while i have the HV battery out of my gen1

Ok,

Scott and I figured out the business aspects:

I'm shipping them shortly and will send him the tracking number. There are three modules:

• #27 - 3.670 Ahr, 8/19/2012
• #28 - 3.907 Ahr, 8/13/2012
• #29 - 3.768 Ahr, 8/18/2012

Once Scott gets the modules, his free to use as desired.

In a perfect world, Scott would charge each of the remaining 37 modules to the dV/dt limit and measure the discharge Ahr capacity. But that takes time. So I'm going to recommend the following:
1) parallel charge the pack to 8.0-8.1 V with one of the three modules
2) remove the parallel wiring
3) take a break . . . say at least a weekend
4) identify the next two modules that have self-discharged the most and replace with my two
5) one last parallel charge to 8.0 V
6) reassemble the pack (MAKE SURE THE END MODULE POLARITIES ARE RIGHT!)
7) put back in the car

Drive.

Scott will need a Prius aware scanner that can read out the 19 module pairs. Over the next couple of weeks, he needs to monitor the pairs and look for any that are going divergent:

- relative higher voltage at high SOC
- relative lower voltage at low SOC

As for the two modules taken out of service, use the Superbrain 977 to master charge/discharge using the dV/dt setting for six NiMH cells and reading out the capacity. They will serve as a sample for the pack condition. If another one or two should fail, use these.

This is a 'band-aid' to get Scott more months of service BUT he is going to have to become a Prius safe driver:

• use "Energy Display"
• climbing hills, slow down if the car is drawing current from the battery, a bad thing in a climb
• use "Cruise Control" to avoid variations in throttle that can cause unnecessary charge/discharge
• speeds no higher than 65 mph to keep out of battery discharge regions
• park so trunk and rear seats are in shade . . . HEAT IS THE ENEMY!

The original traction battery modules are OK if they are kept cool but that means avoiding operating the NHW11 in ways that heat the battery. Fortunately, this also delivers great MPG . . . Prius safe operation is battery gentle.

Gen II and later modules are more tolerant of heat but we're talking about a car that takes 13 seconds to reach 60 mph. Drive the car to get 52 MPG and it will last a long, long time. Change lubricants, fully inflate the tires, and everything else will take care of itself.

Bob Wilson

 

I need to explain the tags on the modules:

	Column 1	Column 2	Column 3
0	[tr][th]27[th]28[th]29[th]metric
1	[tr][td2]8/19/2012[td2]8/13/2012[td2]8/15/2012[td]date charge/discharge started
2	[tr][td2]416[td2]302[td2]482[td]D1	milliamp-Hour (maHr) initial discharge to 6.0 V
3	[tr][td2]3357[td2]3670[td2]3802[td]C1	1st	maHr charge to dV/dt
4	[tr][td2]3670[td2]3907[td2]3768[td]D2	2nd	maHr discharge to 6.0 V
5	[tr][td2]4253[td2]4379[td2]4732[td]C2	2nd	maHr charge to dV/dt
6	[tr][td2] * * [td2] * * [td2] * * [td]rest before second set
7	[tr][td2]2394[td2]2353[td2]2414[td]D1	1st maHr discharge to 6.0 V
8	[tr][td2]4213[td2]4985[td2]4857[td]C1	1st maHr charge to dV/dt
9	[tr][td2]4115[td2]4475[td2]4536[td]D2	2nd maHr discharge to 6.0 V
10	[tr][td2]6060[td2]6520[td2]6161[td]C2	2nd maHr charge to dV/dt or charge limit

Source: discharge/charge history tags

Module history:

