Monday, February 13, 2012

OBSESSIVE COMPULSIVE DISORDER AND A123 BATTERY CELLS

What do you mean obsessive/compulsive?

This week we continue our obsession on A123 cells. I fear this obsession on developing a mechanical package for these cells to use in a car is going to cost me my entire viewership of our EVTV weekly show.

But we continue to pursue it for a number of reasons.

FIRST. We think this higher power cell opens the door to a different kind of power pack for electric vehicles. Not dramatically different, but just conceptually different. Our DIY crowd lives in a strange space of time, space, and money restrictions that VW and BMW simply do not face.

As a result, EVTV has lived in a strange grey zone of tension between those who want to build excellent electric vehicles and those who want to do it inexpensively. There is a little army of guys out there still building lead acid vehicles because they perceive the $3000 of lead cost as workable and $10,000 for Lithium as not. They desperately want to be perceived as "pioneers" in the electric vehicle movement. When I tell them they are not only NOT pioneers, but actually damaging to the cause, it drives them into a frenzy of hostility and abuse mostly directed at the messenger ME.

Let me repeat: lead is dead. IT is at this point a NOT very interesting science project. It is not a car. ANd it reenforces false stereotypes about electric cars among the public. If you have a lead powered electric car, please hide it from view.

But the desire for a lesser cost battery, and the willingness to settle for less range, just might be an itch we can scratch with the A123 MD1HD-A 20Ah pouch cell.

Here's why. This cell only has 20Ah of energy in it and in testing, really more like 19 Ah. But it can put out 20C of POWER and so it's POWER density is quite high. To drive a 1000Amp controller and motor, would only require three of these cells in parallel. A 60Ah pack would cover up to 1350 amps actually which encompasses the power requirements of every single vehicle we've done at EVTV including the eCobra.

And so a 100 Ah pack is actually overkill for a normal car with regards to power or instantaneous current requirements. But it is a smaller pack than we normally use.

As an exercise, let's redesign our pack for Speedster Part Duh. This vehicle is limited to 120v by the controller. If we did a 100Ah version of this at 5 cells in parallel, we would actually be a little less than that, probably 90-95Ah. At let's say at 120v, such a pack would of course have 10,800 kWh. But the weight of the cells, not counting any modules, would be more like 180 lbs instead of our current 450 lbs and could conceivably bring our wH per mile down to 200. That's a 54 mile range to 100% DOD and of course 43.2 miles to 80%.

The national average for a daily drive is 39.4 miles and 78% of the population can deal with a car that does 40 miles or less.

The car would be lighter. And the cost of the cells based on our last purchase would be $4770. The cost of the cells we HAVE in it now for an 80 mile safe range is a bit over $8000 with shipping.

And so we have lead acid, with 1200 lbs of cells and about a 30 mile range at $3000, the CALB 180's with an 80 mile range at 450 lbs of cells at $8000, and the A123 at 180 lbs and 43 mile range at $4800.

The A123 fills in nicely and the CAR actually PERFORMS better than either of the other options because of lower cell weight and so lower vehicle weight.



But perhaps most importantly, the initial cost is closer to the lead $3000 cost than it is to the CALB 180's $8000 cost. And the car is not only fully functional, but probably super functional as it will drive better at a light weight than it will with the CALB 180 cells.

IF you then like the car, and need more range, ADDING cells to the system is almost a trivial exercise.




And so we see these cells as allowing a whole new group of our viewers who mostly view, finally get started to build. That's a big play for us and a big play toward the adoption of electric drive for personal mobility worldwide.

REASON TWO. I'm just incensed at the hubris of Miet Ming Chang and the A123 group in accepting $249 million dollars and publicly dissing ANYONE not an OEM qualified in their mind to be an OEM right on their web site WITH my tax dollars in their jeans. I'm further incensed that they would lay off 125 hapless yucks the first time Fisker hits a glitch, saving themselves what $7 million a year in payroll while lunching at the Fish on their $249 million dollar grant.

I'm THEN incensed that they actually make the cells in Korea, and have MADE IN USA printed on them there.

And then I'm enormously entertained that FISKER has now missed the milestones on THEIR Department of Energy LOANS and have shut down their production and layed off 65 workers. This is a company A123 invested $30 million dollars in, and lo Fisker announced A123's cells to be the very highest quality available.

Now A123 has a rather diminished need for cells.

And so it APPEARS that the Asian factories, facing plummeting demand from A123, are actually selling the cells out the back door to Chinese traders. And there we purchase them.

Several people have alluded to the fact that these may be "reject" cells from QA or "seconds" or otherwise undesirable cells. I would note that EVERY person I have heard this from also sells competing batteries. It is just dehumanizing to watch the "who's ox got gored" scenario play out EVERY TIME in EXACTLY the same way.

Our examination and testing would tend to indicate that the cells we are receiving are brand new and of acceptable quality per the published specification. At some price, the performance is what the performance is and we find 18.5 to 19.5 Ah of energy density and a full 23C tested current capacity very persuasive at these prices.

So reason two is simply that it tickles my fancy to buy cells from China purportedly made in the USA from a US manufacturer who's head is so far rammed up their own ass you would have to e-mail them JPEGS of sunlight if you ever want them to know what it looks like.

REASON THREE has to do with establishing the ongoing market price of these cells. The factory and traders have to make some level of profit to continue making the cells at all. I don't know where the floor is. Our first purchase was a little over 2 months ago at $36 per cell I think. Our latest purchase in December was at $20 per cell plus shipping and paypal charges. Nathan Knoppenberg reported this week a quote of $17.40 FOB china. I asked my guy and he quoted $19.20. As these two quotes were from the same company and our viewers have already purchased several thousand cells from this guy, I kind of went ballistic on him. His response is that I can have them at whatever price I want to pay.

Lower pricing is good. You want to squeeze. But when you squeeze all the oxygen out of the room, understand that the supply might just disappear.. In infant industries such as this, while you are seeking the lowest price, understand that it is generally in your interest for these people who make motors, controllers, batteries, et al to remain in business and for it to actually be an attractive business for others as well.

In any event, CALB cells are at $1.20 per amp hour and TS/Winston/Sinopoly somewhere around $1 to $1.05. At anything like $19 and below, these A123 cells suddenly make sense in a lot of ways. At $15 they pretty much kill off the larger prismatics.

REASON FOUR - it may just be a better battery. A little over a year ago, in our December 6, 2010 blog entry "What heSaid, What I said" a professor Jay Whitacre of Carnegie Mellon University did a video addressing the Carnegie Mellon EV club and talking about batteries, battery management systems, chemistry, etc. This guy has had quite a career specifically with Lithium batteries going back to the NASA/Jet Propulsion Lab and a decade of it. In this video, he several times held up the A123 cell as the "gold standard" by which LiFePo4 cells should be compared. Since then, A123 has done several presentations to DOE as to how they have improved their cells, and mostly this 20 Ah cell. It is a LONG way from the little can cells used in the deWalt power tools.
(Incidentially deWalt as it turns out has a patent on a little BOTTOM BALANCING battery charger).

Now Daniel A Cogswell and Martin Z. Bazant of MIT have published a paper, Coherency strain and the kinetics of phase separation in LiFePO4 that would appear to imply that fracturing of the crystalline matrix in LiFePo4 cells may be DIMINISHED by the use of HIGHER current levels during charge and discharge. In other words, the harder you work em, the longer they last. I have for about a year now thought that the disconnect in the broad electric vehicle community that I feel is about really smart people NOT looking hard enough at these LiFePo4 cells because of their lowish energy density. They don't REALLY know why these cells work, and yet they are moving on to others that have none of the same advantages, for a few ma of energy. How about driving THIS chemistry to its limits first.

Our latest module uses a mold we made from a CALB 180 prismatic cell using Mold Star 30 Platinum Silicone Rubber from a company titled SMOOTH-ON. This took about 16 hours to cure but gives me a very durable but flexible rubber mold to pour cast urethane in. The material needs no release agent when used with urethanes. So you just pour it in and in 10 minutes pull it out.

The urethan resin we used this week (resin du jour as we try a lot of different ones) is their SMOOTH CAST ONYX which has a deep glossy black finish. It takes about 40 ounces in our mold. The bad news there is there is this stuff is $94 per gallon meaning we could do three batteries per gallon or EACH cell would take about $31 plus shipping just of resin. We probably need to find something less expensive.

Assuming about $10 worth of hardware, $31 in resin, and six cells at $26.50 our little battery costs about $200 each for 115 Ah. And our 36 cells in Speedster Duh, instead of the theoretical $4800, is more like $7200. Of course, that is with 120 Ah instead of the 100Ah described at the beginning. But you can see that the form and the cost of the actual modularization of these pouch cells is what it is entirely about. You can wipe out the advantage of them if you have too high a costs in the module. And of course you wipe out the advantage if the modules can't handle the current or lead to early cell death.

So we are doing more work on a module. But we had kind of hoped our viewers would do some work in this area as well.

We also took a first look at Valery Mitzikhov's EMW Bluetooth Dashboard for Android in this episode. MORE ON THAT LATER. This is a fascinating device with a few early version problems.

Yes, the audio is wretched in the first part of the show. Brain forgot to change batteries in the Juiced Link. We would not have even HAD a show but for our new backup TASCAM DR-100 recorder. But even there, he had the microphone pointed at the wall instead of at us and so the sound is like being in a bathroom. We'll do better. I hope....



Jack

155 comments:

  1. Hi Jack,
    I disagree with loosing viewership because of the obsession with A123 batteries. I tune in specifically to see what you've done and how your doing it. :)
    Good luck, keep up the fight, and keep the A123 cell updates coming!

    ReplyDelete
    Replies
    1. Jack,
      Can you tell us the company that quoted the $19.20/cell fob China?
      Thanks

      Delete
    2. This comment has been removed by the author.

      Delete
    3. OOPS! Typo, keep quoting GBP instead of USD.

      victpower.cn is $19.2/pcs but told me they can only drop to $17.4/pcs if I buy in lots of 500 or so.

      I'm on Skype with him this minute. Last quote was $19. I did remind him I want 180 for a first order, (boxes of 36). Said I'd get back to him.

      Delete
  2. First another great Show! I am happy that Brian's father is doing well. I wish them well.

    As for the casting, I think a UHMW mold might be a good solution. I have used these in the past to cast urethane products. I would be happy to machine a mold for you on my CNC mill. I would probably make a three or four piece mold. I would think that bolting the power tab to the top of the cavity in the mold will give you a more consistent cell size and ensure that you do not cover the tabs with resin. That way the cells can hang from the top within the cavity (maybe a little tape to hold the cells together) On a level table this would eliminate the need for the spacers.

    I also think that you can use some fillers to help to reduce the volume of casting material, and potentially help with heat dissipation. I have found that ground Marble (Calcium Carbonate) is a good choice.

    The cheapest thing I have ever used with great success it talc powder. There are also some poly foam products that are cheap and will work well if you are not worried about heat transfer. These could probably take about a pound out of the cast weight.

    The biggest problem with fillers is that they may make the pour too thick. It would also be necessary to put in a little additive to slow the reaction down a little to let the air bubble get out.

