diy solar

diy solar

EVE LF280N Capacity from Docan Power

Delta Sierra

New Member
Joined
Aug 22, 2021
Messages
26
Im posting here and deleting my previous post. I initially ordered 8 of the new style LF280K cells from Docan Power but emailed Jenny Wu @ Docan before purchase making sure the cells I was to receive were not going to be bloated. She advised the cells do have a slight bulge to them and recommended I take delivery on the older style LF280DK (EVE 280N I believe, as the LF280DK is just their internal naming for these cells.) as these cells have no bulge and satisfy a few of my other concerns.

I received them via ups promptly and they were in great physical shape. Barcodes show the oldest one manufactured on 12/1/2020 and newest one on 2/26/2021 so looks like they have been sitting around for about a year before I got them. I assembled them into two separate 12.8v packs with an overkill BMS on each of them. I ended up using .125" thick aluminum plate cut just wider than width of cells and just shorter than their height, one plate onto each end of the group, with 8 hose clamps banding them together tightly. Also used some 2mm thick rubber sheet in between each cell and the plates on the end for protection.

Wanting to go with a flexible connection between each cell rather than a rigid bus bar, I chose to arrange them physically in parallel and built my own 'bus bars' with 2/0 AWG cable and tinned copper lugs, hydraulically crimped. They actually have about 33% more cross sectional area than the 2mm x 20mm bus bars that came with the cells, can move slightly to account for cell expansion / contraction as well as have excellent contact area when tightened- unlike solid bus bars.

I initially connected in parallel with bus bars and top balanced them at 3.65 volts with a cv/cc charger until amps went to near zero. I then used my cables to series connect the cells for each pack, attached the bms and torqued each nut to 3 Nm.

For the capacity test, I put the assembled pack into my RV and connected my Victron BMV-712 smart monitor shunt in line between inverter neg / trailer neg & C- lead from bms so the shunt measures everything, without anything else bypassing shunt. I then reset history of smart monitor, connected 1000w heater to inverter and started logging data into excel every 5 minutes for the 3 hours it took to run the battery to the 2.5v per cell cutoff. Average amp draw through the tests were around 92 amps.

Ended up getting only 275.8 AH out of battery #1 (one test) and 272.9 AH out of battery #2 (ran two tests, exactly same result). These AH numbers are from the Victron app where it says "Consumed AH". Under history tab, discharged energy shows 3.5Kw for all the tests. In excel, I ran a formula that added up all the wattage drawn from the pack from each sample point and got 3514 watts for battery #1 and between 3477 (test 1) and 3487 (test 2) for battery #2.

I'm assuming the watts I would be expecting from these packs would be approximately 3.2v * 4 * 280 AH = 3584 watts correct? So- bottom line question here- did I get screwed on buying these cells or no?

I do realize that there is energy loss on the battery side of the bms that the shunt is not measuring, for instance the battery and bms temps raise approximately 6 and 12 degrees respectively and of course all the connections are not zero resistance either and get slightly warm to the touch (never hot). I'm assuming over the course of a 3 hour test it would be fair to expect 50 to 100 watts of energy loss from this?

So, perhaps I got a fair deal? Please let me know if my methodology for capacity testing these cells is correct and what your thoughts are as to my purchase from Docan :) I emailed Jenny Wu, will see what kind of resolution if any at all she may have for my situation.

Attached a few pics, they will eventually both be in a plano box wired in parallel for 7 kw of power and eventually in series for 24 v

batt2.jpgbatt1.jpg
 
Last edited:
It's possible that your shunt isn't perfectly calibrated. Unless you're using lab grade, calibrated equipment, I would expect a +/- 10% error margin, but hopefully less from a Victron. Even if it is accurate, unless you paid for legitimate "grade A" EV cells, then getting over 270ah from a 280ah cell isn't too bad.

273ah is 97.5% of 280ah, and 276ah is 98.6%. I wouldn't be upset.
 
The other interesting data points to look at are the individual cell voltages when the pack reaches its low cut-off point. This will give you an idea of how closely matched in capacity the cells are.

Also, in this post I gave results of measuring my battery with different meters that I have (none of them lab grade) and had a 4% spread in the capacity results on the same test, depending upon which meter I choose to believe.