• November 2008 - I elected to replace my perfectly fine, NHW11 traction battery pack with a ReInVolt pack using NHW20 modules. The total cost was ~$1,700 and entirely optional, a traction battery upgrade not a failed pack replacement. But my earlier study of the modules indicated these would over time become scarcer and more valuable and I felt this represented a good buy. Our NHW11 requires 38 but each crashed NHW20 provides only 28 modules. So it takes 1.36 crashed NHW20s to make the modules needed for one NHW11. We are lucky that there were ~10x NHW20s so the salvage rate was large enough to sustain the traction battery failure rate of our NHW11s. But this would not last as eventually the NHW20s would begin to see traction battery failures as they are prematurely aged by 'lead footed' Prius owners. Also, I felt a little hypocrite because I did and do recommend ReInVolt packs because I know the technology they use. FYI, ReInVolt had stopped selling rebuilt packs using NHW11 modules, a decision I throughly endorse.
• August 2012 - while my wife was in a physical therapy center, I had a lot of time on my hands at home. So I started a survey of all of original battery modules. They were numbered 1-38 from the control electronics to the end. So modules #27, 28, 29 were from the inner set. Before the pack was replaced, module pair #11 (modules #21, #22) were frequently seen as the lowest pair. Sure enough, one of these modules did NOT take a charge when I surveyed the modules four years later.
• Discharge/Charge protocol - the MRC 989 has three charge limits: dV/dt, temperature, and charge-limit. It was configured to put no more than 6.5 Ahr or 6,500 maHr. At no time did any module thermally stop the cycle but I was using 650 ma., 1/10 C, to minimize risks. But notice how the dV/dt stop was not a hard value. I believe the 6-cell modules have the dV/dt value happen at different times but with six cells in series, the MRC 989 can not get a solid signal that all six cells have reached their full charge. So instead, I find the discharge value the only credible capacity measurement.

So we still have a question of what capacity to claim for these modules. I choose the first discharge capacity in the first cycle. A conservative value, it is also the first, full discharge after the maximum rest period. I would rather give a low value and have the battery exceed that capacity than a high value and disappoint because the module had been left unused or cycled.

Unsupported by any literature, I believe the NiMH capacity is as much an area as a volume of reactants. To understand this requires understanding the chemistry/physics of what goes on inside the cell.

There are two electrodes, the nickel hydroxide layer and metal hydride matrix. Nickel is a multivalent element that has more than one ionization level. So there are two versions found in the cell:

• Ni(OH){2} + OH
• NiO(OH) + H{2}O + e

Now the brilliance of the metal hydride is it absorbs the free hydrogen WITHOUT reacting to it. In effect, the hydrogen gas molecule is stuffed into cavities of the metal hydride matrix. Simply amazing stuff!

To capture the hydrogen gas molecules, I believe the metal hydride surface as to be prepared. The NiMH literature discusses the need to 'active' the cell without details. I found one Korean patent that discussed using elevated temperatures to accelerate the 'activation' but I never found an accurate description of the activation protocol. But there was RC hobby literature discussing the advantages of discharge/charge cycling. My water rehydration experiments indicate this is an important aspect BUT I can not tell if it is charge/discharge at minimum SOC or near peak SOC nor the voltage swings needed. The 'activation' protocol remains an area worthy of investigation IF we plan to sustain these modules into the future decades.

Well hopefully this helps explain what little I've learned about these wonderful beasties. Unlike other battery chemistries, there is little if any accumulation of inert reactants over time. Properly maintained with sufficient water, these modules in Prius service should have a life measured in decades. But make them get hot; leak the hydrogen gas which dries out the electrotye, and; they will hard fail.

Bob Wilson

 

Hi Scott,
Since you charged all modules in parallel and do not know the capacity, it is hard to point to modules that should be paired, so my suggestion would be to follow your idea of swapping the Bob's modules for the 3 worst.
Then you can take the 4 or 6 modules from the ends and place them in the middle of the pack, then re-assemble the pack (remember to start with the correct polarity and place all modules alternating). Over time you might find other weak module pairs so at some point more work might be needed, but I think you need to start using the pack as a series string to find if there are still issues or that it will work well for some time to come... Might be many years.

 

My prius is an 05 with 159K miles. A couple of months back at around 157 miles my HV battery failed. Red triangle and then safe/turtle mode. Being a fix it now and ask questions later type plus busy as heck salvaging old /neglected homes I ordered a module off Ebay and spent a day replacing the bad one. Clueless for sure. My prius is for sure not pristine as I use it to haul about everything except for plywood and drywall. The 4x 8 sheets won’t fit. If I could handle them would sell my truck. It has run great for the 2k plus miles since HV work done. Did not balance anything just slapped the new module in. Here is the weird part. Overall mileage before cell replacement for the first 147K miles was around 48. Since it is around 56. Works for me.



I plan to drive the abused but much loved old girl from Alabama to Oregon to do a bicycle tour and then meander my way home via touring the Southwest.



Am I asking for trouble?

