Dorman HV Battery with a new lease on life.....

There are many threads on the forum discussing Dorman and many other brands of remanufactured HV batteries. I've provided my fair share of information in some those threads. Being as objective as possible, I'm starting a new one to discuss a Dorman battery I recently took into inventory. It'll take me a little while to get the thread caught up to "now" time, so bear with me a bit. Much of this was previously discussed in the "Dorman Battery Experiences" thread. I'm going to copy the relevant posts into this thread so it stands on it's own. I'm only going to copy my posts from the other threads, so some may look a bit weird since they may be replies to other members posts. I plan to update this thread fairly frequently to track this battery.

This all started with forum member @Stimp having HV battery problems. A Dorman remanufactured HV battery had been installed in the car by the previous owner. We arranged for installation of a TMR-JWAP battery on May 4, 2019. And so begins the the tale of the new life and times of this Dorman battery....

 

So, this post takes off immediately following post #265 in thread "Dorman Battery Experiences"

Here's a bit of follow-up on that Dorman. It's serial # 02942 Built on 3/7/14. I think the PO thought it was installed in the car sometime in 2015. It has 27 modules date coded 29XE (October 29, 2003) and 1 module date coded 149O (September 14, 2013). After completing the swap, test driving and then re-installing the interior, we pulled the cover off the Dorman to see how many different years of modules it would have. A quick look and we thought it had all 28 matching with 29XE. I was stunned and actually looked at it again. I have no idea how we missed that one module that appeared to have no serial number. It was one the modules that has the serial number partially under the clamp crossbar. I didn't catch it until I was recording serial numbers for the capacity testing. All the modules had one end colored in with black marker. Except the new one. IDK how the heck we overlooked that also. It must have been some kind of eye game, as all three of us didn't catch it.

I did some minor disassembly and placed the module pack on the bench for initial discharge testing. I measure and record each module voltage prior to starting the initial discharge. I always discharge first to measure how much energy is stored in each module in the 'as removed from the car' condition. It gives a good indication of overall balance between modules. Afterwards, the modules were charged, then a discharge/charge cycle was performed starting at 530am this morning. Here's the results. The left table has the modules in order as installed in the pack, starting with block 1 (away from the ecu). The right table has the data rearranged with the modules sorted from lowest initial voltage to highest. The numbers under the discharge columns are mAh measured when discharged to 5.8v. This way, you can easily see how a modules voltage is not a sure indication of it's condition, or how many "units of energy" it can store. If one of the $399 "come to your house and fix your battery" Craigslist posters had "fixed" this battery, does anyone think the 149O module would have been replaced? Even with the highest voltage (7.72), it's a brick, barely better than the module that has the failed cell. I'd imagine 566E would have been replaced and then @Stimp would be given the car back. Guaranteed failure again within a month or two. Looking at the mAh values on the second discharge, there's nothing there to brag about. It's showing it's age. Quite a wide range of capacities, even though they were all charged identically. Nothing unexpected. We'll see how it looks after more cycles. I have some time available to invest in this, so maybe after several cycles and data, I'll replace the 2 obviously BAD modules and then hit it with a full prolong sequence and then check the capacities again. It will be interesting to see what capacity can be recovered. There will be no further data for module 566E. I've already disconnected it from the chargers. That puppy was swelling like crazy trying to take a charge....I'll probably disconnect the 149O (2013) module after tonight's cycle. No sense wasting electricity on it.





Notice the module in middle doesn't have serial number in same place as others. Stimp, how the heck did we not notice, lol? You can see some of the serial number between the green temp sensor wires. A common Gen 3 location.

 

Here's today's update on the module cycling. This covers the initial discharge and three follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
Many modules are still recovering capacity by hundreds of mAh, but 572E looks like it's struggling and may be near the end of it's rope. Another potential candidate for the core pile. I'll give it another chance or two.

 

Here's today's update on the module cycling. This covers the initial discharge and five follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
Many modules are still recovering capacity by hundreds of mAh, even 572E has come up by 1000mAh over the last 2 cycles. I looked back through my data and realized 572E had a charge cycle that stopped about 2000mAh early, probably due to a false delta V trigger. This is what caused the low discharge reading on the fourth discharge and why it's trailing behind the others. I'll think I'll keep it cycling for a bit and see where it ends up. Most likely the module is fine and will recover like the others.

