got some young 2012 modules

The computer looks for voltage and not capacity. That's why it so critical to hook up all modules in parallel for about 4-8 hour prior to assy (I would recommend 8). And... the computer is looking for a voltage difference of greater than .3v before it throws a code. I doubt you will see a variance greater than .15 with those matched capacities.

 

Since the Prius sees only module pairs, it won't even see individual modules. Also, we do peak charge and very low discharge to measure capacity which is not normal Prius charge and discharge levels.

As others have pointed out, the 40-80% range is key to long life.

Bob Wilson

 

Highest and lowest together so they 'average out' for the 19 sense lines of the traction battery controller.

Put the strongest pairs in the middle of the pack and weakest pairs on the end. We have seen a pattern of module failures in the middle where the heat accumulates. Putting the strongest there means their extra capacity delays the eventual pack failure(*). The weakest on the ends ensures they get the most benign environment.

Bob Wilson

* - I know that "heat is the enemy." It is possible to conservatively drive our 2001-03 Prius, the NHW11 model, in such as way that the traction battery temperatures and internal module pressures never stress the terminal seals, the weakest part of the traction battery. But this requires keeping ordinary speeds at 65 mph and below and climbing tall hills at a speed so the traction battery is NOT providing additional energy. These are achievable objectives BUT like Casandra, I know it will be ignored.

 

Hey Bob,
I'm having a hard time understanding something here. If you pair the strongest modules with the weakest modules, how will it be possible to put the strongest in the middle and the weakest at the ends? Do you see what I'm saying?

 

You can either sum the two and use that number to rank the pair. Alternatively, you can average the pair and use that number (turns out both operations will give the same ranking of pair strength.) Let me give an example:

• 6,500 mAhr
• 6,400 mAhr
• 6,300 mAhr
• 6,100 mAhr

So pairing strong and weak:

• 6,500 + 6,100 -> 12,600 / 2 -> 6,300 mAhr :: goes on either end
• 6,400 + 6,300 -> 12,700 / 2 -> 6,350 mAhr :: goes in middle

The divide by "2" does not change the ranking of the pair as strong or weak.

The pairing rule, strongest with weakest, means the voltage swing as the pack charges or discharges for that pair will appear to the traction battery controller to have a similar swing as a less-strong, and less-weak pair (I know the English reads confusing but it is how my mind works.)

Once the pairs are made, we treat them as a module assembly and rank them so the strongest ones are in the middle where the heat stress is the largest. The weaker ones are on the end where they have more cooling because they are not sandwiched between a bunch of module pairs.

Bob Wilson

 

Suggested ordering:

	Column 1
0	[tr][th]new pair[th]mod A[th]mod B[th]pair capacity
1	[tr][td2]1[td2]26[td2]29[td2]6213.5
2	[tr][td2]2[td2]31[td2]25[td2]6217.0
3	[tr][td2]3[td2]30[td2]19[td2]6220.5
4	[tr][td2]4[td2]2[td2]32[td2]6226.0
5	[tr][td2]5[td2]24[td2]33[td2]6228.5
6	[tr][td2]6[td2]5[td2]38[td2]6230.0
7	[tr][td2]7[td2]15[td2]9[td2]6233.0
8	[tr][td2]8[td2]3[td2]34[td2]6239.0
9	[tr][td2]9[td2]17[td2]23[td2]6242.0
10	[tr][td2]10[td2]14[td2]22[td2]6253.0
11	[tr][td2]11[td2]4[td2]21[td2]6241.5
12	[tr][td2]12[td2]1[td2]20[td2]6236.5
13	[tr][td2]13[td2]8[td2]12[td2]6232.5
14	[tr][td2]14[td2]18[td2]10[td2]6229.0
15	[tr][td2]15[td2]11[td2]13[td2]6226.0
16	[tr][td2]16[td2]27[td2]35[td2]6223.0
17	[tr][td2]17[td2]36[td2]6[td2]6217.0
18	[tr][td2]18[td2]7[td2]16[td2]6217.0
19	[tr][td2]19[td2]28[td2]37[td2]6204.0

This ordering achieves:

• within a pair, strongest and weakest - to make balanced pairs
• ordered pairs, strongest in middle and weakest on each end - weakest on coolest ends

Before rebuilding the pack, benchmark the car with the existing pack:

run 10 miles, two-passes, at a cruise control set speed between 35-40 mph​

record outside air temperature, midnight to 4AM is an excellent test time​

Don't forget to parallel balance all modules to a common 8V limit.

