balancing strategies as they relate to BMS

[Gazoo]

Over voltage protection is not necessarily the same as charge termination logic.
What @Gazoo is describing sounds like tail current termination which would be great.
A tail current of 100 mA is low but fine for our purpose.

Thanks. That's what I was trying to explain. The low tail current termination is a complaint some have with the Riden when used as a charger. I bought it for it's charging feature and besides the low tail current it worked perfectly charging my 12 volt batteries. It has a ton of other features I may never use. I do like the OCP and OVP. So there is some redundancy built in. I did test the capacity of my batteries after charging and they tested the same. No surprise as far as that goes.

 

[Gazoo]

I don't know why the Riden isn't called a battery charger. Where the wires are connected it says for charging batteries.
I'm sure there is some technical reason not to call it a charger. But the word charger pretty much explains what I am doing in one word. Power supply is two words and it doesn't explain what I am doing with it.

The reason is it's main function is a power supply. But it can be used as both a power supply and a charger. And yes it can be confusing.

Something else I realized is that 48V at 12A isn't really all that wimpy. Riden 48V x 12A = 576W Mean Well 3.3V x 120A = 396W
The Mean Well is more powerful for parallel configuration charging. Or charging single cells. But it is faster for me to charge 16 cells in series with the Riden than it is to charge them in parallel with the Mean Well. And I get the BMS when use the Riden.

The wattage is actually much higher. Charging a 48 volt pack requires a charging voltage of 58.4 volts to fully charge the pack. So the power needed from the supply is 700 watts as I posted above.

I agree with challenges comes risk. And I also like challenges. But there is no need to take risks when you already have everything you need to negate the risks.

 

[GXMnow]

I may have missed something, but I was trying to follow the top balancing logic and it put a question in my head.

Everything has some resistance. In this case, it is the electro chemical resistance inside of the cells that I am thinking about. I have not worked much with LFP yet, most of my experience is RC car LiPo, Samsung 18650's and my LG Chem NMC cells. All of these do have very low internal resistance, but it is still there. If I am fast charging at a relatively high current, and then just cut it off, the cell voltage drops quite a bit to the real internal cel voltage. The absorption or saturation charge time where the voltage is held for a period of time, allows for the internal cell chemical voltage to "catch up" with the terminal voltage. I do not see any benefit in cutting off the charge and not doing at least some saturation charge at your desired voltage. So my question is... Is there a reason some here are just doing a quick shut off through a power supply or BMS and not doing an absorb phase? Will's parallel balancing is all about a long top absorb charge to get the cells internal chemical voltage to all match. From what I understand about how LFP cells react, I would think holding all of the cells at the desired top balance voltage until the current falls near zero would be the best option. Then when they are separated and allowed to rest, they will all drop back some, and should still be very close. If one fall more, it probably needs to be held at the set voltage even longer to catch up. In the end, the resting voltage should all be the same if the cells are at the same state of charge, under no load, and at the same case temperature. Am I missing anything?

When I built my 14S e-bike batteries, I had to manually balance the cells as I did not have a balancing charger for 14 cells or a BMS yet. I used my RC car charger that can do 1S to 6S and decided to split the pack into 3 sections, two 5S and one 4S. I balance charged each of the sections until they all sat in absorb for at least 15 minutes and the current dropped to just 0.02 amp. After the pack sat, I checked and all 14 cells were balanced within 3 mv on my Fluke meter. I had to do it one section at a time since the charger outputs are not isolated. Trying to charge a cell in the middle of a wired pack requires that the charging source be fully isolated, or you end up smoking things. It is bad enough working with 5 AH cells. Having a mishap with 100+ AH cells is a whole different story. I had one little slip while checking cell voltages on my 360 AH pack. My clip lead to my meter fell off, and on it's way down just taped one of the buss bars. It burned 1/4 inch off the end of the alligator clip and turned it into flying molten metal bits. Please be careful when working with all this power.

 

[Gazoo]

Am I missing anything?

I don't think so if I am understanding your post. Will is using a traditional power supply. The Riden power supply has a charging feature. So it can be used as a power supply or as a charger.

