Started my top balancing

[Substrate]

@BretS - don't worry, it's a common mistake and for a one-off situation, not likely to lead to long-term damage.

You'll notice that Will uses much beefier supply to do his balance charging, so he can get up to 3.65v quickly and get the process over and done with. That's the key - don't spend a lot of time like days or weeks doing so.

What many miss without experience is disregarding the cumulative effects of TIME, and concentrating solely on voltages. Because that's immediately determined with a voltmeter and seems tangible to work with. TIME spent damaging your bank, especially if it is cumulative cycle-after-cycle isn't physically seen until later down the road.

So if it isn't seen right at the start, it doesn't exist.

So no worries. But had you come to me after buying that bank, I would have said "Did you also budget for a much larger charger supply to do this job that is at least capable of 0.2C at a minimum?

Long way of saying that you're good with this one-off event. But if you build another one and I catch you using a supply that is not at least 0.2C capable, well.....

 

[BretS]

@BretS - don't worry, it's a common mistake and for a one-off situation, not likely to lead to long-term damage.

You'll notice that Will uses much beefier supply to do his balance charging, so he can get up to 3.65v quickly and get the process over and done with. That's the key - don't spend a lot of time like days or weeks doing so.

What many miss without experience is disregarding the cumulative effects of TIME, and concentrating solely on voltages. Because that's immediately determined with a voltmeter and seems tangible to work with. TIME spent damaging your bank, especially if it is cumulative cycle-after-cycle isn't physically seen until later down the road.

So if it isn't seen right at the start, it doesn't exist.

So no worries. But had you come to me after buying that bank, I would have said "Did you also budget for a much larger charger supply to do this job that is at least capable of 0.2C at a minimum?

Long way of saying that you're good with this one-off event. But if you build another one and I catch you using a supply that is not at least 0.2C capable, well.....

Well, it would be nice for beginners if that point was made somewhere in the tutorials. When I went to buy a power supply, they didn't have the one that Will recommended at the time, so I went with another 30V, 10A thinking it was equivalent. I even posted about what I was using and how I knew it would take a long time, and not one person said that was not good. And now I'm concerned about the cells sitting there at 3.585V waiting on the rest to get done. I won't be able to discharge them until I get the entire system installed in my 5th wheel. I'm working on it, but not going to be able to get it operational for probably another week.

I plan to set my absorption to Victron's custom LiFePO4 charge profile recommendation from Will's sticky: 12V Battery: 14.1V
Do I also use his recommendation for Inverter Cut-off: 10.7V for a 12V??
Also, I need to determine correct settings for my REC active balancer...

 

[Just John]

Well, it would be nice for beginners if that point was made somewhere in the tutorials. When I went to buy a power supply, they didn't have the one that Will recommended at the time, so I went with another 30V, 10A thinking it was equivalent. I even posted about what I was using and how I knew it would take a long time, and not one person said that was not good. And now I'm concerned about the cells sitting there at 3.585V waiting on the rest to get done. I won't be able to discharge them until I get the entire system installed in my 5th wheel. I'm working on it, but not going to be able to get it operational for probably another week.

I plan to set my absorption to Victron's custom LiFePO4 charge profile recommendation from Will's sticky: 12V Battery: 14.1V
Do I also use his recommendation for Inverter Cut-off: 10.7V for a 12V??
Also, I need to determine correct settings for my REC active balancer...

So far you have one person saying everything is terrible.
I say, "meh, not so bad".

Eve at least gives lots of charge rates and discharge rates, and how capacity is different at different rates. I don't see anything in the spec sheet that says a 10A rate is bad.

 

[FilterGuy]

So far you have one person saying everything is terrible.
I say, "meh, not so bad".

Eve at least gives lots of charge rates and discharge rates, and how capacity is different at different rates. I don't see anything in the spec sheet that says a 10A rate is bad.

There is absolutely nothing bad with the 10A supply. It is just a mater of how long it will take to charge the cells. Furthermore, if you pre-charge in serial, the amount of time to top-charge charge is reduced drastically.

 

[Just John]

There is absolutely nothing bad with the 10A supply. It is just a mater of how long it will take to charge the cells. Furthermore, if you pre-charge in serial, the amount of time to top-charge charge is reduced drastically.