After installing the rebuilt pack, benchmark the car with the new pack:

run 10 miles, two-passes, at a cruise control set speed same as the first set​

try to run when the outside air temperature is within +/- 5 degrees of first test​

This will provide an objective view of the new pack effect on MPG versus the original pack. At a constant speed between 35-42 mph, the car will cycle between "engine-OFF, battery discharge mode," and "engine-ON, battery charge mode." There are more stressful, traction battery benchmarks but such tests would push both packs to the margins and one might fail . . . boring.

Bob Wilson

 

Ok, here you go:

	Column 1
0

GOOD LUCK!
Bob Wilson

 

I think Bob read your mind.

 

Had a problem with the last table, here you go:

	Column 1
0	[tr][th]new pair[th]mod A[th]mod B[th]pair capacity
1	[tr][td2]1[td2]26[td2]16[td2]6226.0
2	[tr][td2]2[td2]27[td2]32[td2]6229.5
3	[tr][td2]3[td2]11[td2]12[td2]6233.5
4	[tr][td2]4[td2]5[td2]35[td2]6235.0
5	[tr][td2]5[td2]24[td2]9[td2]6237.5
6	[tr][td2]6[td2]31[td2]33[td2]6239.0
7	[tr][td2]7[td2]4[td2]34[td2]6242.0
8	[tr][td2]8[td2]17[td2]21[td2]6243.0
9	[tr][td2]9[td2]14[td2]20[td2]6255.0
10	[tr][td2]10[td2]18[td2]22[td2]6261.5
11	[tr][td2]11[td2]8[td2]23[td2]6248.5
12	[tr][td2]12[td2]3[td2]10[td2]6242.0
13	[tr][td2]13[td2]15[td2]13[td2]6240.0
14	[tr][td2]14[td2]30[td2]29[td2]6238.5
15	[tr][td2]15[td2]7[td2]38[td2]6235.5
16	[tr][td2]16[td2]2[td2]1[td2]6234.5
17	[tr][td2]17[td2]6[td2]19[td2]6232.5
18	[tr][td2]18[td2]36[td2]37[td2]6227.5
19	[tr][td2]19[td2]25[td2]28[td2]6221.0

The spreadsheet is a little tricky but here you go.

So looking at your future pack, it is in effect not just new but improved over the original due to the lower internal module resistance of the new style. Awesome!

Bob Wilson

 

I can't answer that except to suggest survey the modules and then post the 38 modules as a set with the numbers. Tell the buyer you'll equalize the charge before shipping and then it is their problem.

You might also wrap up your charger and resistor set and offer them as a bundle with the 38 modules and not bother with the survey. In effect, selling a 'do-it-yourself' kit.

If I remember correctly, I sold you a couple of replacement modules? I would be interested in what the after-service capacity looks like compared to the tags I had on them when I shipped them to you. Not quite 'apples-and-oranges' as we use different smart chargers. Still, another data point.

You have a unique window to document the effect of the old and new pack on your mileage. The key is as carefully controlling the test environment with the old pack so you can replicate it with the new pack:

• speed - the same speed in all cases but it needs to be just under 42 mph to get charge/discharge loads or over 65 mph to get high-speed metrics
• temperature - anything within +/- 10 degrees around 70F would be great. Closer is better.
• route - same route with or without hills but mostly without traffic and traffic lights
• distance - 10 miles on a reset trip meter, two passes, one in each direction

GOOD LUCK!
Bob Wilson

 

Ready to assemble and install imo. Good luck and good job!!!!