When doing capacity tests, the amp hour result is the best one to look at as it negates losses from power used by other devices or losses in connections. This is because the electrons have to flow through all parts of the circuit regardless of which component is using the power, and the amp meter is just counting electrons. Whereas if you look at watts, a bad connection can create voltage drop and thus a lower watt hour reading at the meter.
 
It's possible that your shunt isn't perfectly calibrated. Unless you're using lab grade, calibrated equipment, I would expect a +/- 10% error margin, but hopefully less from a Victron. Even if it is accurate, unless you paid for legitimate "grade A" EV cells, then getting over 270ah from a 280ah cell isn't too bad.

273ah is 97.5% of 280ah, and 276ah is 98.6%. I wouldn't be upset.
I see what you are saying about the shunt however, the victron manual states current and voltage measurement are +/- 0.4% and 0.3%, respectively. Can't imagine being anywhere near 10% off.

I did pay for what was advertised as Grade 'A' Eve 280ah cells. I do agree with you that I am getting pretty near advertised capacity though. However, I do see Will is passionate about stating if your batteries don't pull at least their advertised capacity, you got ripped off :(

I appreciate everyone's response thanks guys!
 
I see what you are saying about the shunt however, the victron manual states current and voltage measurement are +/- 0.4% and 0.3%, respectively. Can't imagine being anywhere near 10% off.
The 10% is general. Your farthest number is 2.5% off. The closer is 1.4%. that can be the difference in a few degrees difference during testing. It's CLOSE by any standards. It's going to be exceedingly difficult to prove to the seller that those cells are lower in capacity if they're that close to spec, without using professional measuring equipment.
I did pay for what was advertised as Grade 'A' Eve 280ah cells.
They wouldn't have been made in different years if they were a matched grade "A" pack. Did you get test reports that match the serial numbers? Cells aren't "matched grade A unless you get a test report for that specific cell from the manufacturer. The CELLS, individually, might be grade "A" but if they were manufactured that far apart they most likely won't "match".

Most of the actual EV cells I have come from the same S/N set, they were made on the same day. A handful are even sequential.
I do agree with you that I am getting pretty near advertised capacity though. However, I do see Will is passionate about stating if your batteries don't pull at least their advertised capacity, you got ripped off :(
I don't think you got "ripped off" unless you actually paid for matched and batched cells.
I appreciate everyone's response thanks guys!
Happy to attempt to help!
 
Testing cell capacity by testing an assembled pack can be good in some ways and bad in other ways. It gives you the capacity at the battery level, which is what is meaningful for how you would actually use it. That's good. Unfortunately, the capacity of the battery is driven by the lowest capacity cell. One cell will get to 2.5V before the others, and that will cause the BMS to cut off discharge.

This is also why it is good to know the capacity of individual cells, so that you can assemble batteries with more closely matched cells. You may have four cells with over 280Ah, and four with less. If you put the higher-capacity cells together, that battery will actually have more than 280Ah of capacity. When the lower capacity battery cuts off the higher capacity battery may be able to sustain your loads a bit longer. If you put lower capacity cells in with the higher capacity cells, the batteries will both cut off earlier.
 
There's some error when you multiply by a flat 3.2v but cut out at 2.5v. Depends upon what voltage you started at and how fast the curve drops to 3.2v and then falls off to 2.5v at the end.
 
Testing cell capacity by testing an assembled pack can be good in some ways and bad in other ways. It gives you the capacity at the battery level, which is what is meaningful for how you would actually use it. That's good. Unfortunately, the capacity of the battery is driven by the lowest capacity cell. One cell will get to 2.5V before the others, and that will cause the BMS to cut off discharge.

This is also why it is good to know the capacity of individual cells, so that you can assemble batteries with more closely matched cells. You may have four cells with over 280Ah, and four with less. If you put the higher-capacity cells together, that battery will actually have more than 280Ah of capacity. When the lower capacity battery cuts off the higher capacity battery may be able to sustain your loads a bit longer. If you put lower capacity cells in with the higher capacity cells, the batteries will both cut off earlier.
Thank you Horsefly, this is a really good point!