Parallel top balancing the cells using it as a power supply or using it as a charger will achieve the same result. Same if charging a pack. The benefit of using the Riden's charging feature is there should be no chance of overcharging the cell since it will terminate the charge at the selected voltage. The set current at charge termination is 100ma. There is no way to change the set current but I think at that point the cells would be well saturated.

Having said that I have tested it using it's charging function with a 14.6 volt battery twice. Both times it worked perfectly. I don't know how it's going to behave when I get around to top balancing my 8 280ah cells. I suspect it will be ok and cut off like it's supposed to. I know it will take a long time I am ok with that and I will be keeping an eye on it. I will also make sure to check the voltages of the cells after they have reached their resting voltages. Going to be interesting.

 

[ArthurEld]

I may have missed something, but I was trying to follow the top balancing logic and it put a question in my head.

Everything has some resistance. In this case, it is the electro chemical resistance inside of the cells that I am thinking about. I have not worked much with LFP yet, most of my experience is RC car LiPo, Samsung 18650's and my LG Chem NMC cells. All of these do have very low internal resistance, but it is still there. If I am fast charging at a relatively high current, and then just cut it off, the cell voltage drops quite a bit to the real internal cel voltage. The absorption or saturation charge time where the voltage is held for a period of time, allows for the internal cell chemical voltage to "catch up" with the terminal voltage. I do not see any benefit in cutting off the charge and not doing at least some saturation charge at your desired voltage. So my question is... Is there a reason some here are just doing a quick shut off through a power supply or BMS and not doing an absorb phase? Will's parallel balancing is all about a long top absorb charge to get the cells internal chemical voltage to all match. From what I understand about how LFP cells react, I would think holding all of the cells at the desired top balance voltage until the current falls near zero would be the best option. Then when they are separated and allowed to rest, they will all drop back some, and should still be very close. If one fall more, it probably needs to be held at the set voltage even longer to catch up. In the end, the resting voltage should all be the same if the cells are at the same state of charge, under no load, and at the same case temperature. Am I missing anything?

When I built my 14S e-bike batteries, I had to manually balance the cells as I did not have a balancing charger for 14 cells or a BMS yet. I used my RC car charger that can do 1S to 6S and decided to split the pack into 3 sections, two 5S and one 4S. I balance charged each of the sections until they all sat in absorb for at least 15 minutes and the current dropped to just 0.02 amp. After the pack sat, I checked and all 14 cells were balanced within 3 mv on my Fluke meter. I had to do it one section at a time since the charger outputs are not isolated. Trying to charge a cell in the middle of a wired pack requires that the charging source be fully isolated, or you end up smoking things. It is bad enough working with 5 AH cells. Having a mishap with 100+ AH cells is a whole different story. I had one little slip while checking cell voltages on my 360 AH pack. My clip lead to my meter fell off, and on it's way down just taped one of the buss bars. It burned 1/4 inch off the end of the alligator clip and turned it into flying molten metal bits. Please be careful when working with all this power.

Thanks for the input GXMnow. Your experience is often very interesting to me. The subject of long top absorb charge is very confusing. I know there is a big difference between just cutting off at top voltage and doing the long absorb. So many bad things happen when at the top of the knee so I want to do it right.
One thing that I am doing is a capacity test on each cell by discharging them with a capacity tester. And I don't know how the long absorb affects the capacity. I have to assume a long absorb will add to capacity. Then there is the subject of parallel top balancing. If I charge each cell separately I am going to get different capacity results than if I charge them parallel. I don't know if it is enough to matter. I will probably find out someday.
Thanks for the warning about being careful. I still haven't made a single spark. But I'm sure you saw I burned up a meter. It got very hot and the screen got damaged. That is the only reason I knew something was wrong. I saw the screen mess up and I touched it and it was hot as hell. I disconnected the cable that I knew caused the problem but I still don't understand what was wrong. Now you talk about just dropping a lead in the wrong place causing terrible results. For some reason I fear is that a small short will happen that I don't notice and cause a fire later when I am not near the batteries. I don't know why that idea keeps haunting me. I'm sure there are other more dangerous things that can happen.
I did buy a couple fire extinguishers per smoothJoey's suggestion.