    Smooth On has some fillers, but I did not see any listed that I thought would help with casting cost or mechanical properties that much.

    The only other suggestion I would make is to tap the holes in one of the side pressure plates. This will eliminate the need for the nut and one of the Nordlock washers and make the assembly a little faster.

    Anyway, good job on the cells. I do applaud your efforts....

    ReplyDelete
    Replies
    1. If you made a mold that Jack liked and featured successfully in his show, you might find a small bit of commercial interest. Could such a thing be at least somewhat mass produced? A mold for the mold?

      Delete
    2. My machine shop is a hobby. I really would not want to turn it into a business, but I would be happy to make a few molds that the community can share....

      Delete
  3. Hey Jack,

    I would like to add my voice to the "here because of a123 cells" crowd. I found this site because I was investigating those cells (always hated the idea of lead batteries for an EV, figured with the proliferation of Li-Ion out there, there had to be a good alternative to lead.. then I found out about LiFePo4 and that lead me to the a123 pouches). I'm not going anywhere before you get a workable system in place for making excellent a123 modules :)

    Something I was thinking might help with your casting.. Why not just some plate aluminum drilled/bolted together to make a "tank". You can still clamp the front on (or use the poly so you can watch it/demo), but then have some kind of hanging device/bar across the top. Bolt your terminals to it to keep the cell off the bottom. Perhaps as noted by Jeff, tape them together for compactness and maybe tape/glue some spacers to assure minimum thickness on all sides. Maybe put a small drain hole near the top (side) so all the modules end up the same size.. just pour till a little dribbles out the drain.

    Sure, you'd not have the "ribs" on the case, but do they actually do anything on hard plastic casings?

    And at $100/gal for the current plastic, I'd agree that something else might be more cost effective (though, I have no suggestions at present). Though, with a decently sized case (minimize thickness and waste) and a good/effective filler, you might get a lot more than 3 cells out of a gallon.

    Keep up the good fight, guys! And thanks for all your efforts in getting the kinks worked out.

    ReplyDelete
    Replies
    1. Oh, the top bar would obviously not be aluminum but would have to be some kind of plastic.. something non-conductive. Perhaps the entire thing could be made out of poly, would that cost less than the silicon mold?

      And with both the aluminum or poly, you could cut small channels in the surface if you really wanted the ribs.. just a lot more work than the "pour and go" of the silicon, but maybe a lot cheaper.

      Delete
  4. Hi Jack,

    Great show. Those a123 packs are really coming along. I do have one question regarding the old design with the threaded rod. You are still using cast resin, and now it is successful, do you still think it was the heat from the curing that messed up the threaded rod design?

    Thanks

    ReplyDelete
  5. Jack, just wanted to let you know I think the A123 cell packing is very important and very interesting. I look forward to every episode to see what you're doing with these cells.

    ReplyDelete
  6. Hi Jack,
    what a coincidence that you mention the Daniel A Cogswell and Martin Z. Bazant paper just a few days after Mindset of Swizerland translate their July 25th press release into english http://mindset.ch/en/wp-content/uploads/2012/01/PR_Akkutechnologie_engl..pdf , claiming twice the energy density of the Tesla roadster's pack, by an innovative charging method; or did you actually see this? Would be nice if Mindset wouldn't join the likes of Phoenix, Fisker, Think.
    Rgds
    Wolfgang

    ReplyDelete
    Replies
    1. 260 Wh/kg? Achieved through a "newly developed method of charging of pre-conditioned lithium ion cells".

      Maybe I could also pre-condition my 1 liter water bottle so it can hold 2 liters.

      Delete
    2. Sam, believe it or not cells need pre-stressing so there is a slight lithium plating. It's also done to ensure the cells function equally.

      More like altering your water bottles neck size

      Delete
    3. http://www.physorg.com/news187554124.html

      New charging method could greatly reduce battery recharge time


      http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=5684431

      Improving battery charging performance by using sinusoidal current charging with the minimum AC impedance frequency

      Delete
  7. Jack:
    To reduce costs, we need a less expensive material to contain the cells; how about a wooden or plastic box with slots, use the same method of terminal connections you use now and mold the tops and tabs, and the bottoms of the cells with plastic.

    Advantages: easy to replace cells; you can make any size battery desired; space between cells for cooling, expansion; less space used for the battery; no damage caused by the curing process and a lighter container. There may be more.

    ReplyDelete
    Replies
    1. Like the Croatian module? I wonder the same.

      Delete
  8. Hi Jack & Brian

    Dear Jack, I am convinced that you're dead wrong. I am sure that everyone is interested your development of the battery pack A123
    It is crucial how the batteries preforms, and the same applies charger. I'm glad to see it.

    With respect to your new instrumentation, I thought it looked really good, I might think that the Imprecision @ 4 - 5% is probable due to the user of a aluminum block instead for copper. Gauss field will not be the same, but as you also mentioned, calibration would solve the problem. one at 0 amp and one at 100 0 amp, and so on. this way you can get an OK linear

    Do you have the web address to the dealer of the instrumentation?
    I would like to know more about this equipment if possible :)

    Thanks for a great show

    Best Regards Allan

    ReplyDelete
    Replies
    1. http://www.emotorwerks.com/emw3/product/ev-dashboard-by-emw-basic-edition/

      Delete
    2. Many thanks for the info, it's great with an EV Community
      Many thanks :)

      Delete
  9. Keep up the good fight Jack. Remember the OBSESSIVE or a better term would be passionate people are the ones who succeed.

    Thanks for another GREAT Show!

    Chris

    ReplyDelete
    Replies
    1. Jack, I am very interested in the A123 battery project, watch every show and am a great fan of the "crowd source" energy that you have generated with EVTV. Please keep up the good work! Success seems close at hand and will be great for EV'rs all over the world!

      Delete
  10. Jack,

    Thanks for the mention. I feel you are absolutely correct that you don't need the Nordlocks at all with the way you are potting the thing. The bolt head and nut can't possibly move once it is potted so save the $4+ in Nordlocks per unit. Someone else mentioned tapping the terminal block, I am not sure this would reduce the cost. My one thought was to use steel for the clamp piece instead of aluminum since it seems like there might be an issue with the aluminum bending overtime due to thermal cycling since it is soft. If that happens you would lose the clamping force and resistance would go up causing a failure. The thicker terminal block carries all the current and the clamping blocks are just there to hold the tabs firmly against each other and the terminal block. A piece of steel could be thinner and still stronger than the aluminum. And while I haven't priced it probably cheaper as well.

    This looks like a very good package now! All the rest is just incremental improvements so it is easier to assemble, costs less, lighter weight and consistent size.

    I think this is very important from the standpoint of getting those people who are unwilling to spend the $8000 dollars on batteries to build a 100 mile range car. I went through mental turmoil and ended up with a 100AH 54 cell pack simply because I wanted the 800 amps that could be delivered by a 100AH cell. I didn't need that large a pack for the range but I was afraid to size it any smaller because I don't want to damage the batteries during hard acceleration. A 3P arrangement of these A123's would have done the job from that standpoint. I would probably do a 57S4P pack of A123's if I was doing it today. There is always the next car!

    Doug

    ReplyDelete
  11. Like everyone else, I love these battery trials and tribulations. I’m fascinated with these cells and look forward to the next episode.

    Jack keep up the great work, Brain keep the fire extinguisher handy!

    Charlie B.

    ReplyDelete
  12. Jack, I thought of a simple way to make all of your batteries exactly the same height AND keep the resin from covering the terminals. Simply drill out holes in both pieces of your polycarbonate to accommodate two sticks of your high-density polypropylene which will be bolted to the battery terminals. If you position the holes correctly, you can hang the cells from a precise height and move them to exactly the center of your casting space, perhaps using rubber bands or zip-ties to keep the battery cells generally pressed together. That way you can also eliminate the spacers that mar the appearance of the final unit. The holes will provide spillways for the resin, which you could collect pretty easily using cups molded from modeling clay stuck on below the holes. If you drill the holes in just the right spots, you can easily align the unit for precise repeatability by making sure the polycarbonate sheets sit firmly against the bench when you put the unit up vertically.

    ReplyDelete
  13. Jesus, Jack... I was cringing in my chair waiting for yet another shower of sparks as your knife bridged across the two terminals as you were trimming away the plastic. Please, for the sake of my heart, use a bit more caution when touching the sparky stuff on your batteries and your bench! :^)

    BTW, I'm just finishing up my 914 conversion and am using PakTrakrs and wrote an Android app for it. I'll be making the app free for download and open sourcing the code for others to take and improve on. My gauges are fully dynamic..., but the layout is fixed with SOC, voltage and temp, plus a bar graph along the bottom for current flow. Another tab shows the voltage on each battery and the main tab integrates the GPS and Google Maps to track your trip and show you how many miles you have left on your charge. Blog is at porsche914e.blogspot.com. Mike Browh

    ReplyDelete
  14. I like the idea of Jacks silicon mould (mold) material as a multiple cell carrier!

    After considering the box method I've gone off that. Instead, moulded (molded) strips made or cut to length as required.

    They can be siliconed end to end for wider packs. The cells retained in the same manner to stiffen the strips. It will keep them secure and light weight without incurring those thermal stresses which bothers me.

    ReplyDelete
  15. I like the Idea of a small pack that charges quickly, even at level two. I would happily settle for a 60ish km range, particularly if I could top up, or fully charge the car quickly from a regular plug (that’s 230 volt here in Ireland). That coupled with the knowledge that using the power and charging fast may not hurt the life of the cells are positives for me. I hope more of us feel the same way. Thank you Jack, for all your work, you may ignite the movement further than you think with this.

    Padraic

    ReplyDelete
  16. LocTite or epoxy would probably work as well as Nord-Locks in this application. Instead of casting, A123's stuck together with four big dollops of silicone adhesive in each corner and slightly wider ~.063" aluminum plates stuck to each end. Battery box wold be slotted to receive glued up packs and possibly equipped with fans for heating / cooling as needed. *Could* even be filled with a fluid other than air, with proper design considerations, for maintaining desired temperature.

    ReplyDelete
  17. Hello fellow ev enthusiasts! Thanks again Jack and Brain for all your efforts. I recently got a quote for the A123 cells at $20 without haggling, for 300-500 cells, and $19.50 for over 500 cells. I am too chicken to buy them yet, as I eagerly watch these episodes to see what may be possible with these cells. I would LOVE to have these cells in my 986 Boxster preferably without a BMS this time. I already love the car, even though it only has about 25-30 mile reliable range using K2 LiFePO4 cylindrical cells. Capacity seems more stable lately, charges pretty quick, and so I rarely drive my Camry hybrid anymore. If I could upgrade performance and range with these A123 cells, I would be completely happy! :)))

    ReplyDelete
  18. This comment has been removed by the author.