Jenny from Docan Power got back to me and asked if I have tested the cells with either an EBC-A40L or EBC-A20 battery tester and generated a graph. I am actually looking unto buying a CBA V Pro like what Will uses to test battery capacities, minus the amplifier. For a little over $200 I would be able to test my batteries very accurately (over 99% accuracy) and like horsefly said, be able to assemble matched packs. Only bummer is the max draw is 200 watts. But then again, testing each cell individually this would still be around a .22 C rate.

I looked into the EBC-A40L on amazon and they are over $300 without many reviews, the EBC-A20's that I've seen on amazon are about half the price of the CBA V but there are no reviews on them and don't look like they would be very accurate.
 
Thank you Horsefly, this is a really good point!

Jenny from Docan Power got back to me and asked if I have tested the cells with either an EBC-A40L or EBC-A20 battery tester and generated a graph. I am actually looking unto buying a CBA V Pro like what Will uses to test battery capacities, minus the amplifier. For a little over $200 I would be able to test my batteries very accurately (over 99% accuracy) and like horsefly said, be able to assemble matched packs. Only bummer is the max draw is 200 watts. But then again, testing each cell individually this would still be around a .22 C rate.

I looked into the EBC-A40L on amazon and they are over $300 without many reviews, the EBC-A20's that I've seen on amazon are about half the price of the CBA V but there are no reviews on them and don't look like they would be very accurate.
I don't have the EBC, but I will say that I reluctantly got a battery tester, and thought it was expensive for something I would only use once, but I end up using it all the time. I can use it to test the ratings of power supplies, usb chargers, alkalines, and my vape batteries.

I would definitely recommend getting a tester, it's a cool gadget to have around.
 
A strange phenomenon I've noticed with this battery is that when charging is almost done (around 98% capacity) and voltage starts rapidly climbing up through 13.8 and beyond, voltage on cell #3 is about 30 to 40+ millivolts higher than the others (cells 1,2 & 4 remain pretty close).

Once the voltage cutoff is reached and things settle down, the cells are all very close in voltage. However as soon as it starts discharging same thing happens, cell #3 much higher volt than the others.

So, what I am trying is rapidly charging from 98% to full, then discharging back to 98% and repeating over and over to allow BMS to balance that one cell. Hopefully they will all reach voltage cutoff at same time eventually.

Strange as I top balanced this pack initially.
 
I don't have the EBC, but I will say that I reluctantly got a battery tester, and thought it was expensive for something I would only use once, but I end up using it all the time. I can use it to test the ratings of power supplies, usb chargers, alkalines, and my vape batteries.

I would definitely recommend getting a tester, it's a cool gadget to have around.
Oh that's good to know :)

Yes and I've seen it can also be used to test power supplies and solar panels too
 
Get the little fan tester from amazon its more accurate than the victron shunt, I have the 500A.
 
A strange phenomenon I've noticed with this battery is that when charging is almost done (around 98% capacity) and voltage starts rapidly climbing up through 13.8 and beyond, voltage on cell #3 is about 30 to 40+ millivolts higher than the others (cells 1,2 & 4 remain pretty close).

Once the voltage cutoff is reached and things settle down, the cells are all very close in voltage. However as soon as it starts discharging same thing happens, cell #3 much higher volt than the others.

So, what I am trying is rapidly charging from 98% to full, then discharging back to 98% and repeating over and over to allow BMS to balance that one cell. Hopefully they will all reach voltage cutoff at same time eventually.

Strange as I top balanced this pack initially.
Instead of charging fast, trickle charge it (<20w) so that #3 is always balancing and not climbing in voltage, this will allow the other cells to “catch up”.

That or try and bleed off cell 3 with a resistor.
 
Instead of charging fast, trickle charge it (<20w) so that #3 is always balancing and not climbing in voltage, this will allow the other cells to “catch up”.

That or try and bleed off cell 3 with a resistor.
Hi! Yes I actually ordered a 6ohm resistor and wired it up with some alligator clips and have successfully used it to bleed off some power from the cell.

What I've noticed is I the more I let the battery sit at the top of its charge, the less meaningful individual cell voltages become. It helps if I discharge it to 80% or 90% then start another charge cycle, making note of which cells are high. Then, I can apply the resistor to those cells for like 2 minutes each and repeat the process. It is working well.

My theory here is that the cells are not matched at all so even though I top balanced them very well, all that did is get them all to 100% capacity. Which, with differing capacities means when you charge back up from a discharge some cells have a head start vs others so they reach their cut off voltages at different times, thus requiring me to bleed off excess energy from some cells. I guess this is kind of like a bottom balance?
 