 

Here's today's update on the module cycling. This covers the initial discharge and seven follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
All modules are still recovering significant capacity. The higher capacity modules improvements are slowing and the lower capacity modules should start catching up as more cycles are performed.
I expect to see, very soon, a dip in capacities, followed by a good jump. I've added a row at the bottom showing the average capacity of each column, with 566E and 502E removed from the calculation.

I've highlighted "green" modules/capacities that have exceeded 5800 mAh.

80% of rated capacity would be ~5200mAh.
85% of rated capacity would be ~5525 mAh.
90% of rated capacity would be ~5850 mAh
95% of rated capacity would be ~6175 mAh

 

ericbecky,

Fully agreed. It's no different than any other high volume automated procedure. It's all about the equipment and process. A hobbyist could build a car in his garage if desired. Ford/GM could build a million or more in the same period of time. There's proper facilities, technology and equipment for every process. It just depends on scale. These companies wouldn't be in business if they weren't doing something right and filling a need.

The problem I'm seeing is the lack of attention to detail. Small parts missing, cooling air flow path foam inserts ripped off and not replaced, temp sensors not fully mounted and falling off, etc. That's where I see their process having weaknesses. Every company and individual makes decisions due to financial constraints. A company is in business for one reason, and one reason only....to make money. That means efficiency. Employees have a big impact on that. A person who is interested in what they're doing, and has a bit of passion for doing it right, will do a better job every time.

 

Here's today's update on the module cycling. This covers the initial discharge and nine follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
Modules are still recovering significant capacity.
Had a storm induced power blip during the charge portion of the last cycle. Some chargers stopped their cycle, most carried on unfazed and one bricked a channel, losing its readings, but went back to normal after cycling power to the charger.

I've added a row at the bottom showing the average capacity of each column, with 566E and 502E removed from the calculation.

I've highlighted "green" modules/capacities that have exceeded 5800 mAh.

80% of rated capacity would be ~5200mAh.
85% of rated capacity would be ~5525 mAh.
90% of rated capacity would be ~5850 mAh
95% of rated capacity would be ~6175 mAh

I'm glad to see they're still showing significant improvement each cycle. My condolences go out to anyone trying to do this with a single channel hobby charger. Even a single 4 channel charger would probably take over a month at 24/7 to do what is displayed below.

 

5/13/2019 update on the module cycling. This covers the initial discharge and eleven follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
Modules are still recovering significant capacity. They're like energizer bunnies.
Had a storm induced power blip during the charge portion of cycle ten. Some chargers stopped their cycle, most carried on unfazed and one bricked a channel, losing its readings, but went back to normal after cycling power to the charger. The three channels that stopped the charge early had a bit over 2 hours of additional charge performed. I was occupied with other things and wasn't able to stop the additional charge when planned, so those 3 modules ended with higher total mAh charged than the others. That's in the notes under the spreadsheet. The module that was connected to the charger channel that faulted was just reset and had a new cycle started. That's why it's discharge is so low on #11.

I've added a row at the bottom showing the average capacity of each column, with 566E and 502E removed from the calculation.

I've highlighted "green" modules/capacities that have exceeded 5800 mAh.

80% of rated capacity would be ~5200mAh.
85% of rated capacity would be ~5525 mAh.
90% of rated capacity would be ~5850 mAh
95% of rated capacity would be ~6175 mAh

I'm glad to see they're still showing significant improvement each cycle. My condolences go out to anyone trying to do this with a single channel hobby charger. Even a single 4 channel charger would probably take over a month at 24/7 to do what is displayed below.

 

For comparison, the information below is from a battery pack removed from a 9k mile 2015 wreck. This pack was used to build a Gen 2 battery. It was installed in a fellow forum member's car in Savannah, Ga in April of 2017.

 

5/14/2019 update on the module cycling. This covers the initial discharge and thirteen follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
We finally got the dip. Much later than expected, but the modules are still recovering. Energizer bunnies, indeed.
I wasn't able to start this evenings cycle until late due to a company golf outing. But yesterdays continue to look promising..
I've highlighted "green" modules/capacities that have exceeded 5800 mAh.
In the big picture, the test results of these 26 modules are on par with MUCH newer modules. I'm impressed...

80% of rated capacity would be ~5200mAh.
85% of rated capacity would be ~5525 mAh.
90% of rated capacity would be ~5850 mAh
95% of rated capacity would be ~6175 mAh

 

5/15/2019 update on the module cycling. This covers the initial discharge and fifteen follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
We got another dip. Not unusual with this many cycles. #15 did it's discharge (a little dip) but the charge was stopped about an hour early. This was because I started the previous set of cycles late. The charge was still in progress when I needed to go to work this morning. Iwent ahead and stopped that cycle, recorded the data and started a new cycle. That's why #16 is lower. Nothing nefarious going on, they had all charged between 6100 and 6600 mAh instead of the full 8200.