    ReplyDelete
  19. Jack & Brian,
    Very exciting progress with the A123 cells. I too as a few others considered having a level bar in the top in the mold or at the top of the mold to both get a consistent height and level of the pour up to the contacts. Also, do you think it is important to use different metals for the top contacts as the OEM do? I see copper and aluminium? for the top contacts in their pictures.
    I also have a question of anode & cathode. Is it true in these LifePo4 cells that when charging, the Anode becomes (+) and the Cathode becomes (-), but when discharging, the Anode becomes (-) and the Cathode becomes (+)? My experience with Vacuum tubes,diodes, and gas lasers; the Cathode is (-). But it seems in a galvanic cell or a rechargeable battery, the Cathode and Anode switch depending on charging and discharging. Still learning after so many years. Thanks for your excellent teaching Jack & Brian!

    Jim

    ReplyDelete
  20. Klaus,
    One downer about loctite, if it gets between the plates, it expands when set and becomes an insulator.

    They won't guarantee it for government work.

    ReplyDelete
  21. I can't speak for Klaus but I think he is saying the loctite is on the threads of the two bolts clamping the tabs together. I think that might be a good way to save a couple of bucks per each battery of cells.

    ReplyDelete
  22. Jack - was researching sources for the A123/20Ah cells and I came across a 40Ah LifePo4 prismatic pouch cell from Guangdong,China which ships from the ShenZhen port.
    Specs:
    1. Typical Capacity: 40Ah
    2. Cell Outer Dimension: 8.7x150x440mm
    3. Nominal Voltage: 3.2V
    4. End Voltage: 2.5V
    5. Charging Current (max.) : 1C(=40A)
    6. Discharging Current (Max.) : 2C(=80A)

    Minimum Order Quantity: 5 pieces

    Here's the link:
    http://www.alibaba.com/product-gs/390736256/3_2V_40Ah_High_capacity_LiFePO4.html

    ReplyDelete
  23. Loctite is fine used properly but for good reason its very rarely used on electrical assemblies.

    If the threads are going to end up potted then obviously there is no need to use anything.

    ReplyDelete
  24. Is there such a thing as nonconductive saran wrap? If so, why not wrap up these cells to immobilize them instead of using expensive resin and molds? Just a 3am idea after losing sleep over "The Girl with the Dragon Tattoo."

    ReplyDelete
  25. Good point on runny LocTite getting between the tabs, Andyj. Better then to stick with something more viscous like epoxy, even after tightening everything down to keep it from getting it between the tabs during assembly. Along a similar note, Jack mentioned that the 1/8" aluminum end pieces that pinched the tabs together bowed when he cranked down on the screws. Thicker end pieces or U or L sections would probably make for less bending = more uniform contact. A 1500+ amp draw test would confirm the design.

    ReplyDelete
  26. Interesting. We are developing something new here. Crowd sourcing? VideoSourcing? Video collaboration?

    In any event, some good ideas. I tried tapping the off plate yesterday and it worked pretty well. I got very good pressure on both plates. It didn't strip out at all. And it reduced the profile a bit.

    Saran wrap/shrink wrap. A couple of things here that apparently aren't getting across on video. This is TWO problems being addressed by one saran wrap/shrink wrap solution.

    First is isolating the pouches. They have SOME structure but not much. By the time you bolt even a light piece of aluminum on top, the tail is kind of wagging the dog. IN an automotive environment this WILL be DISASTROUS. Any wiggle on those tabs will develop a stress crack almost before we leave the drive way. This is WHY I'm big on the resin pours despite horrific expense at this point. They lock everything in place and EVERYTHING has to move as a unit.

    Puttting them into the mold, holding them together by shrink wrap is not a bad idea, but not really the problem. The problem is of course centering the battery cell in the mold where we get equal resin all around it. That's what the spacers are for - centering. They leave a uniform gap on both sides so we get a uniform layer of resin on both sides.

    Molding rubber and urethane resin are both horrendously expensive. OUr next mold will probably cost $300. Molds aren't bad if they are reused.

    The resin tends to be about $100 per gallon. My hope is that if we ever land on a design, we can start then to try different less expensive resins, including epoxy, maybe some kind of ABS, etc. And walk down the chain toward less expensive.

    The other element is the nature of these things tends to be quart = $50 / gallon =100 five gallon = 200 55 gallon = 1000. You get the idea. The cost goes down a right smart if you can buy in quantity. But the shelf life on these materials is very poor. So you have to order 55 gallons and USE 55 gallons right away.

    So we need to accurately host our batteries within a quality mold, and we absolutely need to cast them into SOMETHING that will isolate the relative movement of pouch cells and terminals.

    Our terminal design is evolving and looking pretty good at this point. This week I'm sliding them further down the tab and folding the tabs over the sides. It looks pretty good.

    Jack Rickard

    ReplyDelete
  27. This just in:

    A123 CEO says,“In general, batteries will come down by about half in price from 2010 to 2015,”

    http://www.electric-vehiclenews.com/2012/02/batteries-to-rival-gas-peaker-plants-by.html

    Has he seen your show? :-)

    ReplyDelete
    Replies
    1. Apparently the man is a prophet. Indeed, they appear to have come down by half between November of 2011 and February of 2012.

      Good show A123 CEO.

      I doubt he's a regular viewer. Which might explain his "Lost In Space" status in the world of electric cars. A $30 million investment in Fisker. Bet he wishes he had THAT back.....

      Jack Rickard

      Delete
  28. I wonder could you have the spacers moulded in? This would probable mean moulding a cell in plastic alone, and subsequently drilling in to it to make the protrusions in the mould you would make by using this to make a mould. A channel might have to be dug horizontally alone the sides and bottom (it may be better to use spacers here, longer ones). Maybe a precision fit and masking tape over the terminals could mean pouring from the smooth side and a more precise shape, with terminals more centred from cell to cell.
    There are problems with this. It means that there will be areas where the pouches will be exposed. This might not be such a problem, aesthetically at least. Also it may be more difficult for the cell to be removed from the mould.

    I hope you can understand my description.

    Padraic

    ReplyDelete
    Replies
    1. I do. You're making it a little harder than it is, but yes. That's the idea. IF we add a bit to the mold in some specific spots to support the cells, like at the bottom and across the back, we don't need the little plastic bits.

      I'm thinking perhaps some small bolts threaded through the rubber to reduce that profile even further. This would leave some small holes in the case but we could always fill them with resin later.

      jack Rickard

      Delete
    2. Perhaps toothpicks with markings on them could be used to center the cells. Have them half pushed in when you drop the cells in and then gently push in.
      After the pour take them out before fully set and the holes might even disappear.

      While you're at it, put your brand into the mould. EVTV on the sides would look nice.

      Delete
  29. Jack & Brian,

    I believe you shouldn't worry about losing audience due to shows about batteries. It is a wonderful scope. Everything else, concerning EV conversion is more or less off the shelf item. By all means, batteries were as well, until last december.
    You are diving into very exciting area. Reducing the cost of the most costly part is serious improvement of anybody's project. Someone else has been trying out as well, but your efforts and findings are shared.
    EVTV is genuine in it. Great chefs sell cookbooks while their recipes are purposely missing out an ingredient, or something is changed in order not to be reproduced.
    I am thrilled with the last episode, and from december on, show's quality is steadily rising, never mind technical glitches. Thanks a lot.

    ReplyDelete
    Replies
    1. We have had some technical glitches, but ok then. We'll continue to pursue it. I agree batteries are the key element of the cars. But they are inherently not very video exciting if you know what I mean. Watching one charge or discharge is not much fun and I cannot picture watching me with a bandsaw do BAD impressions of a machinist for comic relief are really much of an improvement.

      But the topic itself is of course of interest.

      It's just Amp Hour meters and drill presses aren't really very "visual."

      But I guess our show has never been about titty blond chics and aerial HD shots anyway. If it is, we've REALLY done a poor job of it.

      Jack

      Delete
  30. Obviously everyone will agree these tabs must not work harden through movement or vibration. The flip side is cooling, filler weight and the space it occupies.

    Jack, have you still got that opened up Thundersky to show us how the leaves are connected and what lived in there??

    I'm an expanding 'styrene foam guy when the will to think evades me. Nobody should "go there" with silicon oil bathed around the tabs and a water cooling pipe running through, lol

    Still liking the idea of casting 30 cell (~100V) wide strips by 1/2"thk out of a silicon rubber to make pack sides. All parts tacked with silicon adhesive. This way my cells can be added to or removed in parallel modules.

    The advantages are:
    Easy servicing/balancing.
    Light weight, acceleration or more capacity for touring. (Has little dead weight & CdA).

    Just need to grab some darned cells!

    ReplyDelete
  31. Here is a photo of my Hi-Power Cell. Pretty much like the TS but not the quality of the TS. They pretty much hang but the cell packs just rest on the bottom of the case for support. The cell packs fit the inside of the case perfectly so there is no wiggling and no hanging and no real movement. As built they are designed to last a long time.

    http://tinyurl.com/7ndptwr


    Go here for the rest of the photos: http://tinyurl.com/87d8tut

    ReplyDelete
  32. Yes Jack, bolts, a good idea.

    What about a tight fit lengthwise, with the terminals pushed up to the end of the mold, and with the terminals protected with tape?

    If this was easy, and possible, the lenght should be very similar for each mold.

    Padraic

    ReplyDelete
  33. Jack, I had a thought, not sure how good it is. Watching you pour the liquid into the mold and talking about trying to get the tops even made me wonder if this could be done upside down. If the mold is made such that cells go in upside down and are at the bottom with a couple of slightly recessed areas for the tops of the lugs to fit into and maybe a bolt hole for each lug where a bolt could be used to just snug the lugs down to the mold to prevent the liquid from getting in there, this would allow all the tops to be the same with the lugs just sticking out of the top of the plastic once completed and you could then just fill up from the bottom to a fill line. The completed ones that are a little too long, carry them to the band saw and take a little cut off the bottom provided there is enough material poured over the bottom of the cells to trim. I have no experience with molding so I don't know if this is even doable but I thought I would throw it out there and maybe someone can expand on it. I hope I explained this to where everyone can understand it.

    Thanks for a great show.

    DC

    ReplyDelete
    Replies
    1. Since your talking about casting them upside down I thought I'd throw out there something I have been thinking, and that is that to save money you might not actually have to cast the whole battery. If you say just cast the top 1/5th or 1/6th of the battery you have made it so that the tabs are protected, the aluminium blocks are secure and the cells are firmly joined to everything. You would have saved a huge amount of money. For the bottom of the battery that isn't covered you could use a cheaper type of plastic box or wrap it in something.

      I don't necessarily advocate this. It's just an idea to think about if later on the cost of the casting needs to be reduced.

      Delete
  34. Ecarzine, Pete?
    Thanks for the photo's matey, most appreciated.

    Not small bolts in there!

    Noted with interest the whole shebang is wrapped* and taped before being slid into the case.

    *dialectic sheet?

    ReplyDelete
  35. Andyj,

    It's Pete. I had my other name on and did not realize it until I posted the information. Ooooops. The sheets I suspect are fully dielectric and it also holds in the electrolyte. Almost no electrolyte was outside of those packages. This is what you'd find within the A123 Pouches except the pouch is in effect what that plastic is for the prismatic cells. I like the pouch because it can be vacuum packed and makes for a very tight and flat cell.

    Pete :)

    ReplyDelete
  36. No worries Pete,
    The electrolyte being the expensive bit, maybe not brimmed up to the bolts but enough to keep the plates wet if overstressed maybe?

