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

[John Frum]

I'll try to bring the high ones down with a load. In the mean time, I'll be getting the balancer over the weekend. Why does @Will Prowse not care so much for the active balancers? Or did I misunderstand him in his balancing video?

I'm not privy to your thoughts or to Will's.

 

[LivingTheDream]

I'm not privy to your thoughts or to Will's.

I guess what I meant was, are there draw backs to using a balancer?

 

[John Frum]

I guess what I meant was, are there draw backs to using a balancer?

Yes.
They can undo the top balance of your pack.
I read through some of your old posts and it seems you grossly overcharged these cells at some point.
I think an active balancer is probably the way to go.

 

[MisterSandals]

I turned on the charger and within a few minutes, here is where I am already. It's as if I didn't even try to bring these voltages together.

Here you have cells in positions 3,4,8 and possibly 6 that are different (lower capacity). I strongly recommend that you combine them with the 4 or 5 low capacity outliers in your other group.

Just a little effort matching your cells into like 8S sets will reduce what a balancer needs to do. I spent a LOT of time doing this and it works really well.

 

[LivingTheDream]

Yes.
They can undo the top balance of your pack.
I read through some of your old posts and it seems you grossly overcharged these cells at some point.
I think an active balancer is probably the way to go.

Really? I'm not sure when that would've been. I watch them like a hawk. What makes you say that (so I don't do it again)?

 

[LivingTheDream]

Here you have cells in positions 3,4,8 and possibly 6 that are different (lower capacity). I strongly recommend that you combine them with the 4 or 5 low capacity outliers in your other group.

Just a little effort matching your cells into like 8S sets will reduce what a balancer needs to do. I spent a LOT of time doing this and it works really well.

I'll do this tomorrow and see how that evens things out.

 

[MisterSandals]

I'll do this tomorrow and see how that evens things out.

You labelled your cells and noted all that you’ve done and the voltages and the cell position, right?

 

[time2roll]

I'll order some oxguard for the final build out.

Any local hardware store should have it. Get the smallest tube as it will last a lifetime for a hobbyist.

 

[LivingTheDream]

You labelled your cells and noted all that you’ve done and the voltages and the cell position, right?

Yes, on this set. I didn't do a very good job on the first set though, but I'll hook them back up and make note of the results in order to match them up.

 

[DerpsyDoodler]

Ok, I just sanded the terminals and bus bars. Hooked everything back up and I'm still out of whack on three of the cells. I'll work on bringing those down a bit and see what happens. I'll order some oxguard for the final build out. I'm not sure if the lugs are tinned, just what came from Doshan.

after you sanded them, did you clean them off with isopropyl alcohol?

 

[LivingTheDream]

after you sanded them, did you clean them off with isopropyl alcohol?

Yes sir

 

[John Frum]

Really? I'm not sure when that would've been. I watch them like a hawk. What makes you say that (so I don't do it again)?

My mistake got you confused with someone else.

 

[LivingTheDream]

Hello everyone,

Thank you all for your insight and help! I decided to go ahead and charge the pack in series until one of the cells hit 3.650 and then switched over to parallel. I took them up to 3.650 and then cut the power until they dropped down to 3.550 or so. Then I reapplied power to 3.600 and just let it trickle charge from there for about three hours. Took off the PS and let them sit overnight. Now I just disconnected the whole set and took the readings in the photo below. So far so good with a 0.013 differential as noted in green. These cells were I-P. You can also see that cells A-H are still sitting pretty at 0.048 differential which is only slightly higher than the 0.039 from three days ago.

 

[LivingTheDream]

Here you have cells in positions 3,4,8 and possibly 6 that are different (lower capacity). I strongly recommend that you combine them with the 4 or 5 low capacity outliers in your other group.

Just a little effort matching your cells into like 8S sets will reduce what a balancer needs to do. I spent a LOT of time doing this and it works really well.