It should be interesting where the next set of capacities end up, as they'll be back to normal cycle values...as long as there are no more storms or golf rounds to distract me.. This experiment should be drawing to a close soon.

I've highlighted "green" modules/capacities that have exceeded 5800 mAh.
In the big picture, the test results of these 26 modules are on par with MUCH newer modules. I'm impressed...

80% of rated capacity would be ~5200mAh.
85% of rated capacity would be ~5525 mAh.
90% of rated capacity would be ~5850 mAh
95% of rated capacity would be ~6175 mAh

 

5/16/2019 update on the module cycling. This covers the initial discharge and seventeen follow-on full discharge/charge cycles.
566E and 502F are down for the count and will be cores.
It looks like we're coming to the end of the road for this experiment. Cycle 17 is rockin' the house. Cycle 18 dipped a bit, but not unexpected or significant at this point. I'm doing the evening cycle tonight but don't plan to start another one tomorrow morning. I think they're pretty much maxed out on the recovery and they are looking absolutely spectacular.!!.

Even 572E has caught up and exceeded most of the others.

I'll document the mornings results and then just leave the cooling fan running all day. I have plans for tomorrow evening, but sometime this weekend, I'll replace the 2 bad modules with some that match the rest in the pack and reassemble the battery. I'll throw it in the car for some testing.

I've highlighted "green" modules/capacities that have exceeded 5800 mAh.
The average capacity of these modules has significantly surpassed anything I would have guessed. Did I mention I'm impressed?

If a fellow DIYer was trying this with a single channel hobby charger (and assuming it's one with a 10 watt discharge instead of 5 watt) this would have taken over 166 days going 24 hours a day. A single 4 channel charger would still be sitting on almost 42 days worth of work.

80% of rated capacity would be ~5200mAh.
85% of rated capacity would be ~5525 mAh.
90% of rated capacity would be ~5850 mAh
95% of rated capacity would be ~6175 mAh

 

Last cycle for this pack...the final three cycles are fairly tight. I'd like to think this can show what cycling is capable of doing. I WOULD NOT expect similar results from every pack. That's guaranteed not to happen. ....I'll rebuild the pack this weekend with a couple similar replacement modules and throw it in the car for testing when I have a chance.

 

This isn't a problem for me. I have 1000+ to choose from.

The two I selected as replacements are from the original HV battery from the black 2005 prius I purchased for my son last August. That battery had 4 bad modules, but the others tested very well. They had 8 cycles performed and then placed on a shelf. They've been sitting since September, so it's been over 7 months that they've been sitting. All 24 good modules were still within .05 volts when I checked them. I feel safe saying there are no self discharge problems with these modules.

I've already installed the two replacements and performed an initial discharge plus one full cycle on them:

227G1A02509B tested at 6864 mAh
227G2A02851B tested at 7015 mAh

using the same exact settings and methodology as I've been using for quite a longtime. I just started assembling the battery this evening. Pretty confident it's going to be a totally kicka** Gen 2 battery. I'm looking forward to installing and testing it. I currently have a Gen 2 battery in my car that was built with July 2015 date code modules. That battery was built very recently and was in my car for testing. That battery will be going up for sale shortly. Will be interesting to see how this 2003/2005 module battery will compare. These are some pretty good "old" modules.

 

So, the above post was my final post in that thread about this battery. From here on out, posts will be "live".

With the exception of performing 19 conditioning cycles, I tried to perform this rebuild just as if any DIYer might.

Some items noted during the original disassembly:

1. the ecu had it's identification sticker removed and a number written in it's place with a sharpie marker. I have no idea what version ECU this is. Not a performance issue, just an annoyance. -1/2 pt for this.
2. the foam air flow blockers that should be at the ecu end of the pack had been ripped off and not replaced. -1 for Dorman. These foam air blockers minimize (but do not eliminate) cooling air flow being lost. No blockers=excess air flow going to electronics section instead of down between the battery modules.
3. the inside of the module terminal covers had Gen 1 terminal sealant splashed on them in many places. This is some sticky gooey nasty stuff. Apparently they are Gen 1 covers, but NBD, as they are physically identical.
4. there were two different sizes of nuts used on the module terminals. Again, NBD as it should have no impact on performance.

