Top Balancing...Do I need to go to 3.650V?

[MisterSandals]

Then all of a sudden one spikes to 3.650 with a .149 differential.

Can you describe how you arranged the cells and where this runaway cell was positioned? Did you create batteries based on their settle voltage? What was the battery voltage when one cell began to significantly diverge?

EDIT: Wait wut? How high are you trying to charge your battery, to the theoretical limit before they come apart?

 

[LivingTheDream]

Can you describe how you arranged the cells and where this runaway cell was positioned? Did you create batteries based on their settle voltage? What was the battery voltage when one cell began to significantly diverge?

EDIT: Wait wut? How high are you trying to charge your battery, to the theoretical limit before they come apart?

I'm charging them in series with 8 cells in 8S.

- With the first batch, cell six was the runaway and tripped the BMS. I put them in parallel and then took them up to 3.615 before taking them off to settle.
- With the second batch, it is cell three that's the runaway. I am getting ready to put them in parallel to repeat the process of the first batch.
- I will then match the cells based on your recommendation above in post #16.

I was just curious if this runaway is normal and if I'm doing this correctly. Definitely gun shy. LOL!

 

[John Frum]

I'm charging them in series with 8 cells in 8S.

- With the first batch, cell six was the runaway and tripped the BMS. I put them in parallel and then took them up to 3.615 before taking them off to settle.
- With the second batch, it is cell three that's the runaway. I am getting ready to put them in parallel to repeat the process of the first batch.
- I will then match the cells based on your recommendation above in post #16.

I was just curious if this runaway is normal and if I'm doing this correctly. Definitely gun shy. LOL!

Just to be clear you should top balance to 3.65 volts but your your normal charge profile should be lower lower than 3.65 volts per cell.
Anything over 3.45 volts per cell will get them full, it just takes longer.
I suggest splitting the difference to 3.55 per.
That should stop your BMS from tripping on cell over-voltage and allow your bms to maintain top-balance.
It my opinion that you don't have to break the pack or do any curation.
Unless you are going to capacity test the cells its pointless anyway.

 

[MisterSandals]

With the first batch, cell six was the runaway

With the second batch, it is cell three that's the runaway.

Did you arrange the cells into batteries as i suggested based on settle volts? Or just random cells into 8S groups?

 

[LivingTheDream]

Just to be clear you should top balance to 3.65 volts but your your normal charge profile should be lower lower than 3.65 volts per cell.
Anything over 3.45 volts per cell will get them full, it just takes longer.
I suggest splitting the difference to 3.55 per.
That should stop your BMS from tripping on cell over-voltage and allow your bms to maintain top-balance.
It my opinion that you don't have to break the pack or do any curation.
Unless you are going to capacity test the cells its pointless anyway.

Ok, I had them in parallel at 3.615 and all settled overnight disconnected and are reading in the 3.440 range. This was on the first batch, so the second one I'll stop at 3.550. I understand that this is only for the initial top balance. Once I start hooking everything up, then I'll start diving into the settings that they should run at.

 

[LivingTheDream]

Did you arrange the cells into batteries as i suggested based on settle volts? Or just random cells into 8S groups?

This initial set up and charge were just random cells as they came out of the boxes. I was going to use your suggestion of arrangement once I had both sets top balanced.

 

[John Frum]

Ok, I had them in parallel at 3.615 and all settled overnight disconnected and are reading in the 3.440 range. This was on the first batch, so the second one I'll stop at 3.550. I understand that this is only for the initial top balance. Once I start hooking everything up, then I'll start diving into the settings that they should run at.

You misunderstand me.

Top balance is different than your in service charge profile.
The in service charge profile should be 3.55 volts per cell.
For the provisioning you want to take the pack until the first cell hits 3.65 volts.
That way you can see how far it shot up from the rest of the pack.

 

[LivingTheDream]

Ahhh, ok. So I'm good taking it to the first cell hitting 3.650 and then switching to parallel and taking the pack to 3.615?

 

[time2roll]

Initial charging in series will most likely produce one cell that runs to 3.6 well before the others. The spike does not indicate anything initially except that it was at a slightly higher state of charge from the get go. At the top end there is very little extra charge that creates the spike as the voltage rises exponentially at the end. This is the very reason for the parallel top balance.

 

[John Frum]

Ahhh, ok. So I'm good taking it to the first cell hitting 3.650 and then switching to parallel and taking the pack to 3.615?

If the delta between the high cell and the low cell is .150 when the first cell hits 3.65 I think you should skip putting them in parallel.
Declare it a provisioned battery and let the BMS fine tune the balance going forward.
What BMS are you using?

 

[MisterSandals]

Initial charging in series will most likely produce one cell that runs to 3.6 well before the others.

Make note of this cell and any cells that start to diverge before the rest. These are good candidates for being the lower capacity cells.

Good notes and observations will be key to forming the best possible batteries.

This initial set up and charge were just random cells as they came out of the boxes. I was going to use your suggestion of arrangement once I had both sets top balanced.

Once you start messing with the cells like this diminishes the top balance.

 

[MisterSandals]

Declare it a provisioned battery and let the BMS fine tune the balance going forward.

I think that any effort he makes in grouping the 16 cells into 2 groups of 8 similar cells will be highly beneficial. His initial responses indicate he has significantly varying cells. Now is the time, with a little patience and methodology to match the cells as much as possible.

 

[LivingTheDream]

If the delta between the high cell and the low cell is .150 when the first cell hits 3.65 I think you should skip putting them in parallel.
Declare it a provisioned battery and let the BMS fine tune the balance going forward.
What BMS are you using?

I'm using an Overkill.

 

[John Frum]

I'm using an Overkill.

Good, glad its not a Daly.
At least you have a fighting chance.
If the cells drift too much you may need an active balancer.

 

[John Frum]

I think that any effort he makes in grouping the 16 cells into 2 groups of 8 similar cells will be highly beneficial. His initial responses indicate he has significantly varying cells. Now is the time, with a little patience and methodology to match the cells as much as possible.

Missed or forgot that bit.
If they are poor quality mis-matched cells then they should be individually charged/discharged for capacity.
I guess we are trying to save the OP some time?

 

[John Frum]

I re-read the thread, now I remember that OP is using a generator to do the top balance/provisioning.

 

[John Frum]

@LivingTheDream where did you get these cells?
I think that will give us an idea of what quality they are.

 

[LivingTheDream]

Good, glad its not a Daly.
At least you have a fighting chance.
If the cells drift too much you may need an active balancer.

Good to know! LOL! I did something right.

I'll take the readings off my first set and post them here in a minute.

 

[LivingTheDream]

@LivingTheDream where did you get these cells?
I think that will give us an idea of what quality they are.

I got them from Amy Zheng at their Houston warehouse. I was in TX and had such a bad experience here in the group buy, that I drove to their facility so I could inspect them with my own eyes. They look great, no dents, QR intact, etc.

 

[John Frum]

I got them from Amy Zheng at their Houston warehouse. I was in TX and had such a bad experience here in the group buy, that I drove to their facility so I could inspect them with my own eyes. They look great, no dents, QR intact, etc.

Those cells should be fine.