The "Brave Combo" Lithium Pack

[xyster]

I'm probably going to stick with SLA's for my more occasional-use scoot too. If used about 3x per week or more, the pure economics of batteries tilt in favor of other chemistries. As most folks here know, lead has good calendar life, but poor cycle life.

 

[safe]

How do you plan to separate the two strings for charging?

There would be more than two strings. With 120 batteries you have 4 sets of 3 tubes that have 10 cells per tube:

4 sets X 3 tubes per set X 10 cells per tube = 120 cells.

And since there is NO SOLDERING I can every once and a while give the tubes their "routine checkup" and pull all the cells out and test each one individually.

Would it be easier to identify the "weak cell" before or after charging?

I would think before charging, but it's almost as though for the "checkup" I might need to run a separate single cell charger on each cell to see how much capacity each one really can hold. Then from that I would identify the "weak link".

Ideas?

I might begin with the "overall" tube voltage and if it's "fine" then I can skip it and only do the "checkup" on ones that are showing problems. I was figuring to use a charger for each tube, or maybe for every two tubes (24V)... not sure yet.

Just thought I'd repost this so you might see it.

As for chargers I'd need 6.... for each set of 4 I'd need 2 then times 3.

 

[xyster]

There would be more than two strings. With 120 batteries you have 4 sets of 3 tubes that have 10 cells per tube:

By 'strings' I meant cells wired in series. You're planning a 2p60s pack, right? If yes, then how are you going to separate for charging the two strings of 60 wired together in parallel?

 

[safe]

You're planning a 2p60s pack, right?

No. It's a 48 Volt pack. So there are tubes that equal 12 Volts each combined together 4 at a time. Then there are 3 sets of those.

So I could break the 48 Volts into two sets of 24 Volts. (Deans Connectors for each pair of tubes, 24 Volt, makes it really easy, just unplug the wires and hook them into the charger)

In the end the math gives me a need for 6 chargers @ 24 Volts.... something like $80.

 

[xyster]

Ok, so it's going to be 3p40s, the 40-cell strings each broken into 4, 10 cell tubes.

Unless you don't mind disconnecting and reconnecting every time you charge each wire that joins any two tubes in parallel, you're going to have to install diodes. In general, nickel batteries can not safely charge in parallel like lithium and lead can.

Later in the afternoon, I'll sketch a schematic of how I'd wire it.

Would it be easier to identify the "weak cell" before or after charging?

During discharge, the weakest cell will have the least capacity, and/or the most voltage droop. Seems to me to find the weaklings, you'd have to individually attach a load like a resistor, than measure the voltage droop, and also measure how long it can sustain the load before resting voltage without the load drops to ~1.0 volts. The usable range for nickel batteries is 1.0 to 1.4 volts.

 

[Matt Gruber]

safe
please refresh my memory, what are your goals here? is this a new bike?

 

[xyster]

Ideas?

Ok, below's my doodle of how I think I would do your 3p40s NiMH pack if I was going to forego diodes. With this configuration, it should be OK to connect all 12, 12 volt chargers simultaneously (so long as the chargers are all isolated -- i.e. two prongs not three). Plus, there's only one switch to flip, and nothing to disconnect.

I recall other people with NiMH packs writing that more than 10 unmatched cells per charger led to intolerable imbalance over time, so I used 10 cell subpacks for this schematic, but it should work the same with 20 cell subpacks for 24 volt chargers.

Note: the big fat black power wire isn't connected to the ports on the right, like it looks from my sketch it might.

 

[safe]

Unless you don't mind disconnecting and reconnecting every time you charge...

During discharge, the weakest cell will have the least capacity, and/or the most voltage droop.

Come on, 6 Deans Connectors (one for each charger) is not too much to do when you recharge. Right now on my current bike I use three Deans Connectors to charge or to run... so it would be mosty the same idea. In the end the only difference is the time it takes to "reconnect" the 6 connectors after charging that would be any different. (verses plugging and unplugging into a port) I'll take the "Keep It Simple Stupid" route. I like the idea of having the subpacks "physically" disconnected from everything else so that it's a "no brainer" circuit. You literally CAN'T screw up something that simple. (and the Deans Connectors even prevent you from plugging them in wrong)

As for testing the cells, it would be best when they are empty? Or full?

 

[safe]

please refresh my memory, what are your goals here? is this a new bike?

Two bikes actually. I'm going to build two of them in parallel and that way the fiberglass molds can be used for both. My current motors are a MY1020Z3 (the one with the geardown) which I was going to "overvolt" to 48 Volts. Then the other bike would use a 1200 Watt motor. I'm going to build these bikes so that the motor mounts are themselves bolted onto the frame so if I want to play with other motors in the future I can just create the mounts for them and then just bolt them on to test them out. That Transmagnetics motor is VERY TEMPTING, though at this point I'm not willing to cough up the extra $350 for the motor and controller. So the idea is to begin a "platfiorm" for a road racer and be able to test theories on one, while being able to ride another. Back and forth testing and development.