The battery was built, using all the components originally in the Dorman, the 2 replacement modules and correction of the minor issues noted above (except for the ecu sticker).

1. The voltage sensing wire harness and tabs were carefully removed and cleaned.
2. The busbars were already nickel plated and cleaned up easily. +1 for Dorman.
3. The main cables had been soldered on the ends where the copper cable is crimped to the lug. This does wonders for preventing corrosion from working it's way up the cable. +2 for Dorman on this.
4. All the sensor tab crimps, temperature sensor wires, cable plugs and ecu ports were inspected and looked to be in great condition. +1 for this.
5. All relays and other components were inspected and tested if possible.

Several days after assembly, the battery was installed into my 2007 for real world testing.

 

A few days later, on June 3rd, I performed the HV Battery test provided by the Hybrid Assistant App. I performed this after driving the car for several miles to ensure the engine was fully warmed. All the windows rolled down and air conditioning at coldest setting with fan on max. It was a warm day and it provided a 12 amp draw from the HV battery. The current initially flickered between 12 and 10, but then steadied out at 12 amps. Here's some screen captures from the report.

On the first screen capture, it shows the voltage of each of the 14 blocks. A perfectly matched set of modules would appear as one solid line, since they would all just lay on top of each other. A stronger or weaker module can easily be seen on this style graph, as it's curve will diverge from the others in the group when it holds voltage or loses voltage at a different rate. These 14 blocks are very well balanced.
SOC started at about 66% and I ended it at about 45%, when the engine kicked on. You can see the nice linear reduction of SOC. The blocks were not even close to the 14.4 volt NiMH plateau when the test ended.




The below screen capture shows the highest block voltage and the lowest block voltage over the entire discharge. The light blue line in the middle is the average block voltage. A weak module or block will cause the "lowest" line to be lower and the average line will be skewed more toward the "highest" line. A well balanced battery will have the average line perfectly split the middle. A large gap above with a small gap below indicates a single very strong block. A large gap below with a small gap above indicates a single weak block.


Unfortunately, the engine kicking on at 45% caught me by surprise. I expected it to occur around 43% and had planned to stop the test at about 44% as the test had just passed 1300 mAh discharged (one of the best I've ever seen using this test). The min voltage in the below screen capture is due to the sudden high current load (54 amps) occurring when the HV battery was starting the engine. The max voltage recorded was due to the hybrid system starting to charge the battery (20 amps) after the engine started.

 

Obviously, a typical DIY person is not going to have the equipment or time to perform 19 conditioning cycles on a set of modules. Without going back and getting exact dates, I seem to remember this took 10 days of bench time using (7) 4 channel chargers doing 2 cycles per day. Someone who buys one 4 channel charger (~$200) would be looking at ~70 days of 24/7 to completely duplicate this. It's a bit unreasonable to expect anyone to do that, but I wanted to express that Dorman actually had a pretty good battery here. It lasted almost 5 years from the time it was built. Replacement of 2 bad modules (of which one was a Gen 3 module) and some seriously overdone reconditioning has resulted in this battery being one of the strongest I've tested. I would never, ever, have expected this from a remanufactured battery, especially with modules manufactured in late 2003. Think about that. Almost 16 year old modules. This battery deserves to be installed for a while for additional testing.....

 

I'm aso going to put this up. I decided to do an acceleration test. Maybe not the best choice in the world considering the rain we've had the last couple days, but....yesterday morning, on my way to work, I had stopped on the shoulder of the highway to take a phone call. Since I had just reloaded my OBDLink app the night before, I decided to try the acceleration test with this battery. I did this previously with a battery I built using Gen 4 modules. It was enlightening. I'll have to find that one and post it again. Anyway, the test is self starting and it initiates as soon as it detects movement of the throttle pedal. After waiting a few moments for a large gap (we have those here in SC!!) I put the pedal to the floor and almost immediately had traction control engage. I went several feet and it engaged again when I hit a bump where fresh pavement started. So even with traction control engaging twice to kill the throttle, here's the results.........................I'll provide an update when we have a few dry days and I'll take the 100 pounds of tools and 70 pound bag of deer corn out of the trunk. Maybe I'll even take my golf clubs out of the back seat also, lol.



And here's the one from a year or so ago when I was testing a Gen 2 battery built using Gen 4 modules:

 

Ahhhhh, yes, deer corn... No, we do not hunt. We merely feed. We have about 20 deer in our area that come roaming through. We have bucks, does and fawns in our front yard every summer. We've named many.

This is Jaggar--




Jaggar's son Clover and Clover's wife n kid