 

[ArthurEld]

I agree with challenges comes risk. And I also like challenges. But there is no need to take risks when you already have everything you need to negate the risks.

Yes Gazoo you are correct. I have enough challenge following the known best path. I don't need to be blazing new trails.
I still have it in the back of my mind to use the 17th BMS lead to monitor charging a single cell. But I might not ever try it.

 

[GXMnow]

DON'T! The balance leads need to be in order, one cell voltage apart. Get them out of order and you may let out some of the magic smoke that makes it work. Having any out of order will mean a negative reverse polarity voltage. If you want to monitor, charge, or load a single cell in the pack, that device must be completely isolated from the main pack.

 

[ArthurEld]

DON'T! The balance leads need to be in order, one cell voltage apart. Get them out of order and you may let out some of the magic smoke that makes it work. Having any out of order will mean a negative reverse polarity voltage. If you want to monitor, charge, or load a single cell in the pack, that device must be completely isolated from the main pack.

Thanks again GXMnow. I won't try it. I have enough to learn sticking to the basics.

 

[Gazoo]

Yesterday I started my parallel top balancing and I set the riden to charge my cells at 12 amps and 3.4 volts. But I was only getting about 5.8 amps to the cells. The cells were reading 3.3 volts shortly after I started the charge. I let it run over night and when I checked today the cells were reading 3.311....lol. And the current had started to drop.

So I paused the charging and set the voltage to 3.5 volts. Now its charging at just shy of 11 amps. But the nice thing is when I paused it, it didn't lose the amp/wh's displayed going into the cells. It picked up where it left off. So as of now I have put 122ah's or 417wh's into the cells. I know it's still going to take awhile. When the current starts to drop again I will bump it up to 3.6.

Also I taped the temp sensor to the mid top of the power supply. The display shows 87 degrees and it's 72 in here. I think all in all that's good.

 

[ArthurEld]

Yesterday I started my parallel top balancing and I set the riden to charge my cells at 12 amps and 3.4 volts. But I was only getting about 5.8 amps to the cells. The cells were reading 3.3 volts shortly after I started the charge. I let it run over night and when I checked today the cells were reading 3.311....lol. And the current had started to drop.

So I paused the charging and set the voltage to 3.5 volts. Now its charging at just shy of 11 amps. But the nice thing is when I paused it, it didn't lose the amp/wh's displayed going into the cells. It picked up where it left off. So as of now I have put 122ah's or 417wh's into the cells. I know it's still going to take awhile. When the current starts to drop again I will bump it up to 3.6.

Also I taped the temp sensor to the mid top of the power supply. The display shows 87 degrees and it's 72 in here. I think all in all that's good.

Sounds good Gazoo. I put my temp sensor in the same place as you. One of the videos showed the guy wad it up and stuff it in the Riden.
I'm spending my time capacity testing which is going to be a long dragged out side project.

So far it seems like you'd have a tough time damaging a cell with the Riden set to 3.65V using the battery connections. I haven't done anything with the other connections. It might be best to keep the current high all the way to cell 3.65V then do the absorb. I've got lots of experimenting to do.

It will be a while before I am ready to get back to balancing. Good thing I'm not in a rush.

edit: added cell 3.65V. Meaning I want to charge at 12A until the cell reaches 3.65V. Then do the absorb.
I think when you charge larger packs at 12A you can just charge until the pack reaches 3.65V. Then no absorb because the amps are spread so thin it's already like absorb.

 

[Gazoo]

So far it seems like you'd have a tough time damaging a cell with the Riden set to 3.65V using the battery connections. I haven't done anything with the other connections. It might be best to keep the current high all the way to cell 3.65V then do the absorb. I've got lots of experimenting to do.

Yup...I agree. I want to see what the difference is on the display vs the battery when the current starts to drop. Just curious. It's still holding shy of 11 amps and the voltage at the cell is 3.318. Still a long ways to go. When the current goes below 10 amps I will increase the voltage to 3.65.