The battery "tubes" will be removeable and I'll just use one set of "tubes" at first for the two bikes. (then I can always buy more later)

 

[Leeps]

I would stick with xyster on the charging.
As far as checking the cells. your checking for capacity, this basically means charging the cell and then putting a constant current load on it and measuring the time it takes to empty the cell. Beyond that differences in cell impedance can get you into trouble, if one series string had a lower impedance than the adjacent one the load will be unbalanced and one will drain faster than the other, adding ten milliohms (depending on your expected current drain of course) in series with each pack will tend to force them into balance, the best way is to even out the strong and weak cells across the series strings, to check for this your looking at how much voltage drop you have on a cell when you place a load, you dont want all the strong cells in one series string. Something else that can get you into trouble is differences in self discharge rate, the only thing that you could possibly do to correct for that is to charge before every ride if there is a random oddball cell in the mix, best thing would be to throw that cell away and replace it with one that behaves better.
Joe

 

[xyster]

Come on, 6 Deans Connectors (one for each charger) is not too much to do when you recharge.

I initially used Dean's on my bike, it kept vibrating apart so I took it out.

As for testing the cells, it would be best when they are empty? Or full?

Full. If they're empty there's nothing in them to test.

 

[safe]

I initially used Dean's on my bike, it kept vibrating apart so I took it out.

Hmmmm... that's strange. Mine are really tight, almost too tight. They must loosen up over time?

 

[safe]

....adding ten milliohms (depending on your expected current drain of course) in series with each pack will tend to force them into balance

On the surface that doesn't sound like a very smart idea. Here you go to all this effort to install thick cable to keep the resistance down, then you place a resistor as an "add on"?

You want to justify that concept further?

I figured that "balancing" would be a thing I did during the routine "checkups" of the "tubes" when I would pull all the cells out and test them. After testing them if some are bad they get replaced, otherwise they go back in until the next "checkup". I know it's not "automatic" and "fancy", but my guess is that once you got it dialed in right it would most likely stay pretty stable over time. Manual inspection and separate cells give a lot more freedom to diagnose problems than a soldered subpack.

 

[xyster]

Mine are really tight, almost too tight. They must loosen up over time?

Mine fit snug from the beginning, but there was never a 'snap' to them like with powerpoles or even regular blade connectors. The heavy wire on one side draped down a little, consistently pulling the dean's apart within a couple miles.

 

[NickF23]

sounds ok so far. A few points/questions?

1: why all the 12 volt chargers? wouldn't 6 24 volt charger or 3 48 volt chargers be simpler?

2: If your not going to test the individual capacity your cells before you make the packs (the normal way) you could just make 3 48 volt strings and discharge at 1C until to 90% dod and replace the any cells with an unmatched voltage. and return them to bspace/allbattery.

3: watch out for reversed cells. These shouldn't be a problem if you find them quickly could cause damage to the whole pack if they are left in on charge.

4: what is your average and peak discharge rate? don't forget about voltage sag/cooling. heat kills nimh.

5: Bob mcree's expermiments with batteryspace/all battery cells on the power-assist group are well worth reading. I think he has some knowledge of the different kinds of d cells that are sold and there discharge rates/resistance. He sometimes posts this forum so he might be a good person to ask.

 

[safe]

1: why all the 12 volt chargers? wouldn't 6 24 volt charger or 3 48 volt chargers be simpler?

4: what is your average and peak discharge rate? don't forget about voltage sag/cooling. heat kills nimh.

1. I was going to use (6) 24V chargers.

4. My pack overall would be 1440 Watt/hours and 48V 30Ah. If my peak amp draw is about 75 Amps (for the 1200 Watt motor) then that's less than 3C which is supposed to be okay for NiMh that is supposed to do near 5C. Don't forget that I have gears too, so there wouldn't be many times when I'm forced to "bog down" the motor for long. In first gear I would be able to climb a 15% slope while still in the center of the powerband as far as rpms. That's important because the "nasty" heat is in that "death zone" of the lowest rpms. As long as I ride the bike so that the motor tends to "run free" (and more efficiently) it will produce less heat and draw less current. I like to "keep the revs high" most of the time.

On the 750 Watt motor the peak would be only 40 amps, so just barely over 1C.

 

[xyster]

1: why all the 12 volt chargers? wouldn't 6 24 volt charger or 3 48 volt chargers be simpler?

Safe's planning to use unmatched cells. They'll stay in balance better the fewer that are charged in series together. I recalled a post on another forum from an experience member who advised charging in no more than 10 cell blocks. Makes sense. Keep going down in number and you'll have one charger per cell, and then the entire pack would be balanced with each charge. Of course that'd take too many chargers to be practical. But a dozen chargers ain't that bad. I use 20, single-cell chargers to recharge/rebalance my lithium pack.