A ONE, and a TWO and a THREE. A123's Costly Recall: What It Means for the Future of EV Batteries

This week, we were contacted by Tom Brunka of Helwig notifying us that we had the WRONG brushes in our Jim Husted build of the twin Netgain WarP11 motors.

Recall that Mr. Brunka BROUGHT DOWN THE HOUSE at EVCCON during what was supposed to be a Saturday morning dozer/sleeper session at the Electric Vehicle Conversion Convention. A session on the most boring topic we could conceive - BRUSHES. Brunka went over 30 minutes as we sat enthralled. Fascinating session and a LOT we didn't know.

He contacted us to alert us that we had H49 drag racing brushes and he would be much more comfortable if we changed the brushes to his H60 model. These are stronger, harder brushes with a much longer life and that put down a better film while working better for light current loads.

That may be counterintuitive on a motor pair that will likely suffer the indignities of 1000 amps EACH during acceleration.

Apparently the H60's are well able to handle that 1000 amps, but also carry light loads of 20 or 30 amps when idling/coasting. In truth, that's what the Escalade will be doing 99% of the time.

And so he actually supplied us with 16 new brushes at no charge to ensure success with the Cadillac build. Thank you Mr. Brunka.

In every bit of good news, there is a downside. First, we had already done three days of running the motor with the brushes we had. But about that time we received an e-mail from a viewer who had a catastrophic brush failure of the H49 brushes after a total of 120 miles on his car.

Because of the torque converter and mounting, installing the motor on this vehicle will be non-trivial. Worse, our usual technique of putting the motor and other components in where we can easily "drop" a motor on this build is simply not feasible. Once it is in, it would be DAYS of work to paw it back out.

And so we did the change.

I also have a disassembled Netgain WarP9 laying around that we used to discuss motor dynamics and the inter poles of the 11HV. We wanted to put this motor back together with some of the newer Helwig Split red top split brushes that Netgain is now equipping new Netgain WarP9's with now.

So we did a demonstration on the disassembled motor.

As usual, I have a lot of groundless opinions on motors and brushes in my own version of armchair quarterbacking. The reason everyone does this is that it is irresistible. I rankle when our viewers do it, but in truth, I am subject to it myself.

One of the issues is the reversibility of the Netgain motors. If the brushes are offset from center, and if they are radiuses to a round commutator, how could you possibly simply wire the motor backwards to reverse rotation for vehicles such as the Honda, which normally use a clockwise from drive end CWDE rotation when the Netgain is built for a CCWDE rotation?

In any event, a couple of interesting reactions to the show. First an e-mail from Brunka:

Hello Jack and Brian

Very nice presentation on brushes, thank you.

I also liked the way you emphasized that the direction of rotation during brush seating is very important that it matches the direction of rotation when driving the vehicle forward.

Another important point that you made is that it is very important that the brush angle is in the proper direction for your direction of rotation. Yes you are correct we do not want the comm surface rotating into the long side of the brush, instead it should be rotating away from the long side of the brush.

Yes, I did change the brush terminal from a fork lug to a ring lug, we had a couple of reasons for this. Typically an EV application will have the highest value and the most number of overloads per hour.

So in an effort to improve the connection (More contact area and to prevent the spread of the fork when tightened) and to make an easily identifiable difference between the NetGain brushes that are intended for EV applications and those that intended for other applications, we have made the EV brush a ring and kept the fork for the other brushes.

Yes, you have also sold me on the Nord Locks.

Thanks for your help and continued support

Tom

Annoyingly, the new brushes had a full hole connection electrically. The earlier brushes had a forked terminal and you could simply loosen the screw a turn or two and slip ti out. The full hole terminals required you to completely remove the screw, put it through the terminal, and reinstall the screw - kind of significantly complicating the procedure if you are trying to do it on an installed motor and of course at the risk of dropping a screw into the motor.

This has been done to improve the surface area of the connection. Well, ok. Who can be against that at 1000 amps.

I DID notice that they added a mylar sheath insulation to the connecting wires. Gotta love that. And the wires themselves seemed of better material and more flexible. More strands, finer wire. Maybe my imagination.

In any event, we suffered the change on the twin 11's on the bench. Brain was even able to do the bottom ones without dismounting the motors.

George Hamstra of Netgain also followed up with the information that they were absolutely going to the H60 brushes for all new motor builds, though it would take some months to work through current stock which have the H49's. It might be pointed out that there have been THREE total failures noted out of many hundreds of motors sold over the past few years. But in any event, they are going to standardize on the H60's and you will be able to special order the H49's for drag racing purposes.