I think when you charge larger packs at 12A you can just charge until the pack reaches 3.65V. Then no absorb because the amps are spread so thin it's already like absorb.

Well if using as a power supply then yes. If using as a charger it's different but one could increase the voltage to say 3.8 volts and then when the charger goes into CV mode the cells might be around 3.65 volts. But I am curious about that and will be watching the tail end of the charge.

I hope I catch it when it switches from CC to CV mode with the voltage set at 3.65. The difference between the displays voltage and the voltage at the terminals when it switches will be interesting since it is an 8 cell pack. And charging at a low voltage.

I should have set it up to use the data logging....lol. But I got impatient and don't want to disturb it.

 

[ArthurEld]

Yup...I agree. I want to see what the difference is on the display vs the battery when the current starts to drop. Just curious. It's still holding shy of 11 amps and the voltage at the cell is 3.318. Still a long ways to go. When the current goes below 10 amps I will increase the voltage to 3.65.


Well if using as a power supply then yes. If using as a charger it's different but one could increase the voltage to say 3.8 volts and then when the charger goes into CV mode the cells might be around 3.65 volts. But I am curious about that and will be watching the tail end of the charge.

I hope I catch it when it switches from CC to CV mode with the voltage set at 3.65. The difference between the displays voltage and the voltage at the terminals when it switches will be interesting since it is an 8 cell pack. And charging at a low voltage.

I should have set it up to use the data logging....lol. But I got impatient and don't want to disturb it.

I have the Riden voltage set to 3.8
It's in CC mode and the display is at 3.65
The cell is 3.35V
Like you said, I keep changing the voltage when it gets to CV. And I keep a close eye on things.
I have OVP set to 3.7. I will keep adjusting that too.

 

[John Frum]

q: why do cells come partially discharged?
a: {
    I think...
        That the batteries are capacity tested during manufacturer as the first criteria for binning
        That woud mean the cells are fully charged and then fully discharged
        Then the manufacturer partially charges them
            The batteries need to be parially charged so that they have a shelf life
            The target SOC will be whatever the airlines will accept minus a margin for safety
                Not the airline's saftey, the manufacturer's safety
                    from getting their cells rejected on the ramp
                    or worse
                    having their insurance premiums increase
            Because mistakes have consequences this whole process is probably automated and therfore consistent
        These regulations probably were enacted based on airlines having some fires due to explody flavours of lithium ion
}
Consdidering the above {
        I think most people should
            1. examine the cells for damage
            2. give each cell a serial number
            3. note the voltages of each cell
            4. assemble into a serial pack with bms
            5. charge the pack until full or one cell trips the high cell cutoff at 3.7 volts
                If the pack charges to full then
                    you have a balanced pack and can procede to cycle/capacity test the pack
                else
                    The cell that tripped the high cell cutoff will likely have the lowest capacity
                    top balancing of one form or another is probably required
                    then cycle/capcity test the pack or if necessary cycle test the cells indvidually
            The cycle/capacity test is
                disable balancing
                charge to full(since its a top balanced pack the cells should be very close in voltage)
                let rest 1 hour
                record the resting cell voltages
                discharge to empty
                let rest 1 hour
                record the capacity
                record the cell voltages
                charge to full or high cell cutoff
                let rest 1 hour
                record the cell voltages
                pass criteria is
                    cell voltages should be fairly close to the initial values
                    capacity >= to spec
                if pass
                    confgure for normal usage
}

}

fyi @JoeHam

 