    Have read something somewhere, the boxed cells are good at any angle except inverted. So I'm guessing there has to be a bit of it at the bottom. No problems with A123's on this matter which is great for differing fitments. ;)

    Kokams charge curve (lack of) is scarier at the top which is the bit that allows it more charge density... I'll be sticking with these.

    ReplyDelete
  37. I have a question:

    With the prismatic cells there is a hole in the top for gas to vent out if they are over charged. With Jacks first casting experiments the whole battery ended up cracking. Jack knows what he's doing and he's going to under charge these newly cast cells so there shouldn't be a problem, but what happens if they are over charged by accident? Will they crack or explode? It sounds like from the interest in this that several people are going to copy this design when it is perfected.

    I was wondering if it would be possible to mimic how the prismatics are designed so that there is a pressure relief hole in the top? If you had a small area near the middle of the battery where the cells were not cast and then had a route for the gasses to flow to a hole in the top, would that then stop them cracking or exploding if over charged?

    ReplyDelete
  38. And now for something completely different.

    What do you guys think about standardizing on one unit to indicate vehicle fuel efficiency.

    I would propose this unit to be Mega Joule per mile.

    It makes the most sense because it breaks things down to simplest components in terms of energy per distance.

    It also puts EVs, hybrids, gasoline, diesel and every other type of transportation on the same unit, one that makes sense.

    For example the Nissan Leaf has a 24kwh battery and an EPA rage of 73miles.

    So 0.33kwh/mile or

    1.2MJ/mile on the EPA test cycle.

    ReplyDelete
  39. Hi Jack,
    I have reservations about "hard casting" after reading the following article:

    http://144.206.159.178/FT/641/92454/1607530.pdf

    In this article they tested 4 different mfgs 383562 polymer cells for changes in size during charge and discharge. Though these are not LiFepo4 cells they did change in size. They increased on charging and decreased on discharge. This was in relation to the increase and decrease in voltage.

    Snip from the article conclusion:

    Regarding battery enclosure, most surprisingly
    pouch cells gave much smaller thickness change than the metal case cells. This volume expansion phenomenon must be taken into account for power pack designs, modeling studies, and especially for the designs of larger modules where numerous individual cells are stacked and connected.

    I think that maybe you are trying to rediscover the wheel on the packaging issue. If we can find a "right sized" case, you can then assemble the A123 pouches the same as the Thunderskys.

    These folks sell at least one size case:

    http://www.alibaba.com/product-gs/458776451/Car_Battery_Box.html

    My email to them asking sizes and cost is unanswered at this time.

    It might also be possible to cast with a more flexible material.

    Karl

    ReplyDelete
  40. We know we are trying to reinvent the wheel. A123 already has a battery module design Karl. So no news there.

    The problem is that we hAVE to isolate the pouches and tabs from vehicle movement. They will wear stress cracks into the aluminum in HOURS, probably NOT days. LiFePo4 cells are entirely different from the LiPolymer cells you are referring to. That is not to say they do not expand under certain circumstances. Clearly we did SOMETHING wrong with our 40v module.

    Tomorrow I am going to cast 24 cells into a single 12v mold. We'll see how that works out. So far, the six cells in a single unit don't seem to struggle. We're getting full capacity and repeatedly charging and discharging them. They stay REMARKABLY cool even under a couple hundred amps charge or discharge. At 600 amps my test equipment starts to come unravelled.

    So expansion is one issue. But the wear from relative motion between cells and tabs is much more serious at the moment. We have to isolate that or these cells will be useless in a vehicle system. In fact, looking at A123's very professional pack I am unconvinced they have this solved long term.

    Jack Rickard

    ReplyDelete
    Replies
    1. Jack,
      Design issues presented at his point with the assembly of the A123 20ah packs.
      1.The individual cells need to be bonded together to prevent flexing which will eventually fatigue the aluminum tabs of the cell which will eventually cause failure.
      2. The cells expand and contract in size during charge and discharge cycles.

      What I would suggest at this point which would not change the current casting process is change the formulation of the urethane casting resin to become more elastomeric in the form of a high density
      flexible urethane foam...http://www.smooth-on.com/Rigid-and-Flexible/c10_1121/index.html

      Delete
  41. Jack:
    I would like to see you design for the worst case; I see that as when you have a cell that expands and bursts or even a group of cells that expand and burst. You must allow for that particular action by providing space in the enclosure else you are building a bomb! Is that why Thunder sky has a venting system? You bet ya'.

    I for one don't think you need to seal the cells completely in plastic. I go back to the idea of building an non-metal box and fitting each cell in a slot, then fill the bottom up about a half inch with plastic to secure the bottom of the cells, leave the middles of the cells open to the air, and finally clamp the tabs to create terminals and as a last step mold in the terminals, tabs and cell tops down about an inch onto the cell body to secure the tabs from movement...less weight, less cost, more easily repaired,etc.,...and room for the worse case condition, thus no booms!

    ReplyDelete
  42. Well put everyone. I find it amazing how so many of the ideas Pete and I have been emailing to each other are coming out through this discussion.

    You'll note that my urethane rubber casting YouTube videos demonstrate the "hanging from the terminals" concept. And yes, I did say rubber. I am leery about having too hard of casting. Yes, it clearly will shatter if expanded too much. Something on the shore A scale will be much less likely to do so. And it will still provide that kinetic unity that protects against metal fatigue in the tabs.

    The idea of using a much less dense core filler has come to mind. However, as a 'one off' fabricator by trade, I find it hard to balance time against the additional steps and come out overall cheaper than a single step process that has been designed to reduce material consumption overall.

    The balance I foresee is micro modules, of three series. The carryover of 12v units is dubious in merit if they are ultimately to be used in traction packs. The incremental additional flexibility in general pack shape afforded by individual blocks being just that much smaller cannot be understated. Additionally, a 25% per block cost reduction makes the risk of loss a bit more tolerable.

    Ultimately, single cell blocks will be a footnote. There are 60-75% fewer interconnect straps needed, and just as many fewer opportunities for corrosion when the majority of cell to cell connections are embedded in a potting compound. Potting compound itself is reduced by elimating the redundant encasement of each cell as well, benefitting both volumetricly and economicly.

    I can foresee incremental upgrades in total kwh capacity through stepping up voltage too. The big kicker will remain - how man ah wide should the pack be? Nominally 60, 80, 100, 120? After 80, the gains in power are immaterial by virtue of overkill. So then, what is an economically reasonable balance? 200ah wide and you are no longer advantageous. I'm designing around 5 cells for my super beetle; anticipating performance gains through lightness and freedom to increment the voltage on a dc system.

    ReplyDelete
    Replies
    1. Nabil,
      I see more use in the freedom to increment the current capacity on a dc system.

      At the end of the day if you change the voltage, you'll have to change just about everything else.

      I know, it means a lot more connections to do it properly but for bottom balanced packs is it a problem just to parallel? ;)

      After all these cells can take an abusive current with less voltage drop for short trips.

      Delete
    2. A very interesting question. I've been working with 6p. In theory, 120 Ah but in practice, 115 to 117.5 Ah. The concept is to BE over 100Ah.

      The power output is of course as you note, overkill at 2750 theoretically and certainly 2300 amps at the low end.

      On the eCobra, 1400 amps was our theoretical limit, although I would note that LESS sag is MORE power to the wheels whatever the level.

      I also remain convinced that your power output as a percentage of what your power output COULD BE remains a factory in cell life. I could be wrong and indeed there is a paper out there now that indicates HARDER discharges are actually a factor in LENGTHENING the life of these cells.

      But I like the idea of a cell designed by spec to do 2000 amps and I'm going to use it to do 1000 amps max.

      The big thing for me is that there is a breakpoint where a smaller number of cells leads to a very different device to connect them than what I'm using at the moment. It works for six cells. Probably WOULD work for 4 or 5 cells. Probably would NOT work for 3 cells.

      Jack Rickard

      Delete
    3. Our latest will be a 4s6p 24 cell unit for 13.2v - essentially a 12v starter battery. Why would anyone pay $800 for a starter battery? Well, 2700 cold cranking amps? 30 lbs? 20 year life?

      Not certain. But the dubious merit of continuing the 12v tradition is more of an unknown merit. When I see things like that, I go to a belief system in a darwinian evolution of technology. We've had cars for over 100 years. Their batteries are 12 v. Everyone knows 24v would be more efficient and allow lighter more flexible wiring. We started with 6 v. But we are very much stuck on 12v. There is probably a reason. And my knowing or not knowing the reason matters not a whit. The reason stands whether known or not.

      I'm willing to go against such, obviously. But where x = don't care, I usually opt for the extant.

      Jack Rickard

      Delete
    4. 10 or 15 years back it looked like there was a push in the automotive industry to switch to 40 or 48v starter batteries. The advantages would have been a smaller battery, a smaller and lighter starter motor and a smaller and lighter alternator. And all the wiring in the vehicle could have been lighter. I don't know why it didn't happen. I do know that there were problems with headlamps when the switch was made from 6v to 12v. The 12v filaments were thinner and more fragile so the bulbs failed from vibration. It was quickly resolved with a different filament material. It may be that it was just impractical to change everything all at once and the inclusion of a 40A DC/DC converter to get 12v made the whole thing too expensive to go into production for the gains.

      Doug

      Delete
    5. Destabilizing [to corporate control and profits] is probably a better word than impractical. Watch Comprar, Tirar, Comprar [The Light Bulb Conspiracy] for supporting framework:
      http://youtu.be/AkBQKSlsNhU

      Delete
  43. I like your choice of material.

    http://www.youtube.com/user/nabilahmad

    ReplyDelete
  44. If you build them like TS and just have your cells hang they must be done so they fit snugly on all sides and bottom and that the cells are not hanging their weight on the tabs but resting on the bottom like the currently used prismatic cells. You'd need to build a system that will allow precise placement each and every time so zero stress is on the tabs and each and every cell is exactly the same size and that you can do this all by hand. It would also be good to seal the tabs once all the connections are made to help prevent any corrosion of the tabs within the cell. It's not so easy to duplicate in the home garage. A vent would be a good idea but it would have to be a vent that would not allow water or moisture to gain access to the inside of the box to prevent corrosion. An open vent would not be a good idea even if the cells are sealed. I also don't think that a vent system with these cells will be required because they are fully sealed cells unlike the cell packs within the TS case.

    Building a form case and encasing them in softer rubber is an excellent idea and I am thinking of even going as small as 4 cells. If you then need more AH's then just buddy pair them. I propose a very large terminal lug per pack of cells so when buddy pairing them you will not have a current issue across the terminals. This opens up the possibility of even smaller packs for those than want them and if you do happen to loose a cell you are only out 4 pouches rather than 6 or more as proposed. More terminal connections but with the current way to build flexible connections that are going to be tight and secure it is really not an issue except we need to build terminal straps to accommodate the new size cells.

    The cells will remain protected within the rubber and the urethane rubber will also provide a bit of shock protection in the event of an accident and even just normal road shock. Knowing that the urethane plastic is doing just fine with Jacks 6 pack I am sure the urethane rubber will do just fine with a 4 pack.

