBS has the highest Ri of all the Chi-Coms, thus why EV guys do not use them as it robs them blind, shorter cycle life, and heating issues. GBS 100 AH cells run 3 to 4 milli-ohms. You are saying 13 to 14 milli-ohms. OK let you in on a Secret with the PL8. Its Ri measurement is JUNK. Well known fact. They use pulses rather than Delta. Result is make believe readings.If you want to know how to measure Ri, just ask. Your PL8 is Junk when it comes to Ri.
Lastly lets work on your terminology. You used the word SAG to describe Rested Open Circuit Voltage. They are two completely different things. Sag is the voltage drop incurred from OCV to fully loaded Current. It is proportional to Internal Resistance. Example if you have a 3.35 open circuit voltage fully charged up GBS battery with a R= .013 ohms and apply 20 amps the voltage Sags from 3.35 volts down to 3.35 - [20 amps x .013] = 3.09 volts which really sucks as that is a 8% loss of power of Voltage Sag. If you had a good 100 AH cell with only 2 milli-ohms you would only sag down to 3.31 volts or a 1.9% power loss. As you can see from this example if your cells are really 13 to 14 milli-ohms, you have sukass batteries.
OCV or open circuit voltage of a rested battery has nothing to do with Sag. Lithium is no different than any other battery, just different degrees. Example you charge a 12 volt FLA battery with 14.4 volts just like a 4S LFP. Let them rest until the surface charge bleeds off. When that happens a 12 volt fully charged Pb battery = 12.6 volts and LFP around 13.4 volts. Voltage meaningless in both batteries. Only useful if OPEN CIRCUIT and allowed to rest. Apply any load or charge and voltage is meaningless.
On the discharge side you left fuel in the tank. As you noted you only discharged to 2.8 volts. Empty if 2.4 to 2.5 OCV. There is that word again. Two ways to get there. set your set point below 2.5, or even better fast discharge at 10 amps until it complete, then repeat ay 1 amp to 2.5 volts until you rest at 2.5. No big deal as we are only talking 2 or 3 amp hours difference.
So now you are ready to find your Magic Number. You know weakest cells is 103 AH. Do a proper Bottom Balance until the cells are OCV to 2.5 volts. Do this by connecting all cells in parallel and fully discharge them to 2.5 vpc fully rested. 2.4 to 2.5 is right, but every cells should be exactly the same at the bottom. That is done by wiring them all in parallel. Let them rest over night still connected in parallel at 2.4 to 2.5 volts. Don't worry th eline you do not cross is 2 volts. 2,5 is a lot of room for error.
Now reconnect the cells in Series, use your PL8, at Fast Charge mode, no Balance Plug and pump in 90 to 95 AH and Terminate the charge. Disconnect the charger and allow the batteries to rest over night. Next morning measure the rested voltage. That is the Voltage you use on your charge controller to float at.
EDIT NOTE:
Give you a tip. Build you a 4S Balance Plug to plug into the PL8. Gives you the ability to monitor all 4 cells with alarming. Also give you the ability to Top Balance is so inclined later as the PL8 can Balance Charge at 3 amps per cell with the Balance Plug.Last edited by Sunking; 07-25-2016, 12:41 AM.MSEE, PE -
I guess I didn't communicate it well, but this was a charge test only, not a discharge test. The discharges were only to get each cell down to 2.5v so I could run the charge test.
Hmm. Which charge algorithm should I be using in the future?
I didn't "come up with" 200 watts. I measured it with a kill-a-watt in real time and on camera. You saw me enter the settings. I agree the math doesn't jive, but I suspect you're just seeing accumulated losses.
Well, ****. Ok, how do I measure Ri? I'm marginally curious.
Fair enough. I'll work on that.
It takes 10 hours to do a discharge. I'm not waiting around for the phone to ring. The set point is 2.5v, but I think it's rising naturally over time until I finally come around and measure it. Like you said, it's not really a big deal.
I already did a video about bottom balancing, though I didn't wire them in parallel afterward: https://youtu.be/rZGLbAv2MCA I understand the reason for wiring in parallel but I'm not really sure it's necessary with so little energy in the cells, and I really don't think it's worth the trouble to remove the strapping. Not really planning to do another bottom balancing video.