Correct. For me, I bought a 40 amp power supply, just to speed things up. I didn't buy it because 10 amps is bad.

 

[FilterGuy]

Why does everything on here, even Will's video, say 3.65 for top balancing?? I was thinking I was being conservative with going to 3.585...

Well, it would be nice for beginners if that point was made somewhere in the tutorials.

I just posted an update to my top-balancing tutorial with the following updates:

• 5/29/21: Appendix D: Added comments about wire size and type.
• 5/29/21: Section 2: Added note about target top-balance voltages.
• 5/29/21: Section 1: Added note about Power Supply Amps
• 5/29/21: Section 1: Added note that a higher voltage supply would be needed for pre-charging cells for 48V systems.

Some of your questions & concerns are at least briefly mentioned in the updates.

Top Balancing LiFePo4 Cells using a low cost benchtop power supply.

To get the paper, click on the orange at the top of this page. This is a tutorial with detailed steps on how I top-balance my LiFePO4 Cells using a low-cost benchtop power supply. To get the tutorial, click on the orange "Download" button...

diysolarforum.com

 

[BretS]

I just posted an update to my top-balancing tutorial with the following updates:

• 5/29/21: Appendix D: Added comments about wire size and type.
• 5/29/21: Section 2: Added note about target top-balance voltages.
• 5/29/21: Section 1: Added note about Power Supply Amps
• 5/29/21: Section 1: Added note that a higher voltage supply would be needed for pre-charging cells for 48V systems.

Some of your questions & concerns are at least briefly mentioned in the updates.

Top Balancing LiFePo4 Cells using a low cost benchtop power supply.

To get the paper, click on the orange at the top of this page. This is a tutorial with detailed steps on how I top-balance my LiFePO4 Cells using a low-cost benchtop power supply. To get the tutorial, click on the orange "Download" button...

diysolarforum.com

That's great! Hopefully other newbies will find it useful!

 

[Substrate]

Well, prove it to yourself. Charge at a very low rate, from a low state of existing SOC, and stop when you reach 3.45v / cell and no higher. Do a capacity test. Surprise, you are already fully charged. I don't recommend it because of the time degradation.

Do this test on a spare bank you don't care about.

Thing is, manufacturers aren't going to go into this much detail. TMI - too much info which might turn off a potential buyer - who in the end may not actually care or have the ability to do a capacity test.

This assumes that the bank you are testing is not wildly mismatched to begin with. And that the testing process itself which may take so much time degrades the cells.

Imagine something like this appearing in a manual. From a marketer's standpoint, that would contrast severely with the "for your active lifestyle" brochure.

 

[Substrate]

You know what? Don't harm your good bank testing this.

Get a programmable hobby charger so YOU can set the CV and current to whatever you want, and 4 lifepo4 32650's. Baseline them as normal with a capacity check and so forth.

Discharge them - you don't have to totally discharge, but down to at least 30% SOC simulating how you get them off the boat. Now use very minimal current to charge, and 3.45v as the CV limit and stop. Test again.

Scale your results.

 

[Just John]

Well, prove it to yourself. Charge at a very low rate, from a low state of existing SOC, and stop when you reach 3.45v / cell and no higher. Do a capacity test. Surprise, you are already fully charged. I don't recommend it because of the time degradation.

Do this test on a spare bank you don't care about.

Thing is, manufacturers aren't going to go into this much detail. TMI - too much info which might turn off a potential buyer - who in the end may not actually care or have the ability to do a capacity test.

This assumes that the bank you are testing is not wildly mismatched to begin with. And that the testing process itself which may take so much time degrades the cells.

Imagine something like this appearing in a manual. From a marketer's standpoint, that would contrast severely with the "for your active lifestyle" brochure.

Actually, I'm doing exactly this. Difference in capacity is about 2 amp hours.

First 16 done until they drop to .25 amps in.
Next, I'm in the middle of charge until current drops to 14 amps.

Like I said, about 2 amp hours. About.
It's the about part that says you should go ahead and charge to 3.65v and under an amp in. If you want to top balance, you get much more consistent results that way. If it was more consistent, I'd agree with you.

Left is charge to 3.65v and amps drop to .25, right, charge to 3.65v and amps in drop to 14.