@MisterSandals do you you think it's worthwhile trying to match these cells? I've been watching them over the last three days taking readings in the morning and evening. They don't seem to be drifting too much. However, there is one cell in each group that is lower that the rest. I've changed the colors to the four highest of each set.

 

[RCinFLA]

Having difference in cell voltages as long as they are all above 3.45v, at more than several hours after absorb charging has completed, does not mean cells are misbalanced.

It more likely means cells have slightly different surface charge dissipation rate. Surface charge voltage bump at end of absorb charge has very insignificant amount (<0.01%) of cell capacity. The rate cells bleed surface charge does not relate to their state of charge.

An LFP cell is fully charged if its no-load/charge rested (>5 mins) OCV terminal voltage is greater than 3.45v.

 

[time2roll]

I would not spend too much time matching unless individual capacity tests are done and this is an oddly critical system.
Time to assemble and put in service.

 

[MisterSandals]

@MisterSandals do you you think it's worthwhile trying to match these cells?

Let me back up and reiterate that we want to match based on capacity; putting the lowest capacity cells in one battery and the highest capacity cells in the other.
Other than capacity testing all of your cells individually (which i have never done nor do i have a capacity tester), the next best is to watch which cells run away at charge and discharge extremes.

Batch #1:
Earlier I mentioned "positions 3,4,8 and possibly 6 that are different (lower capacity)."
What cell labels are those now? Those are the lower capacity ones in your first batch.

I have been working on a theory as others have also suggested, that "the lower capacity cells self-discharge faster". Is this reflected in your table above? (i cannot tell where cells 3,4,8 and 6 fall into your new labels). Are they cells C,D,F and H?
From your table it looks like B,C,F and H are the fasted discharge cells.

Please confirm labels and how they relate to original numbers.


Batch #2
Lets try to match those with the lower capacity cells in your second batch (presumable I - P labeled above).
The only data we have are the surface charge dissipations over the periods in your table, correct?

 

[LivingTheDream]

Let me back up and reiterate that we want to match based on capacity; putting the lowest capacity cells in one battery and the highest capacity cells in the other.
Other than capacity testing all of your cells individually (which i have never done nor do i have a capacity tester), the next best is to watch which cells run away at charge and discharge extremes.

Batch #1:
Earlier I mentioned "positions 3,4,8 and possibly 6 that are different (lower capacity)."
What cell labels are those now? Those are the lower capacity ones in your first batch.

I have been working on a theory as others have also suggested, that "the lower capacity cells self-discharge faster". Is this reflected in your table above? (i cannot tell where cells 3,4,8 and 6 fall into your new labels). Are they cells C,D,F and H?
From your table it looks like B,C,F and H are the fasted discharge cells.

Please confirm labels and how they relate to original numbers.


Batch #2
Lets try to match those with the lower capacity cells in your second batch (presumable I - P labeled above).
The only data we have are the surface charge dissipations over the periods in your table, correct?

The cells that you were eluding to (3,4,8,6) are from the second batch. So they are now labeled K,L,N,P.

 

[MisterSandals]

The cells that you were eluding to (3,4,8,6) are from the second batch. So they are now labeled K,L,N,P.

That’s a couple more data points. If these were the low capacity cells, they did not show faster surface charge loss. (Dang for my theory!)

I would keep these 4 together and migrate any other low capacity cells to this group (and high capacity away) as you discover them.

What is your plan?

 

[LivingTheDream]

That’s a couple more data points. If these were the low capacity cells, they did not show faster surface charge loss. (Dang for my theory!)

I would keep these 4 together and migrate any other low capacity cells to this group (and high capacity away) as you discover them.

What is your plan?

Well, in analyzing the discharge of both sets, cell D and cell M are losing it the fastest. The margins seem so low, but I'm not an expert on this. So I'm taking three cells from batch one, and five cells from the second batch to make a "theoretically" matched lower capacity batch.