Get the little fan tester from amazon its more accurate than the victron shunt, I have the 500A.
I may try that, but I have a feeling the victron shunt is going to be much more accurate than a little Chinese fan tester. After all, the little fan tester itself utilizes a shunt to measure current as well.

I am now using an EBC-A20H battery load tester to test each cell by itself. The first one from my better functioning pack tested at 285ah so I will keep at it and update here.

I have a feeling the BMS itself, combined with all the connections of an assembled pack, has more energy wasted than one would think vs testing individual cells.
 
My theory here is that the cells are not matched at all so even though I top balanced them very well, all that did is get them all to 100% capacity. Which, with differing capacities means when you charge back up from a discharge some cells have a head start vs others so they reach their cut off voltages at different times, thus requiring me to bleed off excess energy from some cells.
This should not theoretically be the case. Look at page six of the beginner top balance document. The idea is that if your cells have different capacities, top balanced cells will all be discharging at the same rate. By the time you go to recharge, they will be at different levels (in terms of how much capacity is left), but they all went down at the same rate. So when you recharge, they should all get back to 100% at the same point. This theory doesn't quite hold in real life, but it is pretty close.
 
My theory here is that the cells are not matched at all so even though I top balanced them very well, all that did is get them all to 100% capacity. Which, with differing capacities means when you charge back up from a discharge some cells have a head start vs others so they reach their cut off voltages at different times, thus requiring me to bleed off excess energy from some cells.
That is correct, that a top balanced pack is not matched. I saw a good example that explains the difference between top balance and bottom balance with cells of different capacities. Imagine had a group of sticks of varying length. the length is the capacity and bottom balancing is analogous to holding the sticks in your hand and taping them on a table until the bottoms all line up. The top will be uneven and the shortest stick will hit high voltage first and presumably trigger the BMS to shut down. The analogy also works with top balancing if you turn them over and line them all up at the top. Then the weakest cell will hit the bottom first and also trigger the BMS. Either way the capacity of the pack will be limited by the weakest cell.
I have not seen any evidence that a top balanced pack will not remain balanced at the top as long as you do not disturb that balance by over discharging. Just like the group of sticks of uneven length, the cells which are different capacities can never be considered matched. In my mind, matched means equal capacities under similar tests.
 
I may try that, but I have a feeling the victron shunt is going to be much more accurate than a little Chinese fan tester. After all, the little fan tester itself utilizes a shunt to measure current as well.

I am now using an EBC-A20H battery load tester to test each cell by itself. The first one from my better functioning pack tested at 285ah so I will keep at it and update here.

I have a feeling the BMS itself, combined with all the connections of an assembled pack, has more energy wasted than one would think vs testing individual cells.
I thought soo too, but I had tested 4 cells together with JBD 150 BMS and I only read 271 AH and 3498 WH, when I tested cells by themselves on the fan tester It was giving 275-277 890 WH

Victron was the worst said 267 - 3471 WH for the 4 cell pack with bms.

So Fan tester dirtectly on cells = highest ah
fan tester with JBD test = middle ah
and victron 500a shunt = least ah
 
Last edited:
This is the one I have it has v-sense wires:


my main wires are 12 guage and the v-sense wires are 24 guage futaba servo wires.

I don't think you can beat it for the price.
 
I thought soo too, but I had tested 4 cells together with JBD 150 BMS and I only read 271 AH and 3498 WH, when I tested cells by themselves on the fan tester It was giving 275-277 890 WH

Victron was the worst said 267 - 3471 WH for the 4 cell pack with bms.

So Fan tester dirtectly on cells = highest mah
fan tester with JBD test = middle mah
and victron 500a shunt = least mah
Maybe I'm misunderstanding. The Victron Smart Shunt measured the lowest capacity? Does that suggest somehow that the Victron Shunt is less accurate?

To me it sounds like there is something that is not as accurate as the other. Or in this case, if one is correct, the others are not. For me, I'd put my money on the Victron Shunt, vs the other two.
 