I've already found the "Magic Number" as you call it for this pack. I have footage, but it's pretty boring so I'll probably just show the graphs and talk about them in the next video.
Yeah, you can't "pump in 90 to 95 AH" with a PL8. It only measures AH, it doesn't have a setting for terminating charge/discharge based on AH as far as I can tell. Gotta pick a voltage set point instead. More on that in the next video.
Comment
-
OK your Discharge Test was invalid then because you do not know if they were full or not.
First read up on the manual. But when you set up a Balance Charger, the charger has a Default Charge Termminate setting. Default I think is C/10 So if you select 10 amps charge rate, the Charge Terminates when the current tapers off to C/10 or 1 amp if you selected 10 amps charge current. With me so far?
To fully charge a lithium battery, yours, you charge to 3.65 volts, and terminate when charge current drops to 3% of C. For you C = 100 Amp Hours so 3% is 3 amps. So if you used say a 10 amp rate, it cuts off at 1 amp thus over charging the battery somewhat. But if you selected say 30 amps the charger would cut-off at 3 amps. Perfect.
OK then you have serious metering errors in your kill-a-watt meter. You PS and PL8 ar over 90% eficient. That means the max power going into the charger should be no more than 125 watts if charging at 100 watts.
Why wait around for anything. Go to bed, go to work, go do something as the test does not need your presence. ? If it were me, I would just turn on a 1000 watt load and let them go until the Inverter cuts off on low voltage. Them wire all of them up in Parallel, set the PL8 to 2.4 volts and walk away for a day.
But you are right, no big deal if you Top Balance. If you BB you need to get the Balance done correctly.
More later when I get time.Last edited by Sunking; 07-25-2016, 07:42 PM.MSEE, PEComment
-
OK back to how to measure Ri of a battery. It is extremely easy to do and even easier if you have a PL8 to work with. Here is a link to a Sticky located here. However I will go through it personalized to what you have. We have to make one change to the method I linked to because it needs modified to measure lithium cell. We perform the test with Lithium at Partial State of Charge in the range of 30% to 70 SOC range. Just somewhere in the middle. Short story around 3.3 volts.
OK conditions to be met are:
Battery in PSOC range
Battery temp normal romm temp
All voltages are taken directly from the battery TERM POST
Test Equipment and Hardware required are:
PL 8 set up to discharge at 1 and 10 amps.
Good quality DMM with a minimum 3.xxx resolution or better.
Pencil and paper
Calculator
Protein Computer armed and dangerous.
We are going to use the Delta Voltage/current to find Ri testing at 1 and 10 amps. Exactly how a $3000 Digital Low Resistance Ohm meter does it with high current for accuracy and resolution.
Method
Set your volt meter to read 3 volt range directly on the battery Term Post and from no place else.
set up PL8 to discharge at 1 amp.
Connect to battery, and apply 1 amp discharge.
While on Discharge read and record battery voltage and call it V1. Example V1 = 3.297 volts
Record Current 1 as I1 = 1 Amp.
Repeat test at 10 amps.
Record voltage, call it V2 = 3.270 volts
Record I2 = 10 Amps
OK now it is just a little math
1. Find the difference in voltage by taking V1 - V2 = Delta Voltage. so Dv = 3.297 - 3.270 = .027 volts
Solve for Delta Current by taking I2 - I1 = DI = 10 amps - 1 amps = 9 amps
Now solve for resistance of cell R= Dv / Di = .027 volts / 9 amps = .003 Ohms or 3 mili-ohms.
MSEE, PEComment
-
Super easy to do. No more than 2 minutes to test all 4 batteries.
The numbers are usefull as it is a Battery Health Monitor or Age Meter. They are new, thus a refference point in time or standard to measure against. It is one of the most significant parameters. It is directly related to capacity, age, and performance. Very easy to read and interpet. The readings you take today are perfect 100% Healthy batteries. As they age, become damaged, loose capacity the resistance is going to go up and up to useless.
You should measure 2 to 5 milli-ohms on each cell. Your PL8 showed 13 to 15 right? Ignore it. If you measure 10 to 15 this way you gotta problem. Anyway when you see the resistance going up, you know the end is near.