He also alluded to looking at a neutral bias on the brush mountings. And both he and Brunka thought our idea of putting a nord-lock on the terminal screws was uptown. He alluded that he might get a fight from Warfield but he was going to carry the torch.

We also heard from a little bird in Azure Dynamics commenting on our coverage of their bankruptcy. No leak of company private information actually. They are trying to dispose of their stock of Siemens motors and the controllers they built for them. I won't quote the prices but they are VERY attractive. Downside, 300v of course and the controllers are CANbus controlled. And it was unclear if there was any documentation, and probably no support of any kind. Also Brusa chargers at VERY attractive prices, but 300-520v models which we cannot use.

The most significant notice in this weeks show is the news of A123's actually horrifying boo boo with regards to their 20Ah pouch cell. According to CEO David Vieau, in a letter to everyone, one of four tab welding machines in Livonia was miscalibrated, resulting in a hard to detect defect in the cells that could cause premature failure. The wording of this seemed to allude to a failure of the packs they assemble there, which do indeed use "tab welders" to assemble them. But it has been commonly accepted that he was alluding to the cells themselves.

One of the problems we have with the cells from China is that our sources are not very transparent with regards to the source of the cells. Ours are marked MADE IN USA and Vieau's letter would seem to imply that indeed cells are manufactured in Livonia. We had also heard that cells made in Korea were often labelled MADE IN USA. So we just don't know at this point. And we don't know if our MADE IN USA cells are subject to the recall or are even part of the bad cell output.

The problem does seem to arise from the "compression" of the cells together in their packs. We don't compress cells ourselves. Some weight of cells on top of others in Flatten-em series but we only stack them three high. So I don't expect it to be an issue. If it is, at this price I can live with it.

But A123 may not be able to. After suffering the indignity of Fisker's failure to produce cars after gearing up for batteries for 15000 cars a year - what HAS to be 7 million cells, Fisker has sold a TOTAL of 600 cars we're gold. To pile on, A123 had already done a recall on packs to fix a liquid coolant leak. NOW they will have to recall all 600 Fiskers, along with some other vehicle installations, and extend the warranty on the car from 50 months and 50K miles to 60 months and 60k miles. Vieau's probably worst case projection - $55 million. It's a blow that would take many companies to the mat. ANd possibly this one. It will certainly cause any potential OEM buyers to look askance of a company that had TWO recalls on their batteries in the only significant OEM automobile build they had landed. Never mind that A123 had made a $30 million dollar investment in Fisker stock.

BMW, purportedly an A123 customer, announced this same week that they were going to work closely with Toyota on battery technology.

Virtually all the other battery companies selling cells to OEM's are using Lithium Manganese Spinel or the new Lithium Manganese Cobalt Nickel hybrid cathodes. As you know, I think the LiFePo4 are much better with regards to cycle life and safety. A123 was really the only American battery company sporting LiFePo4.

Hopefully the WAS in my was wasn't really a literal WAS. We're kind of warming up to the concept of the A123 cells now that we can obtain them at a competitive price.

With all these failures, Tesla continues on the march to an on-time July delivery date. One stock analyst, after a visit to the factory and seeing the aluminum stamping forms and presses already mostly in operation, reversed their call on the stock and raised their rating to $49. Tesla's stock, TSLA on the Nasdaq, remains at a 45% of float short sale. That means that 45% of the total available trading stock in this company has been sold short, and with every increase in price, the pressure to bail on this trade increases. Noting that his stock is now the third most shorted stock on the NASDAQ, Musk has vowed to "make it sting - a lot."

As we have a deposit down on an S model, we wish him every good fortune and hope he makes that stick. Our position in Tesla stock pretty much assures us at this point of a FREE Tesla Model S, but we would graciously accept a bonus for loyalty and long term prediction. eBay once bought me NINE collectible MG automobiles when I was just testing if you WOULD actually receive a good car bought on eBay back in 2000. ONE of the nine was outright fraud. Most EXCEEDED my expectations from photos and descriptions.

So we will be delighted to take delivery of our Tesla Model S, paid for by Tesla stock. If it moves a little higher, I may upgrade to the Signature series.

Currently at around $37. We've actually piled on some September CALLS at $40. We'll see how THAT turns out - very risky actually. I still don't like their battery program.

Jack Rickard