I"m using 1/0 cable connected to the victron may be the guage of the wire is eating some ah?

the fan tester I'm using 12 guage silicon wire, its dubro brand.

the vicron would also shut down when the BMS would cut-off so I had to look in the history tab to see where it stopped, this is 1 thing I really don't like about this shunt is that because it has a red power it needs active power to work, as soon as the BMS goes to low voltage the shunt shuts down. I don't know if it has a delay in writing the data.

I think the victron is a good device to get an idea of state of charge, but not a good tool for capacity testing.

I saw will was using something like this and it has a secondary power supply, seems like this would be more accurate:


this way even if the BMS goes to lvc the shut will still be running.
 
I"m using 1/0 cable connected to the victron may be the guage of the wire is eating some mah?

the fan tester I'm using 12 guage silicon wire, its dubro brand.
There's really two separate matters. All three of the devices you talked about have a sense cable to measure the voltage across whatever is being measured. Without those sense cables, there can be voltage drops that would make it so they wouldn't know when to stop the test.

However, all three devices are measuring the current (amps) over time (hours). The current is the current, no matter how large or small the wire is. So if the wire is small, there may be some voltage drop, but the current through the wire is the current that goes through the tester / shunt. So there is no such thing as losing some Ah on the wire (or as you said, "eating some mah").

As for why the three may measure different current: They all three use a version of a shunt, which attempts to measure a voltage drop across a fairly low resistance piece of metal. To the degree that the resistance of the piece of metal is known to a high degree of accuracy, and the voltage measurement is highly accurate, they should see the same amps going through their respective shunts. However, the resistance of some shunts are not calibrated to a very high degree of accuracy, and/or the method of measuring the voltage across the shunt may not be very accurate.
 
Well, as I was explaining, the sense cables are only so you know when to stop the test. So it doesn't matter too much.

The current is really what you are measuring, which goes through the two main lugs on the shunt.

I wouldn't bother returning the Victron Shunt. It's pretty good as it is - for measuring current over time, which gives you Ah.
 
The other interesting data points to look at are the individual cell voltages when the pack reaches its low cut-off point. This will give you an idea of how closely matched in capacity the cells are.

Also, in this post I gave results of measuring my battery with different meters that I have (none of them lab grade) and had a 4% spread in the capacity results on the same test, depending upon which meter I choose to believe.

When doing capacity tests, the amp hour result is the best one to look at as it negates losses from power used by other devices or losses in connections. This is because the electrons have to flow through all parts of the circuit regardless of which component is using the power, and the amp meter is just counting electrons. Whereas if you look at watts, a bad connection can create voltage drop and thus a lower watt hour reading at the meter.
Hi! Yes this is a good point. Upon voltage cutoff on my capacity test, I had the following voltages on each cell: Cell #1: 2.694, Cell #2: 2.524, Cell #3: 2.811, Cell #4: 2.733

Upon recharging from this point, I can see why cell #3 reaches it's high voltage cutoff before the other cells- it has a head start on charging vs the others.
 
This should not theoretically be the case. Look at page six of the beginner top balance document. The idea is that if your cells have different capacities, top balanced cells will all be discharging at the same rate. By the time you go to recharge, they will be at different levels (in terms of how much capacity is left), but they all went down at the same rate. So when you recharge, they should all get back to 100% at the same point. This theory doesn't quite hold in real life, but it is pretty close.
Yes I agree and initially these were my thoughts exactly. If I a start with all cells at the same exact voltage, then draw same exact power from each cell, then recharge them all with the same exact amount of power each, they would arrive exactly back where they started from- all at same voltage.
However, there must be something more in play here with the physics of the energy flow which I am not comprehending. It could be the assumption that the same amount of energy is leaving each cell, which may not be true. For instance, if cell #3 is at a slightly higher capacity at the beginning, it will always be at a slightly higher voltage than the rest of the cells, which means even if amps are the same, different watts are leaving the cell. There may be other laws of energy working here that I don't understand either.

Another idea is that during top balance I am at very tiny amounts of current, however finishing a charge cycle with bms hooked up I am charging at approximately 30 amps, which could thus exacerbate differences between the cells which would not be apparent during the trickle charge top balance. Also, I am discharging at around 93 amps and charging at 30 amps, discharge could be a different efficiency out of the cells vs charge efficiency into the cells which may affect cells of varying capacity differently. There may also be a slight peukert effect affecting cells of differing capacities differently.