There is just one catch, the TEST conditions must be met to be valid: SOC in the middle, rested cells, room temp cells and room. and same exact test setup and equipment. Making very small measurements requires attention to details. To be more accurate 10 times more accurate instead of using 1 and 10 amps, use 10 and 100 amps. For the high current test you could use an Inverter with a large load. When you test just record both Voltage and Current. Say 10 amps with the PL8 and 63.5 amps using a blow dryer on Inverter. You would need a way to measure the current.
DID YOU MAKE A BALANCE PLUG?Last edited by Sunking; 07-26-2016, 04:28 PM.MSEE, PEComment
-
You make good videos!
If you mean the flat spot on the voltage graph, that is where the charger has gone into CV mode and is keeping the voltage constant by decreasing the current until it decreases to 1 amp at which point the charger terminates the charge. This decreasing of the current reduces the rate at which the battery is being filled up so the SOC graph on the right hand side flattens off. It has nothing to do with overcharging.
What does overcharging mean? It could mean putting too high a current into the battery at any stage of the charge cycle or it could mean going above a certain charge voltage or a combination of both of these. The voltage and current are not fixed or rigid numbers, they should be chosen depending on the battery specifications and the circumstances that the battery is being used in.
Lithium batteries are not like a bottle that we fill up and it overflows if we overfill/overcharge it. They contain a cathode and an anode. When the battery is empty the cathode is full of lithium ions and the anode is empty. When the battery is full the anode is full and the cathode is empty.
If we represent the anode and cathode as two sponges and the lithium ions as water, an empty battery is where the cathode sponge is full of water and the anode sponge is empty. Putting a voltage across the battery to charge it is like squeezing the cathode sponge, the harder we squeeze it the faster the flow of water/current out of the cathode sponge. When the battery is nearly full it gets harder and harder to squeeze the water out of the cathode sponge. We have to increase the voltage/squeeze harder to get more lithium ions/water out. We can't get all the lithium ions/water out of the cathode sponge, there will always be some lithium ions/water left in the cathode sponge, the amount left depends on the voltage that we charge the battery too/how hard we squeeze the sponge.
Simon
Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor
Latronics 4kW Inverter, homemade MPPT controllerOff-Grid LFP(LiFePO4) system since April 2013Comment
-
Thanks! I try (sometimes).
Hmm. ok. I wasn't aware it was possible to put too much current into the battery. I thought it would simply accept what it would accept and that was the end of it. I thought that was why the current tapered off at the end of a charge.
What is doing the squeezing in this analogy? voltage or amperage?
I guess I always thought of LifePO4 batteries like a standup paddle board being pumped up. The voltage is like the pressure of the pump. If you set the pump to 12 PSI, the paddleboard "charge" will complete and the paddleboard will be full of air and usable, but it can still accept more air. If you set the pump to 15 PSI, that's like using 3.6v on a LifePO4 cell. It's the max the cell or paddle board can safely handle. You can increase the voltage or air pressure beyond that, but you'll risk damaging the paddle board / battery from over pressure.
The amperage in this case would be the size of the air tube or maybe the capacity of the pump for each throw of the handle. Toward the end of the charge, it becomes more difficult to move the handle up and down because you're resisting all that air pressure in the paddle board.
But I guess this analogy isn't perfect. I'm hearing from you and sunking that too much amperage can overcharge too. I need to go back to re-read his post. Clearly I'm still a little fuzzy on that.
Comment
-
Charge current is determined by one of two factors; The chargers Current Limit, or the battery Ri. Example lets say we take one of your cells, it is at 3 volts, and we have an Infinite Current Charger set to 3.6 volts. What is the initial Charge Current? You should be able to answer this since you know each cell Ri. .
The current = 3.6 volts - 3.0 volts / 2.7 milli-ohms = 222 amps. In this case the battery Ri limited charge current and yours would be just fine with it. ou would only see 222 amps for a few seconds, then the current will quickly taper off as the battery voltage riss to meet the charger. Despite the 2C charge rate, it will still take at least 600 to 90 minutes for the battery to saturate and be fully charged up. Now tell me what the current would be if the cell voltage was at 2.5 volts, and 3.2 volts?
Now use a 10 amp charger. What is limiting the current? The charge now limits the current until your battery reaches 3.6 volts - [10 amps x 2.7 milli-ohms) = 3.573 volts.
Voltage is always the pressure. Look it does not matter what type of battery we are talking about, with the exception of Nickel based batteries, all batteries charge the exact same way. Ohm's Law does not change between battery types.