277.9​	276​
271.8​	269.5​
278.3​	276.4​
278.7​	In progress
278.1​	276.1​
271.5​	269.4​
275.2​	274.1​
276.2​	275.1​

It takes less than 60 seconds for the cell to get to 3.35v when you put a load on it (40 amp load, either charge method).
I have the csv files if you'd like to look.

The entertaining thing (especially for anyone that claims voltage is an accurate state of charge measure when between the knees) I've been trying to find a good charge method to leave cells in an identical state of charge storage voltage.

Empty cell, charge to 3.33v and wait until amps in drop to 14 amps, wide range there. VERY, VERY wide range.

93.14
118.49
108.41
124.21
129.93
137.16
77.49

 

[FilterGuy]

Well, prove it to yourself. Charge at a very low rate, from a low state of existing SOC, and stop when you reach 3.45v / cell and no higher. Do a capacity test. Surprise, you are already fully charged. I don't recommend it because of the time degradation.

Do this test on a spare bank you don't care about.

Thing is, manufacturers aren't going to go into this much detail. TMI - too much info which might turn off a potential buyer - who in the end may not actually care or have the ability to do a capacity test.

This assumes that the bank you are testing is not wildly mismatched to begin with. And that the testing process itself which may take so much time degrades the cells.

Imagine something like this appearing in a manual. From a marketer's standpoint, that would contrast severely with the "for your active lifestyle" brochure.

Are you talking about the top balance target voltage or the voltage used for charge limit during normal operation?

My top balance voltage is higher than my BMS Over-Voltage Cell limit which is higher than the effective cell voltage for my charger limit (Charger limit divided by number of series cells)

Top balance voltage > BMS Over-voltage > (Charger voltage/number of series cells)

This helps avoid nuisance trips of the BMS during normal operation of the battery.

 

[Substrate]

This would be the top-balance target CV needing to be lowered for an initial charge if started from a low SOC to begin with and there was insufficient current when doing that first run.

My concern are guys paralleling 8, 100ah or more cells right from the box together for that first run. Ie, on your bench in parallel is a nominal 3.2v 800ah battery. Getting charged to do a parallel balance at only 10a from a low SOC. I'm seeing it here in these forums and elsewhere.

That's the equivalent of charging at a 0.013C rate. That's not enough to push cells past 3.45v unless secondary reactions are taking place. Just not good - time is the killer here. Things start to get bloaty in prismatic land.

I think that's all I'm really trying to point out - is that with too little current, you won't be touching any of your bms functions anyway. With solar, of course we can't help it at times once in operation.

It's hard to express for me without using bad analogies. It's much like expecting your discharged 200ah agm bank to charge well when you put a 2.6A maintainer on it. Also not recommended, but some do it because eventually it finishes the job - provided there is no safety timer! - but put the hurt on the battery nonetheless.

 

[FilterGuy]

You'll notice that Will uses much beefier supply to do his balance charging,

Actually, in his video on top balancing he uses a cheap 10A supply.




Also, on his recommended tools page he is now recommending this 10A supply.


Except for the color, it looks identical to the one used in the Top Balance Resource I put out.

 

[FilterGuy]

My concern are guys paralleling 8, 100ah or more cells right from the box together for that first run. Ie, on your bench in parallel is a nominal 3.2v 800ah battery. Getting charged to do a parallel balance at only 10a from a low SOC. I'm seeing it here in these forums and elsewhere.

Yup. 10A at 3.65V is only pumping 36.5W...... That is going to take a long time on a large bank of batteries. That is why the tutorial recommends charging them in series first. 10A at 14.6V is 146W and 10A at 29.2V is 292W.

That's the equivalent of charging at a 0.013C rate. That's not enough to push cells past 3.45v unless secondary reactions are taking place.

Interesting..... I have never heard of (or experienced such a problem). In fact, I am not even sure how that would be true.

Just doing a thought experiment..... If I take a single discharged 100A cell and hook it to a 10A supply set to 3.65V, it will start out taking the full 10A (.1C) but the power supply will probably be current limited so the voltage will not be at the 3.65V that I set it. As the cell charges, the current will go down and the voltage will go up. Eventually the power supply will switch to voltage limit and hold at 3.65V while the current continues to drop lower and lower eventually getting down to .013C and then dropping to zero. So even with a beefier supply, you will end up charging at a very low C rate.... so I am not sure how a very low C rate is a problem.