I finished a charge cycle last night at about 25amps up to high voltage cutoff and all the cells were within 15 millivolts which is pretty close. Today I will do the same and do a capacity test for 3 hours down to low voltage cutoff and see if i get any more capacity out of it.

Meanwhile, on my other assembled pack, I have so far completed individual cell capacity testing (at 20amp discharge rates- the max of the tester i'm using) and so far have gotten 285ah & 282ah from the first two cells. This is all pointing to two things- 1) Just as Docan Power stated, the higher the amp draw you test these cells at the lower their capacity (characteristic of cells that failed to make EVE's Electric Vehicle grade) and 2) an assembled pack has much more parasitic loss through bms and associated wiring which is not measured at my shunt.
 
Yes I agree and initially these were my thoughts exactly. If I a start with all cells at the same exact voltage, then draw same exact power from each cell, then recharge them all with the same exact amount of power each, they would arrive exactly back where they started from- all at same voltage.
However, there must be something more in play here with the physics of the energy flow which I am not comprehending. It could be the assumption that the same amount of energy is leaving each cell, which may not be true. For instance, if cell #3 is at a slightly higher capacity at the beginning, it will always be at a slightly higher voltage than the rest of the cells, which means even if amps are the same, different watts are leaving the cell. There may be other laws of energy working here that I don't understand either.
You can't really use the voltage to compare cells, because two cells can be at very nearly the same voltage but have completely different state of charge.

I've lost track here. Did you top-balance your cells before attempting any of these tests. The only way what I described works is if you get the cells to the same state of charge - by either top balancing or bottom balancing.
 
I thought soo too, but I had tested 4 cells together with JBD 150 BMS and I only read 271 AH and 3498 WH, when I tested cells by themselves on the fan tester It was giving 275-277 890 WH

Victron was the worst said 267 - 3471 WH for the 4 cell pack with bms.

So Fan tester dirtectly on cells = highest ah
fan tester with JBD test = middle ah
and victron 500a shunt = least ah
I have a feeling what you may be seeing is difference in capacity when testing at different amperage draws. I'm assuming when you used the fan tester on the cells directly you were pulling a max of 20 amps, with of course no losses associated with bms or bus bars. This would give the highest capacity reading.

Then, when you used the fan tester to test the assembled pack along with the bms, you introduced losses with all the battery cell connections as well as bms losses and you were also drawing only 20 amps max. This would be your middle ah capacity result.

Finally, when testing with the assembled pack, bms and victron shunt you were most likely drawing much more than 20 amps which amplify parasitic losses and super slight peukert (LiFeP04 has very minimal peukert effect though). This naturally would give you your lowest ah capacity rating.
 
I"m using 1/0 cable connected to the victron may be the guage of the wire is eating some ah?

the fan tester I'm using 12 guage silicon wire, its dubro brand.

the vicron would also shut down when the BMS would cut-off so I had to look in the history tab to see where it stopped, this is 1 thing I really don't like about this shunt is that because it has a red power it needs active power to work, as soon as the BMS goes to low voltage the shunt shuts down. I don't know if it has a delay in writing the data.

I think the victron is a good device to get an idea of state of charge, but not a good tool for capacity testing.

I saw will was using something like this and it has a secondary power supply, seems like this would be more accurate:


this way even if the BMS goes to lvc the shut will still be running.
Yes I know what you are saying with the victron shunt shutting down once the bms cuts power upon low voltage shutdown. When I do my capacity tests with the victron shunt I log data every 5min or so in excel and once it is very close to shutoff I log it every minute so I know exactly what the shunt reading was right before cutout.

When the power comes back on i've noticed the history tab will greatly exaggerate the ah consumed & Kw discharge vs what it read right before cutout.
 
You can't really use the voltage to compare cells, because two cells can be at very nearly the same voltage but have completely different state of charge.

I've lost track here. Did you top-balance your cells before attempting any of these tests. The only way what I described works is if you get the cells to the same state of charge - by either top balancing or bottom balancing.
Oh yes I did a full top balance on the cells, a few times actually. Connected all in parallel, followed the guide with positive from lab charger on cell #4, negative from the lab charger to negative on cell #1, verified with multimeter charger was outputting exactly 3.65v before attaching, hooked up and waited until zero amperage passing through before disconnecting.