To charge a battery we need to be able to do at least two things. Regulate both Voltage and Current. In other words build a plain ole fashion DC power supply that we can set the voltage and limit current. Nothing more is needed.
Now if we want to charge fast, we need two more things. The ability to measure current,, and change voltage based on the current. We call that a 2-Stage Charger. In Stage we we crank up the voltage, say to 14.4 volts on a 12 volt system. We apply the voltage until the charge current tapers down to 3 amps on a 100 AH battery. Then when the current reaches 3 amps, we lower the voltage to 13.6 volts and FLOAT. All charge current stops, the battery is at 100% SOC and our charger will supply all power for the loads until we turn it off. What kind of battery did we just charge up?
From the voltage and current profile I just went through it has to be either LiFeP04, or a Lead Acid battery. Charger does not care, nor would the batteries as they have the exact same charge profile.
Now if I want slow and gentle all I would need is 1-Stage. Set the charger to 13.6 volts and walk away. Both batteries are still going to get 100% charged up, it will just take a little longer until the batteries saturate to 13.6 volts.Last edited by Sunking; 08-01-2016, 11:08 AM.MSEE, PEComment
-
Current is flow of electrons, like flow of water from a tap.The squeezing is the voltage. If you squeeze a sponge a certain amount you will first get a large flow and then this flow will taper off to nothing, squeeze a bit harder and you will get another flow that will taper off.
Your analogy is wrong, when you are charging a battery you are not putting extra electrons (air in your analogy) into the battery you are just shifting Lithium Ions from a lower energy state in the cathode to a higher energy state in the anode, when you discharge the battery you are shifting the Lithium Ions back to a lower energy state in the cathode. To use the analogy of two ponds, a waterfall and a pump the top pond is the anode, water running down the waterfall is current being discharged from the anode into the bottom pond(the cathode) and the pump is recharging the water into the anode again.I guess I always thought of LifePO4 batteries like a standup paddle board being pumped up. The voltage is like the pressure of the pump. If you set the pump to 12 PSI, the paddleboard "charge" will complete and the paddleboard will be full of air and usable, but it can still accept more air. If you set the pump to 15 PSI, that's like using 3.6v on a LifePO4 cell. It's the max the cell or paddle board can safely handle. You can increase the voltage or air pressure beyond that, but you'll risk damaging the paddle board / battery from over pressure.
The amperage in this case would be the size of the air tube or maybe the capacity of the pump for each throw of the handle. Toward the end of the charge, it becomes more difficult to move the handle up and down because you're resisting all that air pressure in the paddle board.
The problem with too much current is that due to the internal resistance of the battery that it heats up, the heat produced is proportional to the Current squared (Amps*Amps). Heat is bad for the battery because it causes unwanted side reactions. With high charge currents this heat can also get concentrated in hot spots which amplifies the problem.But I guess this analogy isn't perfect. I'm hearing from you and sunking that too much amperage can overcharge too. I need to go back to re-read his post. Clearly I'm still a little fuzzy on that.
Simon
Last edited by karrak; 08-01-2016, 11:15 AM.Off-Grid LFP(LiFePO4) system since April 2013Comment
-
407 amps, which exceeds the 3C (300 amp) max charge rate for this battery according to the specs page: http://www.batteryspace.com/lifepo4-...-2vx4-dgr.aspx
Interesting. This means you could damage a LifePO4 battery by attempting to jump a low voltage LifePO4 battery from a fully charged LifePO4 battery, correct?
Ri limits current on the high knee. Super interesting.
But if you change that voltage to 13.4v or 3.35vpc, you'll only charge the battery 40-60%. I proved that by measuring the AH in at that voltage using the PL8.
Comment
-
Huh. I remember reading batteries weigh more when fully charged, but doing some research now, it appears this is due to oxygen and CO2 venting, not electron movement: https://www.reddit.com/r/askscience/...ss_than_fully/ Good to know. I like the two pond analogy.
So, as I've been building my V3 portable solar generator, I've been thinking about this heat issue. I guess if you could adequately cool the battery you could charge at a higher rate, right? Prismatic cells are probably poor candidates for water cooling, but I guess this is another area where small cylindrical cells shine.
Comment
Comment