    Building the form is them hardest if you don't have proper shop tools. My Hi-Power cells won't fit 6 pouches deep and the width is almost to narrow. Height is fine.

    Pete :)

    ReplyDelete
  45. Jack,
    Your EV Android dashboard: Good roadtest!
    If the fella that makes it heeds your findings he will eventually end up with a great "Killer app".

    I'd love to see that Cycle Analyst you have with a back to back against your equipment Sometime if you wish.

    I know its quite a cheap "Jack of all trades" with the odd glitch here and there but I'm taken with the serial output for a nice screen output, error correction in software and logging.

    Example in software used to cleverly correct any unwanted readings:

    http://whatreallyhappened.com/WRHARTICLES/The%202012%20SC%20GOP%20Primary.pdf

    Sorry for that - at least flip-flop voting is not weather related. >:-))

    ReplyDelete
  46. My dumb idea #635... Cast the tab end in a plastic box with the onyx resin, leaving the terminals exposed. Fill most of the remainder with candle wax (unscented, let's not be silly), then seal the bottom with resin (or perhaps epoxy) to complete the cell. The end product would be similar to prismatic cells. The wax should give a little in case of expansion and help save on materials.

    ReplyDelete
    Replies
    1. DR. Righteous,
      To go with your idea and further refine it. It is possible to purchase "adhesive tooling wax" in various thickness to put between the individual A123 pouches to allow for expansion and contraction of the cells and then a less expensive casting resin could be used. http://www.freemansupply.com/brochures/sheetwax.pdf
      Mark

      Delete
  47. I too LOVE the Android App application for Electrics, which I had hoped someone would develop. And Nabil, I would think that even if you have a hard rubber that has a shock protecting quality to it, then it can transfer vibration and movement, albeit small, I would fear in the long run, would be detrimental to your internal pouch connections and may fracture the tabs.
    But I guess we won't know until there's some real world testing of your enclosure. I hope to see your version put to the test.
    Love this blog! Much appreciated!

    ReplyDelete
  48. I will try to put together in writing a collective solution, where I pinched some from lots of ideas shared in this article's comments. Four pouch cells per battery. Off the shelf battery case, with removable top, with big vent and terminal holes. A T shaped high molecular weight vinyl stand, screwed together, similar to a table flag pole. Its base in the shape of the casing's bottom, rectangular, just a shy smaller providing a tight fit. The "pole", a rectangular strip, or perhaps better a beam.
    My picturing changes as I am putting it down in writing :)
    Its upper part could mount a horizontal strip (beam), almost as wide as the casing. in order to serve as a shoulder for connected tabs. Maybe even cross beams on both sides, to impede swinging motion of the shoulders inside the casing, of the length providing a close fit.
    Once the pouches tabs are clipped together, with stainless steel brackets using nordlocks on both sides (under the bolt and nut), they will be mounted on top of this cross stand. The pouches taped tight with stronger scotch, few strips in order to be inserted smoothly in the plastic casing, as a block. If casing results much deeper than the pouches bottom, the filler will provide they no longer hang from the tabs. Here is where branching might occur.
    This battery could become serviceable or not, depending on the filler. It shouldn't be very expensive either way, because the casing helps the filler economy. Casing material is thin but rigid. It no longer matter if and how much of it is between the pouches and the casing. No spacers needed. Filler could be non destructible, removable agent, a foam, a flower like inert solid or wax as suggested. Even combination of fillers, whatever proves more economical and in the same time safe.
    Scenting might not be such a bad idea, provided a compound that would be released at certain temperature, indicating a bad cell. Training a dog to find it, if to big battery pack, shouldn't be much of a trouble ;)
    This modular unit could perhaps be thin enough to cover the vehicle's floor in extractable trays. These could be serviced, standardized and used as ammunition clips. Vehicle should be provided with several bays, for easier manipulation of lighter units and for range extending. It would be lighter as a city commuter and completely armed for longer trips.

    ReplyDelete
    Replies
    1. Possible candidate as a filler agent might be paraffin. Second paragraph under title paraffin wax, stresses it out as next to the best isolator there is, second only to teflon. A quick swoop over net for paraffin wax pricing leads me to jump into conclusion, this might be interesting solution.
      The problem might be in its melting point, possibly detrimental for pouches' content. Maybe at bottom to fix in place T-stand base, fill the rest with talc, and seal it up with paraffin wax again. Talc is the softest mineral in the nature, so it cannot possibly harm the pouches. It is light and electrically non-conductive. Talc's price, for equivalent to what the cost was in the latest experiment per mold, FOB China gives a 1 ton of talc!

      Delete
  49. We have been playing a little to devise a sketch, Albert Bel and I, and here it is, illustrating better part of the above description.

    ReplyDelete
  50. Paraffin/wax. This is how to make my car into a candle. Hopefully not a Roman candle. In any event, they are flammable and petroleum based.

    The scenting idea is quite original. Electric Vehicle Aroma Therapy. Let's cogitate on that one for a while.

    Jack Rickard

    ReplyDelete
    Replies
    1. Roche Diagnostics stresses out pure paraffin as non-inflammable liquid. Oils from waxes are different story. Maybe we'll get to the point of mixing paraffin with additives preserving the non-conductive qualities it has. I accept your refusal of the proposal. What I certainly haven't had in mind is hurting anybody with it, not even accidentally.

      Coating of the wires is petroleum based, as well as crock shoes. In electric cars we will not get rid of plastic, which is a petroleum based, as pretty much everything around us, from a ball pen, to a wind-breaker jacket, wall-to-wall carpets to a boat.

      New materials are largely in our hands, thanks to petroleum. Lighter, alternative vehicular chassis are based on carbon fiber, derived from side products of oil drilling. This industry will not vanish, even if we stop burning immediately oil as fuel.

      My effort was to put together, picturing, what I understood from this brainstorming. This collective idea of battery assembling I presented, turns cheaper your latest 3 from the gallon design, as well.

      Looking for alternative fillers is meaningful, because your findings are, fillers could surpass in price the guts of the battery. As well, to immobilize and make a block the whole unit is the drive, never setting your car on fire, or anybody's.

      There is a branching point, where someone might prefer getting to the cell/pouch that goes bad, and eventually, one will sooner than the other.

      If dimensions of the off-shelf casing permits clipped together pouches on a T-stand, this model should work pouring your resin in it, resulting in non-reversible and non-serviceable agglutination of the box's content. Less resin usage will be detectable due to the box itself and a T-stand. If instead of the pole it is sheet like stand, stretching almost the whole width of the box that much resin will be saved, volumetrically and in dollars.

      Consider proposed sketch, a model for packaging of A-123 pouch cells, even if the paraffin/wax filler is discarded. One might like a whisky chilled with ice-cubes, but not as much having it diluted, when they melt. How about arriving to the point where battery box filler is solid when needed, with batteries in use and not, when battery needs service?!

      This way, you get the pouches "casted in amber" for their indestructibility and yet possibility to prolong or save the investment, when intervention is needed. Depleted cells are hard to match, I have learned that, but if only one is out of balance, the rest could still have use, even if taken out of the car's battery pack.

      Scenting has intriguing aspect indeed. Due to volatile compounds. Some among them, like amber, are inflammable ;)

      Here, I typed myself smart. The issue was, to try to save what I think is still worth in the proposal.

      Delete
  51. I think we are all thinking too hard.

    Pete earlier showed that normal prismatics have nothing between the cell packs and the clamps. They simply ensure nothing moves. A tough plastic box and lid.

    How about 4mm Correx for the box? simply slit every 4th Correx cell inside so the welded edges of the A123 cell is held in the cut slots. This gives a 1.5mm (1/16") air gap between cells.
    Of course the box will be bent for at least two sides and a bottom. Hot glue any loose ends.
    Same across the top if you want, then attach Dr. Righteous' top end assembly to save on dead weight and mat'ls!

    I'd give it a go.

    ReplyDelete
  52. I wouldn't worry too much about the flammability of paraffin wax. Even without additives the flash point is 390F, which would probably be fatal to the cells anyway.

    There is one problem I hadn't considered tho. Wax expands quite a bit when heated. That could be an issue in a sealed up cell.

    - Doc

    ReplyDelete
  53. The 25KV high voltage gubbins on electron beam welders are dumped into a big tub of pure paraffin. I know, I was working with them when one couldn't get a load up without tripping. That would be a >3" spark flying through the paraffin. Perfectly safe tho...

    Jacks point is turning the car into a candle if the worst happens. It could be scented so

    ReplyDelete
  54. Andyj:

    We may not be thinking hard enough. The analogy with existing prismatics breaks down on a couple of fronts.

    Physically, we have a 7.5 mm pouch cell wagging a very thin tab. In the Prismatics, we have hundreds of very think foils top to bottom AND so many of them that they are completely supported by the battery case sides. They are also wrapped and set inside a UHMWP sleeve.

    So apples and oranges on the isolating the tabs to prevent stress cracks.

    Second, the foils are directly exposed to electrolyte and so are the terminals. This is what you have a copper terminal on the copper foils and aluminum terminals on aluminum foils. In presence of electrolyte, dissimilar metal galavanics is inevitable.

    In purse the pouches are sealed and ALL electrolyte is in the pouches. So noe is on the terminals.

    Jack Rickard

    ReplyDelete
  55. $9.50 a cell. This american in china is posting asking if he should consider this over lead acid batteries for his bike.

    Wow.

    http://www.diyelectriccar.com/forums/showthread.php/looking-replace-upgrade-batteries-could-use-69796.html

    ReplyDelete
    Replies
    1. Batteries dropping to half the price during 2012… said the A-123 official the other day. This is an excellent news drgrieve.

      Thanks for the link. Even Google Translate was enough to amaze me with 59 yuan as a price tag per pouch, and no mention of minimal order.

      Aside form use in vehicles, I believe interesting laptop energy survival unit for field trips can be devised from these cells. Similarly for camcorders, since today's high power bricks are not exactly tiny. At least slick designs are possible with these pouches.

      This news might become a game changer for shoe string budget projects particularly. But if this is viable, than EV conversion becomes a small personal loan for a bunch of people usually gathering for a bear over a soccer game. EV conversion enters the range of common car tuning.

      From now on the battery assembly from pouch cells gets the highest priority. Even if batteries seize to be the most costly part of the EV conversion, they will continue being its decisive item, considering contribution to a vehicular weight and range autonomy.

      I am astonished with the prospects.

      Delete
    2. The tabs have been cut off of those cells.

      - Doc

      Delete
    3. Oh, it was sounding too good to be true having prices in a free fall.

      Delete
  56. Maybe I'm a sandwich short of a picnic here, but if we aren't doing BMS control at cell level, can we expolit the form factor of these cells to do a very large flat battery (~10 ft by ~4 ft by 1.5") that goes under the floor pan of the vehicle containing (say) 10 - 20 cell series strings of batteries connected at the ends in whatever S/P combination met your voltage/current targets? With no load on the terminals, a broad flat rectangular plastic duct containing a string of cells would be doable if you could come up with light connections

    ReplyDelete
    Replies
    1. I like this idea John.