This, of course gets all the cells to the same voltage but they may be of varying capacities. Also since they were brought up to the 3.65v at a very tiny rate (milliamps) individual cells (may not?) show their differences of voltage as much as if they were being charged rapidly i'm not sure.
 
Yes I agree and initially these were my thoughts exactly. If I a start with all cells at the same exact voltage, then draw same exact power from each cell, then recharge them all with the same exact amount of power each, they would arrive exactly back where they started from- all at same voltage.
However, there must be something more in play here with the physics of the energy flow which I am not comprehending. It could be the assumption that the same amount of energy is leaving each cell, which may not be true. For instance, if cell #3 is at a slightly higher capacity at the beginning, it will always be at a slightly higher voltage than the rest of the cells, which means even if amps are the same, different watts are leaving the cell. There may be other laws of energy working here that I don't understand either.

Another idea is that during top balance I am at very tiny amounts of current, however finishing a charge cycle with bms hooked up I am charging at approximately 30 amps, which could thus exacerbate differences between the cells which would not be apparent during the trickle charge top balance. Also, I am discharging at around 93 amps and charging at 30 amps, discharge could be a different efficiency out of the cells vs charge efficiency into the cells which may affect cells of varying capacity differently. There may also be a slight peukert effect affecting cells of differing capacities differently.

I finished a charge cycle last night at about 25amps up to high voltage cutoff and all the cells were within 15 millivolts which is pretty close. Today I will do the same and do a capacity test for 3 hours down to low voltage cutoff and see if i get any more capacity out of it.

Meanwhile, on my other assembled pack, I have so far completed individual cell capacity testing (at 20amp discharge rates- the max of the tester i'm using) and so far have gotten 285ah & 282ah from the first two cells. This is all pointing to two things- 1) Just as Docan Power stated, the higher the amp draw you test these cells at the lower their capacity (characteristic of cells that failed to make EVE's Electric Vehicle grade) and 2) an assembled pack has much more parasitic loss through bms and associated wiring which is not measured at my shunt.

The bigger the BMS the more loss it will have, a 100 a probably you will loose about 4-5 Ah, the 150 5-6 ah, and like a 200 or so 6-7 ah.
 
Yes I know what you are saying with the victron shunt shutting down once the bms cuts power upon low voltage shutdown. When I do my capacity tests with the victron shunt I log data every 5min or so in excel and once it is very close to shutoff I log it every minute so I know exactly what the shunt reading was right before cutout.

When the power comes back on i've noticed the history tab will greatly exaggerate the ah consumed & Kw discharge vs what it read right before cutout.


which is why fan tester is great, under $100, pretty colors on lcd screen and the has RGB colors also no shutdown & v-sense cables.

It should be the standard for the dollar!
 
This, of course gets all the cells to the same voltage but they may be of varying capacities. Also since they were brought up to the 3.65v at a very tiny rate (milliamps) individual cells (may not?) show their differences of voltage as much as if they were being charged rapidly i'm not sure.
The voltage at the top of the charge curve is very indicative of the SOC. Just look at the steepness of the curve versus a lower SOCs where the curve is very flat. Charging rapidly as in constant current has nothing to do with the shape of the curve above 3.6 volts.
Yes they could be varying capacities but there is nothing you can do to change that physical property. Top balancing is just balancing capacity at the top. Imagine a stack of popsicle sticks of varying length analogous to capacity. You can stack them where they are even at the top or the bottom but you can not change their capacity (length).
 
The voltage at the top of the charge curve is very indicative of the SOC. Just look at the steepness of the curve versus a lower SOCs where the curve is very flat. Charging rapidly as in constant current has nothing to do with the shape of the curve above 3.6 volts.
Yes they could be varying capacities but there is nothing you can do to change that physical property. Top balancing is just balancing capacity at the top. Imagine a stack of popsicle sticks of varying length analogous to capacity. You can stack them where they are even at the top or the bottom but you can not change their capacity (length).
Yes, that's a good point. I think what I meant to emphasize is that even though a group of battery cells are all at the same voltage, their capacities could be, of course, still different. So, after a fantastic top balance is completed where all the cells are at 100% SOC and I then proceed to take say around 500 watts out of each cell on a discharge, that could mean leaving cells 1, 2, 3 and 4 at around 45%, 44%, 46% and 45% SOC. Upon charging these to full at this point, some of these cells have a head start on the others as to reaching 100% so will reach their voltage cutoff sooner than the others.

