[kerrymann]

Well I just bought a 1994 Mazda Miata for my next EV project. I am reusing parts from my last EV project, the Leadwing.

http://www.diyelectriccar.com/forums...ad.php?t=58378

For those not familiar with the project it was a 1987 Honda Goldwing trike that I converted to electric mearly to learn and act as a stepping stone for this project. That project was based on some very good, but old 24V AGM aircraft batteries that I got a really good deal on. With 196V and 880amps available from the pack and only one small steering wheel the performance of the bike was predictably scary. All the local drag strips prohibited trikes so I was never able to see what I could get on the 1/4 mile. Anyway, this winter the parts from that project were harvested to go into the Miata. I had been scouring CL for cars for many months till I found this one. I was considering my 1969 bug but it is so original and such good shape I couldn't do it. Considered another bug, a fiat spider, and some others but there is an awesome aftermarket for the miata. Throw in disc brakes, great handling, reasonably safe and the fact that I could get one in good shape for $1300 and it was the final winner.
The project's goals in no particular order are:

1) A fair weather only car
I will commute in about 2-3 months during the summer. I car pool with my wife and as a school teacher for the other 9-10 months out of the year in our prius. Top down driving means no worrying about heat/AC.

2) Range of 35miles at 75mph down to 80% DOD
I can charge up at work so I am aiming for a minimum range of 35miles at 75mph down to 80% DOD. Realistically I will be averaging 65 on the highway but I want to be conservative. I will have to pay particular attention to motor cooling as 80% of my drive is at 60-80mph. I may have to go to a Warp11 or even dual 9"s but I am really trying to avoid this by minimizing aero drag where ever I can. I need to do more calcs but I'll probably start with my current motor and if I have to upgrade, then so be it.

3) 1000amps for 10 seconds
I want a pack that will make the most of my 9" motor and the Soliton1 controller. I would like to autocross and track it occasionally keeping in mind the impact it will have on my battery life. Aiming for 0-60 in 6 seconds but that will depend on how much weight I strip from the chassis.

4) Simplify and then add lightness
I have seen miata's down to 1600lbs with the stock motor and extreme Weight Reductions. No top and side windows for sure but I may go so far as no real windshield. It depends on how crazy I decide to get. Look at my last EV and it's pretty clear I don't want to build anything that is "conventional" I want a curb weight of under a ton but I'll have to work for every ounce.

5) $10000 Budget
The $8000 is for the batteries, BMS, instrumentation, and adapter, cabling, materials, etc. Given my other goals mean I have to go Lithium and a budget to match. Another 2k for a charger.

6) Have fun and learn
As a Mechanical Engineer who has spent over a decade in research, design and development "Have Fun" and "Learn" are one in the same.

My plan for the car is two stages:

Stage 1: Get it Running
A pretty straight froward EV conversion using the motor and controller I tested on the Leadwing. I will get it running and registered keeping the rest of the car pretty stock. Many other people have done miata conversions with adapters and mounts being available off the shelf. The big question for me (as is most converters) is the battery pack. Right now I am leaning toward some gray market A123 cells. Specifically the AMP20M1HD-A. I am talking with one guy and going to order a few for testing.

Stage 2: Get Crazy
The second stage will be stripping the car down to a fair weather only superlight speedster. No top, maybe even no windshield depending on how crazy I get. I will be replacing the front bumper with some custom fiberglass to get rid of the engine cooling scoop but I am toying around with a fiberglass front clip to save more weight. I will fab a real roll hoop/cage for protection. Without the constraints for a top I can make a cage that will provide real head protection. I used to run SCCA time trials and it would be fun to do so again.

First step now is getting the title transferred. I always do this FIRST when starting a project. I don't want to strip it down and then tell me I have to drive it down for a VIN inspection.

[kerrymann]

Picture of the "new" car:

[kerrymann]

And a couple pics for inspiration:

[kerrymann]

Today my clutch kit arrived, aluminum flywheel, etc. I went with a Stage 3 becasue the electric motor can put out more then twice the torque as the stock engine.

This clutch and flywheel is the lightest I have ever held. The flywheel is only 8 lbs which is 33% less then the stock cast iron and I am going to lighten it up a bit more. Weight here is particularly important because it is mass rotating at <8000rpm. I was suprised that in all my searchining for deals on kits adding the flywheel only cost $60 more. It would cost me haldf that to have the cast iron resurfaced.

I don't need the starter ring so I am taking it off to save the weight.