      So modular flat slams you build up in layers under the car to get the amps you want. If you need more voltage you increase the area of each slab like adding pieces to a jigsaw puzzle. Sounds great.

      Delete
    2. Thank you Nickkk - precisely.

      Plus if properly designed much of the stress concentration issue should go away

      Delete
  57. Jack, Had to read your response twice to find the issue.

    I'm not in argument and saying prismatics rattle around and fully aware how they exist in their artificial environment due to dry assembly and the electrolyte stopping rot from dissimilar metals. like the insides of a car engine. Otherwise they wouldn't last a week!

    What I proposed did not preclude say dabs of silicon rtv (bath sealant) to hold the body of the cell and kill vibrations between and around the edges. Nor did it preclude a cast top end which I fully endorsed earlier. Just a means to save on adding too much weight and expense overall.

    Horses for courses eh? ;)

    ReplyDelete
  58. Hey Jack!

    One of your best shows. Love the spontaneity of clamps falling off your mold just at the moment of truth, and circut boards coming apart in the middle of a demonstration. Your "grace" under pressure is impressive.

    Regarding the contact of the tabs in your clamping arrangement, if you make the clamping side of your clamping plates a little convex on your bench grinder, the clamping plates will flex as they clamp, and create pressure along the whole tab stack.

    Mike Kaindl

    ReplyDelete
  59. One thought I've lately, is to use the suggestion of casting the cells terminal side down and only pot the terminals to the pouches, to completely cover the tabs. Tape the pouches to each other in bundles, then to fill out the blocks, rather than parafin, use expanding foam. Very light, electrically insulative, pretty cheap. The pitfalls I see are long term robustness against mechanical shock and vibration, and then heat during expansion and cure.

    ReplyDelete
    Replies
    1. I find expanding foam quite interesting… worth studying.

      Delete
  60. Dear Jack:

    You have NOT lost a viewer here. Personally I'm very intrigued with the A123 batteries. It's nice that you have the technical knowledge and resources to perform the experiments I could only dream of. And you're willing to share this with anyone willing to listen.

    The A123 cells look like an ideal solution to powering my dream EV. The only problem is the package is not usable in an automobile. Too bad A123 is such a stinker when it comes to sales especially their pack units. It says "Modules are not available for consumer use or aftermarket conversion kits." right at the bottom of the web page. I do find it annoying they are like this considering how much of my tax dollars wound up in their pockets.

    ReplyDelete
    Replies
    1. Randy,

      This site is advertising A123 Packs with a minimum quantity of 1.

      http://www.alibaba.com/product-gs/509102505/A123_Systems_Prismatic_Pouch_20AH_AMP20M1HD.html

      Delete
    2. CZTREE,
      Last week I emailed them to get a price quote for 300+ of the A123/20Ah cells and got a reply back on the 15th this month giving me a price quote for 300-cells at $17.30/each, but they did not give a shipping cost. Maybe Jack could give an estimate on shipping 300-cells to the U.S. Also I was told not to give this price quote on blogs and that they were not making a profit, just trying to get rid of their stock...yea right!

      Delete
  61. James,

    They show some packs on their site, I wonder how much they would charge to make up packs using the A123 cells? By the time you buy the aluminum, bolts & nord locks, make the mold and put in the potting, their pre-made packs may be a good deal. I sent an inquiry to get pricing on 5 24 volt packs so it will be interesting what they come back with.

    Regards,

    Randy

    ReplyDelete
    Replies
    1. Randy,
      Some of the ways they connect their cells do not seem to be the best way to me. Sometimes it looks like not enough surface contact area at the connection points. Maximum surface area connections is what is best in my humble opinion.

      Regards,
      Jim

      Delete
  62. James,

    They had a picture of the A123 pack on their Alibaba page and although it is probably wishful thinking, was hoping maybe they were contracted to make battery backs for the Fisker and were stuck with them after the meltdown.

    ReplyDelete
  63. Jack,
    Have you seen the article from the New York Times from Aug 2011 on the manufacture of the A123 Prismatic cells? Here's a link: http://www.nytimes.com/2011/08/28/magazine/does-america-need-manufacturing.html?pagewanted=all

    Some of the quotes from this article are:

    "In 2009, the U.S. made less than 2 percent of the world’s lithium-ion batteries. By 2015, the Department of Energy projects that, thanks mostly to the government’s recent largess, the United States will have the capacity to produce 40 percent of them."

    "A123 is also the first large-scale lithium-ion manufacturer whose domestic operations are up and running, though its pedigree is international. Its battery technology was developed at M.I.T., and for the last several years, the company had been making its lithium-ion cells in factories in Korea and China. When I asked Jason Forcier, the head of A123’s automotive division, why the company went to Asia to make its products, Forcier said he had no choice. “That’s where the supply base was,” he said. “That’s where the know-how was — it was nonexistent in the U.S.”

    “We bought a company in Korea that had the technology around this type of battery and had developed the manufacturing process there. We basically brought that here, copied it exactly and scaled it up.” A123 also brought a team of six Korean engineers to help transfer the technology to the U.S. and sent a team of Americans to Korea to learn."

    So it seems that the Korean plant and the Chinese were both manufacturing the A123 cells before the $375 Million grant to manufacture these cells here in the U.S. when Korean Engineers were brought here to manufacture them for the Fisker Karma EV as part of the plan using the grant money. Still poor planning to manufacture a high end EV in a bad economy, and I'm not surprised at the outcome.
    So are these A123 20Ah Prismatic cells a year or two old now that the Chinese are selling?

    Jim

    ReplyDelete
  64. Jim:

    You know, I just don't know. No, I don't think these cells are old at all. I think they are a few weeks old. I know A123 was gunning up to produce cells at a rate to support 15,000 Karma's a year.

    The Karma's problem wasn't timing in a poor economy, although that certainly doesn't help. The missed introduction every year I can recall. When they do finally do a car, it gets 20 mpg after the electric drive part. Who among tgeh green is going to SEE that as GREEN? ANd the price was steadily increased from $80K to $106K. Model S is no imminent, with full electric, 300 mile range, basically same profile, what is the attraction to this Karma? NOthing. It's not even made in the U.S.

    You are quite correct. The Chinese KEEP insisting that these cells really ARE made in the USA. I just cannot get my mind around the possibility that they are shipping them to Hong Kong, we are buying them for $17 a piece, and paying to ship them BACK, when they originate in Livonia. I could drive up and pick them up. These guys CANNOT be that lost. And so I insist and assume that they are made in Korea and they print MADE IN USA there. That's probably because I cannot face the horror of the reality of this.

    IF they accept $249 million in tax money, THEN send these cells to Hong Kong, where I buy them and ship them back because they won't' sell them to me at the same price they sell them to the Chinese for, then somebody needs to hang by the neck until they are dead, dead, dead.

    Speaking of which. Does anyone know what happened to Manzanita Micro? Their web site appears sickly today.

    Jack Rickard

    ReplyDelete
    Replies
    1. Jack what can be expected in shipping costs say to your place of residence for 300+ of the A123/20Ah pouches? I got a crazy shipping quote of $1500, which is absurd to me.

      Jim

      Delete
    2. Maybe they have a lot of stock in the shipping industry..

      Delete
  65. The horror of the reality of this!

    The elephant is feeling a little rotten.

    ReplyDelete
  66. Manzanita's site site is fully registered to 2018. A phone call should find out whats up. According to Google it was ok two days ago.
    If anyone needs...
    Phone: 360-297-1660 Fax: 360-297-1905. ATTN: Clarice Matz PO Box 1774. Kingston, WA 98346 ..
    ========================================
    Nenad's 59 Yuan (~$9.40) cells with the tabs cut off.
    The website says they should all be 14 to 18 AH.
    If you can cut back enough for a good connection without making the pouches leaky then they could be an absolute bargain when paired up!

    I'm confused now.. Champagne or cider. What's it to be?
    ========================================
    Anyone sold on the idea of LiFepo4 pouch cells but need 100 AH blocks and don't care for huge 'C' rates then for interests sake Google this: "GEB23310260". 1' x 0.9" x 8.5"+1.5" tall tabs. Some sell them in 1 off's.

    ReplyDelete
    Replies
    1. Alibaba.com brings range of offers from mainly three manufacturers: Shenzhen Victpower Technology Co. Ltd., Shenzhen OSN Power Tech and Shenzhen's as well General Electronics Battery Co. Ltd. Their prices drop as low as $2 - 10 / piece, for A123 pouch prismatic cell. I wrote Ms. Alice Chen, last night to inquire, since minimal order was 1000, and before getting the reply, an ad changed to a 100 as minimal order. Some of their other ads state Made in Korea, although Shenzhen is 20 miles from Hong Kong, a port all three of the above mentioned companies ship from. All pouch cells at Alibaba.com have tabs, though price reaches as low as those crippled ones from Taobao.com still in their beta. If Ms. Chen replies I will post again, because something strange is going on with A123 product on the market. Prices seem to be in a free fall. Rather fishy business.

      Delete
  67. Breaking news!
    This last battery I just mentioned about...
    One line of warning on it states:
    "Do not disassemble, cast, puncture or crush the battery".

    Oh!

    ReplyDelete
  68. If the cells are truly made in America, Who is up for Picketing the factory to force them to sell their product to Americans, when it is our tax dollars which made it possible for them to have the factory here?

    ReplyDelete
    Replies
    1. Sign me up. I am tired my monthly reminder letter to my Senator, who is clueless, about A123 not being an American company and only being able to purchase their cells from China. It looks like the Dragon we fed is about to Eat us. But the few, 2 to 5% that profited from off shoring our manufacturing capability are rich and hopefully dead by now.

      Delete
    2. And if their response is fine, no problem. Cells are $145 each. Line forms to the rear....

      Jack Rickard

      Delete
  69. I am going to construct a 13.2V/60Ah backup battery, which will give me an initial experience at constructing one of these LifePo4 batteries similar to what Jack is doing,but it will not be for EV use, just around the home backup power in the event of a storm. So I just bought 12-each of the "A123 LiFePO4 Prismatic APP72161227 PHEV/EV/E-REV 20Ah cells" which do not have the high 30C-Continuous rate of discharge of the "A123 Systems Prismatic Pouch 20AH AMP20M1HD-A Cells", but instead have a continuous discharge rate of 3C with a 10C/10secs max discharge rate. I ordered them from an ebay seller named "victoriz" who had them priced at $34.75/each with shipping at $14.95-EACH from California. So I wrote him and asked if I could get them at $30/each and told him he could ship all 12 of them for about the same price. So he wrote back and said he'd ship all 12 of them for $30/each & Free Shipping. I could have paid the same $1.50/Ah or $30/20Ah cell combined price + Shipping for the 30C cells from China in this small quantity(12-each),but this was quicker and easier. But I'll go with easier this time around.

    Jim

    ReplyDelete
  70. I believe I paid $1265 to ship 300 cells plus $285 in Paypal charges.

    If that helps.