BTW: Lead acid *does* have issues with a very low C rate... but that is a totally different chemistry.

 

[Just John]

Maybe it's just me, but I don't remember reading anything about low C rates being bad for LiFePO4.
It's certainly faster if you spend $200 and get a 60 or 40 amp supply. I did. Also makes a great pack charger.

Edit to add, never did have any problem other than time to get to 3.65v. My experience says the claim it will stall and cause bad things to happen at 3.45v does not reveal that. No bloated cells either.

148.98US $ 30% OFF|Bench Source Laboratory Dc Stabilized Power Supply Adjustable Variable Digital Regulated Power Supply 30v 20a 60v 10a Wanptek - Switching Power Supply - AliExpress

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

 

[Substrate]

Its not the c-rate that is bad by itself - LFP has great charge-acceptance. It's the time spent at the top end, and what the top-end is when taking so much time is the reason for it.

You can read the spec if you find your cell manufacturer. An LFP's "standard charge" is usually spec'ed at 0.2C, and TIME is the reason. They are assuming you are following that minimum specification, and outside of that, you are on your own.

 

[FilterGuy]

Maybe it's just me, but I don't remember reading anything about low C rates being bad for LiFePO4.
It's certainly faster if you spend $200 and get a 60 or 40 amp supply. I did. Also makes a great pack charger.

Edit to add, never did have any problem other than time to get to 3.65v. My experience says the claim it will stall and cause bad things to happen at 3.45v does not reveal that. No bloated cells either.

148.98US $ 30% OFF|Bench Source Laboratory Dc Stabilized Power Supply Adjustable Variable Digital Regulated Power Supply 30v 20a 60v 10a Wanptek - Switching Power Supply - AliExpress

Smarter Shopping, Better Living! Aliexpress.com

www.aliexpress.com

I am curious... Does the PS you linked to hold the constant voltage well? I find the low cost ones drift by around .02V

 

[Just John]

I am curious... Does the PS you linked to hold the constant voltage well? I find the low cost ones drift by around .02V

So far, seems to hold what I set it at to within .005. I'm basing this off the Tekpower I bought, these seem to be manufactured by the same company, just different nameplate. Some of the teardown video reviews, it's the same basic desigm, this is however a newer version that I haven't purchased yet.

My 18 amp Ridens are much more accurate, but of course they are programmable. Off by + or - .002 and very consistent.

 

[jwelter99]

So no worries. But had you come to me after buying that bank, I would have said "Did you also budget for a much larger charger supply to do this job that is at least capable of 0.2C at a minimum?

Long way of saying that you're good with this one-off event. But if you build another one and I catch you using a supply that is not at least 0.2C capable, well.....

This is silly advice. 4 280Ah cells in parallel is 1120Ah. 0.2C is 224A. Most are using, successfully, 10A bench top supplies to balance. That is 0.01C - or 1/20th what you are recommending.

Giving this sort of advice is really bad for the community.

 

[Just John]

Its not the c-rate that is bad by itself - LFP has great charge-acceptance. It's the time spent at the top end, and what the top-end is when taking so much time is the reason for it.

You can read the spec if you find your cell manufacturer. An LFP's "standard charge" is usually spec'ed at 0.2C, and TIME is the reason. They are assuming you are following that minimum specification, and outside of that, you are on your own.

I'll have to process some more data, but at 40 amps charge current it takes approximately 370 seconds from 3.45v before the voltage hits 3.644v and current starts tapering. So about 4 amp hours. That's the thing about LiFePO4, there really isn't any capacity above 3.4 or below 3.1 (that's the knees everyone talks about).

Once it starts tapering current (again at less than 40 amps now) it takes about 60 seconds before it reaches 14 amps input.

So, even with only a 10 amp input, you are not going to spend a lot of time above 3.45 like you are talking about (30 minutes if doing one cell, 2 hours if 4 cells, etc). Based on my data, at 40 amps, about 7 minutes. Certainly if you do a large number in parallel, it gets longer, but unless you are doing more than 16, I wouldn't worry about it, even with a 10 amp supply. If you were going to let them sit for a week or two, I'm sure that would not be good. But I'm looking at actual data.

So, you might hold them at a high state of charge (as defined by you) for 8 hours if doing 8 to 16 cells.