Yes I am trying to make my flywheel cooler, but I am not making a pun. To help I threw it in the deep freeze. The aluminum shrinks much faster then the steel ring so this will give me a fighting chance of being able to get it off.

Now if only I knew where my motor adapter is going to come I would start ripping out the motor.... :-/

[kerrymann]

After 30 mintues in the deep freeze and 3 minutes of hammering, the starter ring is off. The flywheel now only weighs 5.7lbs.

[kerrymann]

No progress on the car as I am still waiting for my motor adapter. It is supposed to be done this week and at which time I will start ripping things out. But I have started the first steps on my pack.

Battery Testing

Last night I received a batch of A123 cells and I am starting testing. For those that don't know these are very good batteries that can put out over 20C and have shown very good life cycle in the labs at more reasonable discharge rates.

http://www.a123systems.com/products-cells-prismatic-pouch-cell.htm

The problem is they don't sell to individuals, only OEMs. So the only way to buy them is on the gray market. There is a lot of speculation about how they get on the gray market; some are rejects that don't meet specs (usually slight low capacities) and many have had the tabs cut to make them difficult to use. The other theory is they are not selling as may as they planned so there South Korean factory is selling them out the back door without corporate knowing about it. Finally some are just outright fakes. But all of them have to be bought from China (since people will sell ANYTHING in china without fear of lawsuits). I bought 8 of them from my supplier to make sure they were good before going off and blowing $5k on the ~220 cells for a full pack. If these check out then I will use them to make up my 12V battery.

I am pretty confident these are real A123s that came out the back door given the serial numbers were sharpied out and some of the other markings.

Romulus Anode?

The first check with a DMM and not a good sign on three. The 3 cells with an extra sticker that says "Romulus Anode" have just over 2V. I am doing a 2 cycle charge/discharge on one of the cells right now so we will see how it comes out when I get home from work tonight.

FYI, Romulus doesn't refer to StarTrek rather it's Romulus, MI where A123 makes they anodes as opposed to the Chinese supplier they used before. I have been doing more research on the fabrication of these cells and it seems a bit odd. They are all labelled made in the USA but common knowledge holds that they are actually made in South Korea. I find it odd that A123 would fabricate components in the USA then ship to Korea for assembly and them ship them back to USA to assemble into packs. I would think they would manage their supply change better then that.


Test Cell

Here is my test station. A CellPro PowerLab 6 hooked to a PC. It's a bit ad-hoc right now as I can't find my other set of test leads and I have to do some more digging to figure out how to get the PL6 to go to 40amps instead of the 10amps I am limited to now. Even though I have selected a battery power source with 40amp max, the highest it will let me select is 10amps. Speaking of the power source it is a couple of left over 24V batteries from the leadwing project with a trickle charger. I am also using a book plus a battery to provide some compression on the sides of the cell during cycles. This evening I will make some improvements to the set-up.

[steven4601]

Building an EV is nice, building a 2nd EV is even nicer.

Where did you order the pouch cells? Did you consider already how to mount/package them?

[kerrymann]

Originally Posted by steven4601

Building an EV is nice, building a 2nd EV is even nicer.

Where did you order the pouch cells? Did you consider already how to mount/package them?

I have a couple of people I am looking at weighing cost and risk.

My current plan is a stacked pack. Turns out very similar to A123's set-up but without laser welding.

Originally I was planning on 3 packs with just 66 cells in series but a BMS makes that less then ideal so I think I will be doing 3 packs of 3P11S. I am still working on the electical connection but am leaning toward a simple clamp and spacer system.

[kerrymann]

Well so far so good. The 2V Roumuls anode cell actually showed the highest capacity at just over 19ah which is much better then expected. Most people have been reporting capacities in the 17.5-18.5 range.




I still need to check the remaining cells since I only have a sample set of 3 so far. Right now I have cell #4 doing infinite cycles and I will check tonight and see how it is doing.

I still haven't spent anytime trying to get the software to go to 40 amps. Anyone have any suggestions? Doing it at 10 amps obviously takes 4 times as long.

One other thought I had is that there are several people on DIY forum that are testing these batteries from different vendors using the CellPro Powerlab6. If we all use the same test settings and post our data we can make a community pool of cell data that people can reference.

[drgrieve]

Originally Posted by kerrymann

I still haven't spent anytime trying to get the software to go to 40 amps. Anyone have any suggestions? Doing it at 10 amps obviously takes 4 times as long.

Post this question in the battery sub-forum, there will be a few folks there that can answer this question.

They can also share their test settings.

Cheers.