    Jack Rickard

    ReplyDelete
    Replies
    1. Thanks Jack. I found an earlier post of yours quoting shipping costs, and I guess I missed it,so apparently their shipping cost to me for 360 of their A123/20Ah AMP20M1HD-A Prismatic cells is the going rate. So I was wrong in my impression of the shipping costs to NW Florida. So yes, that helps,but I'm not yet able to start an EV project. Sure hope to one day sooner rather than later.
      Thanks again Jack, you're the best!

      Jim

      Delete
  71. 180 A123's shipped to the UK, $795 plus Paypal $166 = ~$23.93pc.
    Obviously before 2% import duty & EU VAT (trade extortion) of 20%.

    Jack, you made a salient point with pressing too hard on the sellers. They are under a lot of pressure. I'm content with this price.
    ================================
    If this Iran business gets nasty the tipping point might come a lot sooner and faster. Who hates ya and loves ya baby... Irony.
    Iran has just decided to pre-empt non-delivery of oil to the EU, now! This means fuel prices are going to start screaming for the sky.
    ================================
    Oh Jack, that's a huge bulk of plastic you are adding. no wonder they say they should not be cast... It will crush the fish...

    ReplyDelete
  72. This comment has been removed by the author.

    ReplyDelete
  73. I posted this link before in this thread. After reading all preceding comments I thought it would be wise to post again.
    Changing the formulation of the urethane casting resin to a more flexible high density foam would allow for flexibility and adhesion of the pouch cells.
    http://www.smooth-on.com/Rigid-and-Flexible/c10_1121/index.html

    ReplyDelete
    Replies
    1. No reply suggests itself.

      We'll watch and see if you have something to add that might make this a more interesting possibility.

      Jack Rickard

      Delete
    2. Jack,
      May I suggest nobody responded because they did not go to the link and if they did they did not understand what they were looking at.
      These cast-able thermoset 2-part self-skinning foams can be formulated from the lightest 2 pounds per cubic foot density all the way to the heaviest density of the urethane foam which has minimal entrapped air cells, I would suggest that this resin would have the same density of the cast-able urethane that you are now using.
      It all depends on the formulation on how much foaming agent is added. Determining the proper density foam will take some consideration on the the pouch cells expansion properties.
      If the hardness of the outside shell is a design consideration, the SiO2 mold could be eliminated and instead place the assembled pouches in a custom size container with a lid and then pour in the cast-able urethane foam to bond the container and the pouch assembly together.
      Also, whatever happened to the idea of thin aluminum sheets between the individual pouches for cooling purposes?
      I have more thought too.

      Delete
  74. Mark, when people don't respond to ideas. It doesn't mean nobody is listening unless someone turns up later with a brilliant old idea.

    I suggested a "reverse dovetail" to clamp tabs. Nobody said "good idea" but Nabil actually made one. :)

    Spray foam from a can to absorb vibrations or your common silicon sealant used for bathrooms and kitchens is often used to stop vibration damage on electronic sub assemblies and should be ok to glue cells together. But no matter how good rubbers are, there is flex.
    http://youtu.be/kVIhT__TK9U?t=58s
    A tough Shore 80. Not good for tabs or cell corners. On the other hand, cells change thickness depending on state of charge.

    I'm awaiting idiot paypal to do what it is supposed to do; up my limit. It's that time To make up packs to my own spec. different to anything we see here but will be listening in with interest on best materials and procedures to suit.

    I've gone off rubbers as a block though.

    ReplyDelete
    Replies
    1. Andy, I was amazed when you suggested the idea originally. I had already been through a couple iterations of refinement at that point! But indeed, no one gave it much thought. Not even after I posted my video. Im not sure what my low hit counts mean. I'm sure there is a large contingent of viewers who have never given these comments on jacks blog any attention.

      Delete
  75. Jack,

    I've heard you mention more than once that your PFC75 doesn't shut off. Reading the manual and spec sheet for it indicates that it has several timer options. Is yours set to start the timer at the CV stage and then only run for a few minutes? Maybe something got accidentally switched on it somehow.

    Oh, and manzanitamicro.com is back up and running.

    ReplyDelete
  76. Gizmo:

    Maybe maybe maybe. We have switch 1 and 4 ON and all others OFF. It should terminate on the timer. We had the timer at the first position of 15 minutes.

    It will run for hours at about 1 ampere.

    Jack Rickard

    ReplyDelete
  77. I just wanted to note that with the PFC charger the switch numbers start with 1 on right and 8 is on the left. Also, on is down and up is off. It is a bit unusual. With my PFC-20 have 1, 3, and 4 on and my charger timer works normally. I generally set it for 3.5 vpc and then hold for 40 minutes (3 clicks clockwise from the 3 'o'clock position.) I have an older PFC-20 and it did loose the timer function at one point. As I remember the shipping cost more then the repair.

    ReplyDelete
  78. Hi Jack,

    On your battery casting process, you might consider making a pattern and cast a 1/2 mold that splits on the diagonal vertically. After casting two half molds they could be sealed for casting around your battery pack with your wide shrink wrap.

    On another subject have you tried making your misc copper parts out of Alloy 145 from Mcmaster-Carr. It machines very well, I use it for making EDM electrodes which are cut in a CNC machine to very complex shapes.

    ReplyDelete
  79. Can you use the same or similar plastic casing for the A123 that they use for the prismatics?
    Could one design terminals to fit Calb or other casings, and will the A123 pouches fit inside?
    Who supplies the casing to Calb, Winston, etc? Or do they make the casings themselves?
    Could we buy the casings at a cheaper rate than the resin?
    Could we design a casing, have it made in bulk, and have a cheaper, lighter module than pouches smothered in resin?
    In the casing, are the A123 tabs more likely to fail than if they were encased in resin?
    Feel free to ignore. I’m not adding anything, only asking a bunch of questions.

    Kind regards,

    Padraic

    ReplyDelete
  80. Read where A123 is making a 12 and 24 volt replacement battery for military use and as a start/stop device for autos; the battery has its own build in charge controls so I think they can be charged by an alternator, or in our case, an external charger.

    So, Jack, the idea of making up a 12 battery puts you right in the game. However, how do we make your battery lighter? I watched the shrink wrap process with interest since it holds the cells together nicely. So, seems to me to make the whole battery stable, you need to only mold the tops of the cells to the terminals.and enclose the battery in a light weight box.

    ReplyDelete
  81. This comment has been removed by the author.

    ReplyDelete
  82. Lad,

    Did you not read my suggestion using 4mm Correx with internally slitted correx cells (only one wall) every fourth row to hold the battery edges inside? (16mm).

    The box covers all sides of the batteries. All the batteries sides and edges are 'silco-setted' internally in place and the box itself is hot glue gunned together to seal up completely. The tabs stick out of the slitted top and all gaps glue gunned. Pour the mix into that top tray: The box sides made high enough to retain the mix so no need for a mold.

    Not as tough and foolproof as Jacks but ought to save ~10lb per 30 cells which matters for my build.

    That is what I aim to be doing.

    ReplyDelete
    Replies
    1. Please read the 16th comment in this group; but hey, we are family here so "your idea is excellent."

      Delete
    2. Haha!
      We are singing from the same sheet? Simply defined the material: Correx, can be got cheaply after the election fraud results or a few dollars for an 8' x 4' sheet.

      This stuff will set the cells up accurately with a fine air gap as recommended by Kokam.

      Delete
  83. Hi Jack,

    I was wondering on how you plan to install the 6p4s pack into a car.

    Are you going to put them into a box as per normal?

    I was thinking that you may be able to bolt them directly to supports if you cast a thread into the battery.

    Alternatively in a different direction, assemble an aluminium box not much bigger than the assembled cells, and cast them inside the box. This way they are still immobile but you can use much less marterial - hopefully a tenth.

    The plan then would be to bolt this box to supports. At this moment I'm thinking of having the bolts in the box before recieving and potting the cell assembly. The bolt head would end up being the spacer.

    ReplyDelete
  84. The temptation to a custom pack and the inclination toward a generalized module is a battle pretty much for everyone in the world. I've come to admire the simplicity of the Chinese prismatic with the ribs for air flow.

    A 6p4S looks a little bit like a starter battery. Maybe I'll put one in an ICE car and see how it fares.

    Right now I'm working on the next "secert modjewele" - codename FLATENUM.

    More of a custom module to bolt on under the car after the fashion of Tesla's low slung cell modules. Our flatenum series is shaping up nicely.

    Jack Rickard.

    ReplyDelete
    Replies
    1. I would recommend looking at the dynamics of aluminum composite materials. They are nice to work with. It is a sandwich of foils lining both sides of a remarkably soft plastic. The foil doesn't stretch, so it actually creates a very rigid sheet. A similarly sized and thickness of solid aluminum will bow much more under nothing but it's own weight! The key is that the core material keeps the two lining foils aligned throughout the whole sheet. I'm not sure, but rather optimistic that these principles could yield a very remarkable über flat module. The model S is one fine example.

      Delete
    2. This is the most interesting line for me. Model S has remarkable technical solutions, among which a flat-bed battery stretching across the entire floor. Many of our donor cars simply cannot opt for such solution. Perhaps the other face of the floor, inside the passengers cabin floor space could be used. Making a new hollow floor for batteries, on top of which seats will be positioned, a bit higher then originally.

      Delete
  85. Call me crazy, but I just have an issue with batteries, cables or anything electrical/electronic being on the undercarriage of a car. Water, and in some states salt, does bad things under there.

    - Doc

    ReplyDelete
  86. Ok. You're crazy. But before you wanted us to call you Doc Which is it?

    Yes, and those electric FIELDS and magnetic FIELDS they can do bad things to you too don't they" I mean like, haven't they studied mice and stuff? Didn't they like grow tumors or something?

    I read that cell phones cause brain tumors in mice - but only those on AT&T who suckered in for that two year contract plan.

    And what about that dreadful STATIC CLING thing with the dryer. What's that all about?

    Saltwater certainly does bad things to cars. It rots them out for one thing starting in the wheel wells. But that's how they deal with snow in some areas. Wash your car after having sex is all I can tell you. If you do it often enough, you're car won't rot.

    Jack

    ReplyDelete
  87. http://theunderstatement.com/post/18030062041/its-a-brick-tesla-motors-devastating-design

    it might not be true but if it is it isn't elegant.
    in general I wouldn't design an EV with loads when it's off. when it's off it should be off.
    I remember Martin Eberhard first remarking that the 100 watt water pump ran continuously even when the car was off. that was changed later though but it does lend credence to the idea that TM has been sloppy with parasitic loads. one of the TM pdfs cited in the blog states that the car can drain as much as 7% per day when off. that's pretty terrible with a 53kWh pack. if you can actually kill a 40k$ pack by leaving the car unplugged for a couple of weeks they dropped the ball on the design. that's a nasty vulnerability.

    ReplyDelete
  88. If you have sex a lot in your car it won't rot? That explains everything about my last car.

    A Tesla style battery belly pan to replace VW pans or made for say, the 818... oooooh Matron. Quick! pass me another hankie.

    The Volt n Leaf have under slung batts. I guess its down to applying reasonable automotive sensibilities in the design. Gotta make 'em magnetic mice proof.

    As goes self discharge. Dan, do you know of just one car Jack (or anyone) put together that has no constant parasitic load? Seems its only me who's making up a vehicle with native pack voltage for everything.