 

One more parallel charge but with my three modules:

• one to replace the failed module
• two to replace the weakest, most self-discharged modules

Bring them up to 8.0-8.1 V and then let them rest over your holiday. When you come back, record the last self-discharge voltages and get it back in the car. As for pairing, I'm partial to strong and weak with the strongest ones in the middle and the pairs with the weakest ones on the ends (the coolest parts.)

CARE FULL! Make sure you don't reverse the series voltage!!!

Put the big buss bar on first. It will be on the 'forward side' of the pack with the control electronics on the driver side. Ohm-out the sense wires and make sure they are good before going any further. If a sense wire is bad, fix it. Make sure the terminal nuts are torque tight . . . don't guess . . . do this d*mn job right.

I'm not too worried about the footer screws but do at least one but both is correct. If you are ever rear-ended, you don't want the traction batteries to short to the case and compound the accident with a traction battery fire!

Finally put the two splits on, one at a time. During this phase, use the "one hand rule" and also put on some non-conductive gloves. They don't have to fancy but even if you have them on, treat them as useless.

Patience, calm, and coffee . . . do this d*mn job once and know that it is as good as it gets.

GOOD LUCK!
Bob Wilson

ps. Looking forward to the final payment.

 

You've gotten three modules that need to be charge balanced with the rest before final assembly:

• final parallel charge to 8.0-8.1 V with the replacement modules
  • choose which two weak ones to take out
• let it rest over the holiday
• use the last 'self-discharge' levels to pair the modules

The Prius expects the maximum voltage difference between the lowest and highest pairs to be less than 0.3 V. Use this rule to do the pairing. During your days off, think about how you want to pair them up but the 0.3 V maximum difference between pairs is a hard limit.

Bob Wilson

 

Yes assuming one of them is 'the bad one.'

You are putting a final balancing charge on all of the modules so they will start at the same voltage "SOC". When you come back, we'll check their levels just to make sure the modules I sent are working per the survey from last year. I don't think they need conditioning as much as balancing the charge on all modules and making sure you've got a list to 're-pair' and 're-order' all of them . . . the final assembly.

Don't forget we'll have to check the sense lines . . . they break easily and that can cause a 'snapped line' code (which escapes me at the moment.)

Bob Wilson

 

hello bob! can you give us newbies and non-engineers an idea of how you did the math to arrive at your amp hour (Ahr) readings given your discharge/charge cycle calculations on the tags? it looks like you put each module though 4 discharge/charge cycles correct?

 

I use an MRC 989 configured for 6-cell, NiMH cells, dV limit, 1V/cell discharge low, charge limit 6.5 Ahr, auto-cycle. The unit also has a temp probe in case the battery starts to thermal runaway. It records the Ahr for each discharge/charge cycle so I read them out and write them on a tag.

Bob Wilson

 

thanks bob! but i'm still wondering why it happens to be the third line down on each tag? ex. #28 is 3.907 Ahr and the third line down on tag #28 is 3907-D2 why isn't it any other D1 or D2 on that tag?

 

It has to do with how the MRC 989 works.

Each cycle consists of a "discharge" followed by a "charge." After all the RC hobbist wants to use a fully charged pack. So you specify the number of cycles.

The MRC 989 only has eight, memory read outs. One is used for the discharge and the other for the charge. This is enough to cover four cycles.

In my case, I wanted to have some overlap in the memory with the first two and last two cycles. The first two cycles generates four values and shifts the previous values up. So I specified two cycles which is:

1. discharge to 1 V/cell or in our case 6 V
2. charge to:
   1. dV/dt - the momentary decrease in voltage during NiMH charge
   2. charge limit - there is a 12% overage so I set it to 6.5 Ahr*88% (this was learned experimentally)
   3. temperature probe over 120F

After the first two cycles completed, I would record the date and four values. Then I would start the second charge cycle. Sometimes due to work or sleep, there might be hours between the first and last set but in the overall world of things . . . a wash.

We know batteries respond to a 'rest' period. However, I suspect these voltage adjustments are not significant in the overall scheme of things. The longer the rest period, the less the duration and the more the actual performance characteristics of the battery comes into play.

Bob Wilson

 

ok, i'm using a supermate dc 6 and i have the discharge set to 0.5A at 6V with my charge cut off at 7500. i did 3 cycles of discharge/charge last night and got these numbers this am:

d1 0014
c1 2375

d2 0002
c2 1782

d3 0002
c3 1769

what would my amp hour capacity be for this module? thanks!