Safe way to fully discharge batteries?

Also, how critical is it that the batteries be relatively balanced when installed? Will the onboard charger sort things out after a few charge/discharge cycles, assuming the batteries are all ok? I am using the more recent Vectrix firmware.

Hi Matt

it is very critical that the batteries are balanced.

If they aren't, more will bulge and fail on the first recharge.

I recommend using a west mountain radio CBA to discharge every 8 or 9 cell block.

Any cell that can not do 25Ah or more should be replaced.

It is best to install all the batteries completely discharged, as you know they are all balanced, and they won't drift by themselves while at 0%.

Matt

I am attempting to replace some bad Ni cells in my battery pack with some 'good' used cells.

Hi Matt, that is a good approach to bring new life in a used pack. Do you also install the improved charging software from The Laird?

I have replaced the bad cells (as indicated by swelling or zero voltage), and am at the point where I would like to put things back together. I have heard that it is best to reinstall the batteries fully discharged. They are in their 8 and 9 cell modules, and I am trying to bring them all to a relatively common state of discharge before reinstalling, but am worried about the effects of too deep a discharge. Any suggestions on how to safely and evenly discharge batteries on a module level? I have been screwing around with RC chargers/dischargers, and with some ~6A/12V trouble lights for the modules, and some 1 ohm power resistors for cell by cell balancing.

It seems that you would have made it easier on yourself discharging the cells (actually the entire pack) before adding cells to strings.
Discharging a balanced string is easy, just connect a 12V car light and wait until it turns off.
Since your cells are in unknown state, you must take care not to reverse cells too badly while discharging.
The safe method of discharging cells is to add a load (resistor) across each cell and wait until everything is discharged.
Note that you can fashion low resistance loads from cheap steel wire or coat hangers. If you can get 0.1 Ohm power resistors then you should get each cell discharged within a few hours even if they start fully charged. You can also connect cells in parallel (turn every 2nd cells around in its string and jumper all terminals together) then discharge all together.

Also, how critical is it that the batteries be relatively balanced when installed? Will the onboard charger sort things out after a few charge/discharge cycles, assuming the batteries are all ok? I am using the more recent Vectrix firmware.

The charger will sort things out eventually after a couple charges, but the charger has stupid assumptions such as when power is first applied to it, it thinks that the pack is completely empty so it will push 30+Ah into it, even if the pack was already fully charged. I don't need to tell you the (overheating) effect of cooking 30Ah away at 10 Amps 150-ish Volts in cells that are full.
Since the charged does not monitor cells, only blocks of cells, it assumes that all cells are balanced and will happily overcharge one fuller cell or reverse (and destroy) a lower cell while running. That is why you should indeed start with all cells almost completely discharged (barely 120V on the idle pack, 100% discharged is good) You can easily prop up the pack when it does not reach over 100V to restart the charger - either using a "Freddy" which takes no more than a AC cord, fuse, diode and lightbulb; or by giving each string a small boost from a 12V supply.

Matt,
You are doing OK. This is not an exact science,
since some cells may have 30Ah capacity while others are down to 25Ah
so even if you are capable to discharge them to *exactly* the same level,
as soon as you fire up the Vectrix charger it will hit them all with 10A
until it starts tapering at the top of the charge, by which time it will
have over-cooked the 25Ah cell with up to 5Ah in the process.
You simply want to avoid that it will start with fully charged cells and
still forces 30Ah into a full pack...
If you want to be gentle then install The Laird's software before charging
(you still do have the silver color ESD charger? His SW does not work with
the gold color Chinese Runke charger)
Another possibility is to plug the bike in for about 2.5 hours so you know
that approx 25Ah has been put back in, then unplug it and use an
external charger (current limited power supply) to push about 1 Amp into
the pack overnight. No cell will have a problem dissipating 1 Amp
over-charge, so heat production will be minimal and you will push all
cells together at the top end, where you will be every time that you
plug in anyway (except with The Laird's SW which slightly undercharges to
give the pack a better longevity)

If you have discharged below 1.2V then the cells are already more than
50% discharged, I expect that going down to 1.15V is low enough in charge
to consider them empty and you can start the way back up putting the pack
together and connecting the charger and charge it.
I found a discharge curve of 4 series cells here, clearly showing that
4.6V (1.15V per cell) means that the charge level is below 10%:
http://shdesigns.org/batts/battcyc.html

Can anybody recommend a termination voltage per cell at which to stop discharge, hopefully before reversal or damage (at ~ 4 degrees C ambient temp).

If discharging cells individually, 0.4V is plenty low enough

Theres no danger to a deep discharge at the cell level
easiest way is to put a 1 ohm resistor across every cell.

If you do discharge at the pack level, and some cells reverse, as long as the reversal is not more than 20% of the batteries capacity (6Ah in this case) it should be ok.

doing it repetively, or holding a cell reversed will cause increased self discharge
If the cell get too hot, it will also lose capacity

What I usually do is charge an 8 or 9 cell block at 1A for 30 hours
I then discharge 25Ah (~5A for 5 hours),
any block that has any cell that falls below 0.8V after 5 hours gets replaced

don't put 1A continuous into the whole battery without the fans running (135W into 75kg of battery will cause the batt temp to rise high enough that raised self discharge in the hottest cells will cause significant imbalance)

Matt