[Colonel.lp]

q: why do cells come partially discharged?
a: {
    I think...
        That the batteries are capacity tested during manufacturer as the first criteria for binning
        That woud mean the cells are fully charged and then fully discharged
        Then the manufacturer partially charges them
            The batteries need to be parially charged so that they have a shelf life
            The target SOC will be whatever the airlines will accept minus a margin for safety
                Not the airline's saftey, the manufacturer's safety
                    from getting their cells rejected on the apron
                    or worse
                    having their insurance premiums increase
            Because mistakes have consequences this whole process is probably automated and therfore consistent
        These regulations probably were enacted based on airlines having some fires due to explody flavours of lithium ion
}
Consdidering the above {
        I think most people should
            1. examine the cells for damage
            2. give each cell a serial number
            3. note the voltages of each cell
            4. assemble into a serial pack with bms
            5. charge the pack until full or one cell trips the high cell cutoff at 3.7 volts
                If the pack charges to full then
                    you have a balanced pack and can procede to cycle/capacity test the pack
                else
                    The cell that tripped the high cell cutoff will likely have the lowest capacity
                    top balancing of one form or another is probably required
                    then cycle/capcity test the pack or if necessary cycle test the cells indvidually
            The cycle/capacity test is
                disable balancing
                charge to full(since its a top balanced pack the cells should be very close in voltage)
                let rest 1 hour
                record the resting cell voltages
                discharge to empty
                let rest 1 hour
                record the capacity
                record the cell voltages
                charge to full or high cell cutoff
                let rest 1 hour
                record the cell voltages
                pass criteria is
                    cell voltages should be fairly close to the initial values
                    capacity >= to spec
                if pass
                    confgure for normal usage
}

}

fyi @JoeHam

Good advice. What's your theory then on why it appears so many are getting cells that are way over 30% or 40% charged. Mine are certainly up between 80 and 90%, and I keep reading of people here that have one rogue cell that hits cutoff way before the others.

 

[John Frum]

Good advice. What's your theory then on why it appears so many are getting cells that are way over 30% or 40% charged. Mine are certainly up between 80 and 90%, and I keep reading of people here that have one rogue cell that hits cutoff way before the others.

Yours is the first case that I am aware of, at least recently.
Can you link to other similar cases?

 

[John Frum]

Good advice. What's your theory then on why it appears so many are getting cells that are way over 30% or 40% charged. Mine are certainly up between 80 and 90%, and I keep reading of people here that have one rogue cell that hits cutoff way before the others.

Did you cells ship from origin or from a regional warehouse?
If they came from a regional warehouse, they might be returns.

 

[Colonel.lp]

Yours is the first case that I am aware of, at least recently.
Can you link to other similar cases?

Didn't Arthur say he had one cell that went straight to cut off before the others, and also someone else, I didn't pay attention to the names.

Did you cells ship from origin or from a regional warehouse?
If they came from a regional warehouse, they might be returns.

They came direct from xuba, certainly looked clean, no marks or scuffs apart from a small dent in one.

 

[John Frum]

Didn't Arthur say he had one cell that went straight to cut off before the others, and also someone else, I didn't pay attention to the names.

I recall he had a runner but I don't think the cells arrived at a high state of charge.
Pretty sure he charged them for many hours as a serial pack before the bms tripped.

They came direct from xuba, certainly looked clean, no marks or scuffs apart from a small dent in one.

I have a vague recollection that some manufacturer has a warehouse somewhere in Europe.
You order from the manufacturer and it takes 30+ days for the cells to arrive but if you dig into the tracking they come from Europe. I could be totally out to lunch on this.
My memory is not what it used to be.

 

[ArthurEld]

Didn't Arthur say he had one cell that went straight to cut off before the others, and also someone else, I didn't pay attention to the names.

I did have one cell take off from the others (16s) and hit the BMS OVP. In another day or two I should be able to determine if that cell is different than the others. My cells were all 3.3V when I received them from Deligreen - China.

 

[Colonel.lp]

I recall he had a runner but I don't think the cells arrived at a high state of charge.
Pretty sure he charged them for many hours as a serial pack before the bms tripped.

I have a vague recollection that some manufacturer has a warehouse somewhere in Europe.
You order from the manufacturer and it takes 30+ days for the cells to arrive but if you dig into the tracking they come from Europe. I could be totally out to lunch on this.
My memory is not what it used to be.

Plenty have orded from xuba with good experience, I have no way of knowing if they have a warehouse in Europe, but I've not heard it being said.