One could theorize 'well, they all started at the same point and should return to the same point'. Perhaps, but perhaps not. Remember, we got these cells to their top balanced state by connecting in parallel and / or charging individually to 3.65v. I presume the energy's interaction with the cells is different in the aforementioned way versus now discharging and charging in series with a bms involved as well as many other sources of resistance. I believe this is one of the reasons individual cell voltage will drift over time and we need a balancing mechanism in place to bring them back.

This all points towards my idea that if the cells differ from each other significantly enough (be it capacity, chemistry, age, internal resistance, construction methods or materials, etc) they will, upon charging to their respective 100% SOC, all ramp up in voltage at slightly different times. Looks like there may be no real way of dealing with the resulting 'runners' outside of simply draining power from them individually so they line up (closer) with their counterparts, although they will probably only be able to get so close.
 
I managed to complete testing all my cells, individually, using a ZKE Tech EBD-A20H load tester. This tester connects to a computer and generates the pretty graphs, logging voltage and amps every few seconds. Just a chinese tester, but uses independent voltage sensing leads and is seemingly accurate as verified with my Klein DVM. Each cell took approximately 14 hours as this tester maxes out at 20 amps.

The results:
-------------------------------Battery 1------------------------------
Cell-----Ah--------Energy---Internal Resistance--Production Date
--1---285.1Ah---922.81Wh--------0.17mΩ-------------2/6/21
--2---281.4Ah---912.20Wh--------0.17mΩ-------------12/1/20
--3---282.2Ah---914.89Wh--------0.175mΩ------------2/15/21
--4---284.3Ah---921.66Wh--------0.17mΩ-------------2/8/21

Capacity testing assembled pack avg 94A mixed draw (about 20A from 12v loads, 74A from inverter w/ space heater plugged in), measured with Victron BMV-712 logging data every 5 min until voltage cutoff (every 1 min during last 10 min of test):

275.8Ah, 3514Wh
Individual cell voltage at end of assembled pack test:
Cell---Voltage
--1-----2.896
--2-----2.548
--3-----2.708
--4-----2.856

-------------------------------Battery 2------------------------------
Cell-----Ah--------Energy---Internal Resistance--Production Date
--1---279.6Ah---904.51Wh--------0.175mΩ------------2/25/21
--2---278.9Ah---902.38Wh--------0.175mΩ------------2/26/21
--3---281.1Ah---910.30Wh--------0.175mΩ------------2/2/21
--4---280.1Ah---907.80Wh--------0.175mΩ------------2/18/21

Capacity testing assembled pack avg 96A mixed draw (about 22A from 12v loads, 74A from inverter w/ space heater plugged in), measured with Victron BMV-712 logging data every 5 min until voltage cutoff (every 1 min during last 10 min of test):

272.9Ah, 3450Wh
Individual cell voltage at end of assembled pack test:
Cell---Voltage
--1-----2.693
--2-----2.527
--3-----2.795
--4-----2.723


Battery 1 performs very well, cell voltage stays very close to each other (within 10 or 15mv) all the way to 3.65v cutoff.

For Battery 2, cell #3 likes to start climbing faster than the others around 3.5v with about a 25mv delta over the others. I have used a resistor to try to tame this cell's behavior closer to the others. However, after attaching this 6Ω resistor for only about 10 minutes while charging pack at around 5A, once getting to about 3.55v, the other cells climb fast and leave cell #3 in the dust.

I believe I will have to continue to play a bit of whack a mole with the resistor and repeated run-ups to volt disconnect to average out the differences. I imagine it unlikely to get the delta below 25mv overall but i'll be happy with that, given the circumstances.

I began on this journey with a flawed understanding of quality of cells I would be receiving from Docan but after reading on this great forum I now understand there is only so much one can expect from these electric vehicle battery rejects. I have a JK-B2A8S20PH BMS ordered, should arrive in about a month. I believe the 2A balance current will sufficiently keep these cells behaving well together, much more so than my current Overkill Solar JBD BMS could ever do with it's 50mA balance current. Plus, i'll be able to use it when I go 24 volt with these cells, yay!
 
Back
Top