Acceptable V differences before parallelling 18650's ???

[Matador]

Just got my hand's on 17 used (less than 100 cycles) Makita BL1840 battery packs, each containing 10 x Sony VTC4 18650's (rated 2100mAh@1A... 2035mAh@10A, 30A max discharge, 4.0A (2C) max charge).

I started testing each individual cells for mAh and voltage drift over time.
Makita Packs are from 2014, 2015 and some from 2013. From now on, 7 pack opened and salvage voltage varies from 3.47 to 3.98 volts. All cells in one pack are same voltage up to the second decimal (I use a cheap Mastercrap multimeter, with ±0.01V error margin). From very crude preliminary testing, I assume cells will be in the 30-40 milliohms DCIR range.

I use two intelligent Foxnovo 4S digital Chargers (will discharge cells @ 0.5A/cell and give mAh/cell) to test for discharge capacity and recharge to 4.20V. I can test 4 cells at a time on each device, so I do 8 cells per 10-12 hours (charge, discharge, recharge).
MY GOAL : I want to charge all cells to a full 4.20V, and parallel cells in 10P groups (Goal : 14S10P pack).

I noticed that one of my Foxnovo 4 cell charger systematically brings cells to 4.22-4.23 volts (sometimes 4.21V) charged, while the other Foxnovo 4 cells charger brings cells to 4.20 4.21 volts.

MY QUESTION : What is the maximum tolerable voltage difference acceptable before parralleling cells ?
I am worried some cells (eg those at 4.23V) will massively dump current in other cells (eg those at 4.20V) during paralleling connection, causing harm from exceeding the maximum 4.0A charge rating... (PS : Connections between each cell will be made with 4.6 mm2 cross-sectionnal area pure copper bussbars, allowing for very high currents while making a decent heatsink at the same time).

What do you think : 0.01 V ? 0.03 V ? 0.05 ? 0.10 V ? 0.20V ? We're talking closed to fully charged in the charge curve here.
I mean what is acceptable when cell are fully charged or very closed to it ? 0.01V difference okay ? 0.05 V difference okay ?

Any experience with this ?
Should I buy myself a 3 or 4 decimal voltmeter or my old matercraft still okay ? Should I use my Bestech HCX-D131 BMS (14S, 80A, 168±20 mA-balancing current/cell) to bablance 14 cells at a time before making 10P parrallel stings connections ? That seems like a lot of work....

 

[Inwo]

I've been watching for an answer.
No experience with 18650s at all, but I'll take a shot.
Seem current will only be limited by ir of the two cells.
Say 100 milliOhms?
Only 10 amps/volt, 1 amp/.1 volt, .1 amp/.01 volt etc.
Try it and see.
Or limit current between candidates thru low value resistor or lamp for a while before connecting.
That's what I do on a larger scale using cells with much lower ir.

 

[Matador]

Though about it.
Somone on this forum measured new cell DCIR for VTC4 in the 20 mOhm range (AC impedance 12 mOhms according to manufacture specs).

So for my 10P pack if I have 9 cells parallel and about to connect the 10th cell with a 0.03V difference, it would be too much :

9 cells in parrallel means 0.00222 ohms (DC). (Using 1/Rtot = 1/R1 +1/R2 + ... + 1/Rn to calculate R value of the 9 cells string)
deltaV = 0.03 Volts...

According to ohms law.... deltaV = R x I ... so I = deltaV / R
meaning I = 0.03 volts/0.00222 ohms = 13.5 AMPs

Yes the battery is rated for 4.0A max charge current ...

So according to my calculations, for this particular cell, the maximum tolerable voltage difference for a max of 4.0A "charge" by current dumping is 0.0089 Volts when going from a 9P 4.23V string and connecting a 10th 4.20V cell to make a 10P string. That is deltaV = 0.00222 Ohms x 4.0Amp = 0.0089 Volts.

Does that sound right to the pros on ES ?

The thing is... I'm not sure if the resistance of the 10th cell to be connected should be taken into consideration in the equation or not... I would'nt know how to solve it anyways...

 

[Matador]

I've been watching for an answer.
No experience with 18650s at all, but I'll take a shot.
Seem current will only be limited by ir of the two cells.
Say 100 milliOhms?
Only 10 amps/volt, 1 amp/.1 volt, .1 amp/.01 volt etc.
Try it and see.
Or limit current between candidates thru low value resistor or lamp for a while before connecting.
That's what I do on a larger scale using cells with much lower ir.

Thanks for the input Inwo.
I like that idea about using low value resistors.
Or maybe I can first connect parallel strigs with very very low gauge wire, like 36 AWG (if I can find some) to bottleneck the current, let balance then connect busbar

 

[Inwo]

Good idea, or a td fuse. They have low resistance plus a safety factor of blowing.

 

[Inwo]

I would consider the pack as zero ohms and calculate using the est. res. of added the cell.