    ReplyDelete
  89. speaking of self discharge, some data points.
    I took out some batteries that I've sampled from around the world, mostly around 2008. 3 out of 4 Cens cells had killed themselves below 1 volt and my first 10Ah headway cell was down to 0.8v. never charged or discharged.
    the A123 cells are still strong at 3.3v and so were all the Valence cells at around 3.2.
    I had some Ryder 18650 3.7V cells and they were all good at 3.8V I put it in a flashlight and it lit for 2h45m. I charged it to full at 4.25v and it lit for 2h57m. the ultrafire cell that came with the flashlight only lasts 2 hours.

    ReplyDelete
  90. Dan:

    What? A useful and contributing blog entry from Dan Frideriksen? As I live and breathe.

    You bring up an EXCELLENT point and so does the article. I had no idea that Tesla had this problem, and actually NO IDEA that I DID TOO.

    But I HAVE killed some cells on a GEM with Thundersky's over a STEREO that always uses a wee bit of power. The problem is, it might be parked for five months with no one using it at all.

    Worse, plugging in doesn't help on ANY of our vehicles. Unlike Teslas, our chargers are NOT that smart. NONE of them that I know of has a setting for maintenance with the lone exception of the MANZANITA which I routinely turn this feature off. The Elcon's and Brusa do not even support it - or such is my understanding. Perhaps there is a mode somewhere on the Brusa. In most cases, you have to actually unplug the charger and plug it back in to reset it and get it to go through the charge sequence again.

    And ALL of our cars have parasitic loads Dan. I went through this with the eCobra trying to cater to the owners wish TO have an aux battery. We just killed several batteries within a day or two. They just get bled off to squat immediately. A pack, through the dc-dc converter, simply takes longer.

    Our solution? A maintenance switch. If you are going to park it for more than a few days, we have a switch to BREAK the pack, disconnecting EVERYTHING. We have had no instances of these cells bleeding down on their own though we DID DESTROY 16 cells on the Mini Cooper with two Cell Log 8s devices. Those tiny meters and the TINY amount they draw, over the course of 3 months of inactivity with the car, or let us say, activity related to rebuilding the drive train, was enough to drain down those cells dangerously below the others.

    I cannot stress enough that a TINY 10 ma draw, works out to 240 maH per day, which is 1.75 Ah per week, which is 7 Ah per month, which is three months is 21 Ah.

    ANd unfortunately, what you see in this article is homes being renovated, cars being shipped to Japan, and just inactivity. The Tesla, and our cars for that matter, are mostly toys. And the owners have other cars. And well, things get parked.

    In my case, if I turn the maintenance switch off, everything goes to zero and holds. But with it on, even the little AVC31 board of Modular EV power draws 12v continuously and it HAS to be on for the plug in J1772 charger to work. It will NOT enable charging with out it.

    Most meters have the same problem - constant unswitched 12v or they don't count Ah while charging?j

    PROPERLY DESIGNING an electric car to NOT use such parasitic loads can ONLY be performed in a Dan Frederiksen fantasy world. It is almost impossible in a Jack Rickard reality world from which I do suffer.

    But I can see that this would be a pretty hairy set back in a production car such as a Tesla with non-technical customers.

    How WOULD you prevent such a thing? I suppose the ultimate "daughter" mode would be if the pack gets down to 10% state of charge, you disconnect the entire pack????

    Jack Rickard

    ReplyDelete
  91. Hi Jack,

    I've ordered a bunch of A123 cells few weeks ago in order to make some tests on it and to improve my own experience on Lifepo4 technology.
    I'm especially looking for a small pack, with very high charging rates abilities.

    I've started some tests on these cells built in a 1p16s pack (non selected cells) that I've charged at 0.5C, 1C, 2C, 3C and discharged each time at 2C.

    I've obtained such pretty good results from 0.5 to 2C, but i've noticed huge divergences from 3C.
    On 16 cells, 3 of them at least seems to have a bad IR as they reach the 3.7v HVC point in few minutes with temperature near to 120°f contrary to other that never rise over 95°f..

    My initial goal was to verify the best capabilities of these cells at high charging rates compared with Headway cells but now, I've some doubts about reaching a 6C rate with this cells...
    Does I'm hit up on a bad batch of cells ? I need to verify with an other...

    Apart from your high discharge tests at 23C, have you made some similar charge tests on it ?

    Thanks,
    PY

    ReplyDelete
  92. doing away with parasitic loads is easy and that's how you fix this very serious problem.
    for charging you simply have a momentary button that you push after connecting j1772 that enables the needed systems to come on directly from the pack that is otherwise fully disconnected.
    for the duration of charging whatever energy counter that needs to count is also powered but not after charging. there is no need for parasitic loads. simple

    ReplyDelete
  93. Dan:

    It is EASY as a THOUGHT experiment in a FANTASY. In the real world it's not even possible. Do you want to reload all stereo presets every time you get in the car? How about your onstar info? Your telephone contact list. We actually have TWO DC-DC converters on the Mini cooper. The car can TELL when I'm close to it IF I have the magic amulet on me. But of course it has to have 12v to do that - and let me into the car. Without it, even the rear cargo door is locked.

    The list of things is interminable. It would actually be difficult to compile. In your world of heavy thinking, it just all goes away. In my world, it does not.

    We have tried aux batteries and aux systems. They basically eat batteries. So we run off the big pack. But within days, it would eat the pack if we didn't shut it off or keep recharging it.

    Jack Rickard

    ReplyDelete
  94. Hey PY,
    An exothermic cell when charging or discharging, from what I can find out is a fantastic indicator of a fully charged or heavily discharging cell. I cannot help but think its a better indicator than voltage monitoring. 3.7V is too much for real world use.

    These cells when charged, one pole gets warm and discharged the other gets warm instead. Some cell types can take a massive discharge but the chemistry on the other pole might be ordinary hence a more standard charge rate applies. This of course depends on the cells external surface area so it can lose the heat.

    I hope I'm right. I've only read myself clever on this.

    ReplyDelete
    Replies
    1. Yes Andy, I know that an exothermic cell reaction could be synonym of fully charged or heavily discharging cell.
      But the divergences i've observed for 3 of 16 cells appeared within 2min of charge at 60A, as all the cells were previously uniformly discharged until 2,6v and this after a full charge and top balancement at 3,65v.
      I precise that the pack is well cooled and the 3 divergent/warm cells are not only situated in the center of the pack.

      I heard and read that these cells could handle these +C rates without problem... the reality is different ?
      I will swap the three cells and retry...

      Delete
    2. PY,
      Did you balance at the bottom or the top? I'm reading two different things here.
      expect these cells to have varying capacities.
      There is no real world reason to top balance because if you run any low and the others are strong, you WILL kill the weaker cells. They will force the weak cell into a reverse charge. That means cell death.

      The mantra is never to over or under charge them. When the first cell hits 3.5V, stop. There is nothing to be gained in mileage by upping the stronger ones higher whilst overcharging the weaker cell. If a cell is too weak, consider paralleling stronger with weaker.

      Your pack is as only as strong as the lowest capacity cell.

      ==================================
      I've got another joke.
      Q: How do you know when Dan's cells are running low?

      A: His bow-tie stops.

      Delete
  95. Jack, I want them to store settings in flash memory and not volatile memory. I would hope most things do that already.
    If I say it is possible it is because it is

    ReplyDelete
  96. Hope?

    Dan you're delusional. If you say it's possible it is because you are OFF YOUR MEDS AGAIN. YOU PROMISED TO GO AWAY.....

    Jack

    ReplyDelete
  97. Jack, if you don't attack me there is no conflict. try to be civil.

    ReplyDelete
  98. Dan, you go store your stuff in flash memory. Let us know how you overcome everything.

    I found a lovely little passive low voltage warning circuit. Cheap and simple enough for every cell.
    When the pack V is high and output is low, it still consumes. Like everything else, these can never be allowed to reside on my cells when I isolate them no matter how minuscule the current.

    ReplyDelete
    Replies
    1. I plugged in an old word processor a few years ago and was surprised when it began to daisy wheel a letter I sent to a foreign embassy about 20 years ago. Something which hadn't been electrified for more than a decade had perfectly stored a lengthy letter: why couldn't this type of memory be employed in cars, as I'm gathering Dan F. suggests?

      Delete
  99. This is my first blog ever so I might as well jump in with both feet in my mouth. Please forgive my ignorance, but why not just make a box out of good old-fashioned sheet steel? By casting these cells into a resin block you're eliminating the advantage of easily replacing one bad cell instead of several and adding bulk that makes finding space for additional power more difficult. From everything I've read in this blog the problem is metal fatigue of the conducting tabs from vibration, this is easily solved by making the box small enough to hold the pouches securely. Make the main body as a channel open on both ends and have the end plates attach with bolts and threaded rod to provide pressure to keep the pouches from shifting. You can line the box with some material to keep anything from shorting out and since the structure is in the steel the insulator can be cheap because it doesn't have to have any strength. I I think that I'd try powder-coated, or epoxy-painted steel with pickup bedliner material on the inside to provide electrical insulation. I saw an episode of a TV show called Smash Lab where they planted explosives under the bed of a pickup truck. The bed was mangled, but not as bad as an uncoated control subject truck and the liner did not appear to crack or delaminate. Once you have the cells securely clamped in the case you can secure a plastic bar that fits between the positive and negative terminals of the cell bundles and below the clamp blocks to the steel box and then secure the clamp blocks to this with silicone adhesive to keep them from moving. In this way everything is secured, but can still be disassembled if needed. You'd have to put a lid on the whole thing, but this could be a simple plastic plate or (my personal favorite)more steel with bedliner coating. Any length would be possible without locking yourself into fixed dimensions for 13V blocks. I hope I haven't come across as a total idiot, but whatever I try I can only afford to do once and I'm trying to stay away from anthing too expensive. Let me know what you think.

    ReplyDelete
  100. I would like to think that whatever we tried we only had to do once. Somehow I wind up doing the same thing over and over again until I get it right.

    We're delighted your first blogging experience was here.

    Jack Rickard

    ReplyDelete
  101. I would like to think that whatever we tried we only had to do once. Somehow I wind up doing the same thing over and over again until I get it right.

    We're delighted your first blogging experience was here.

    Jack Rickard

    ReplyDelete
  102. Flatenun series pack!
    Hey I considered that for my project. It fits my floor pan width. Thought against it after. The reason is if I make 120mm high 50V packs side by side (100v) for 6 sets in parallel, it would be more modular and serviceable.

    Considered that glassfibre Hexel sheet for batt. support and floor but it's getting expensive.

    ReplyDelete
  103. I've seen a variant of my question asked, but I am left wondering if it would be possible to add something like styrofoam spheres the size of BBs to the resin mix in order to save money and weight?

    I don't understand what use anyone would have for pouch cells with shortened [cut] tabs?

    I found this 48V/20AH through drgrieve's DIY forum link:
    http://translate.google.com/translate?sl=zh-CN&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fitem.taobao.com%2Fitem.htm%3Fid%3D13805031856

    Is it made up of pouch cells? If so, how are they stabilized and contained?

    ReplyDelete