DIY Electric Car Forums banner

Starting new build: BMW 3-series (E46) - high-performance electric

96K views 143 replies 28 participants last post by  1-ev.com 
#1 · (Edited)
Hi All,

Opening a new build thread here. After converting a Fiat Spider convertible (haha, such a wordplay), EVs are a way to go for me. Fiat thread: http://www.diyelectriccar.com/forums/showthread.php/noob-starting-1977-fiat-124-spider-51822.html

So we are taking one of my other cars (2001 330Ci) and making it high-octane electric. In doing so, we will plan to develop bolt-on designs that can be easily replicated on other E46 bodies.

Target parameters:

0-60 below 5 sec. It just HAS to be faster than stock (which is not super-easy to do with a 330...), EDIT: EVcup qualifier is <5 sec 0-60
125 mph top speed. EDIT: EVcup qualifier is >200km/hr top speed
100 mile city range

Configuration options we decided on so far
Battery: 90 CALB 100Ah cells for a 288V nominal 800A peak pack
Controller: 1000A DC (most likely Soliton1 but if I can get 1400A WarpDrive in less than 4 weeks, I'd try it)
Motor: Kostov 11" 250V SFM, with space left for dual 11" setup later.
Charger: our own design (open-source, http://www.diyelectriccar.com/forums/showthread.php/10kw-60a-diy-charger-open-source-59210.html) - 10kW, 60A
Drivetrain: not a direct drive (see below)
Coupler/transmission plate: from ElectricCarInternational. Can't beat the price and no design fee. Plus good discounts on subsequent orders of same design.

Open Questions:
1. Mechanical drivetrain setup. I definitely want to have a transmission. The question is: do I keep stock manual 5 speed or swap it for one of the automatic sets like http://www.grassrootsev.com/tranadapt.htm. Benefits are: even higher drive-ability and ability to use idling to belt-drive accessories
2. Accessories. How to best set up heating, AC, power steering, vacuum, etc. In a Fiat, I have ignored all of these - after all, it's a small convertible so braking is fine without assist, heating is a joke anyway with top open, no power steering / AC. In a beemer, all 4 of these will have to work. Now, if you do that in a 'traditional EV DIY' way, that's 4 electric motors and subsystems you have to mount somewhere, etc. Why not keep all the original pumps and pulleys and just mount then on the backplate of the motor and drive all from idle of the main motor (if using automatic tranny) or one small high-voltage DC motor?
3. EV management system choice (or build) - I want a big color screen displaying everything I want about the vehicle at all times. It has to look like a production electric car, without any ammeters on the dash, etc. At this point, the thinking is to build our own system based on Arduino for data gathering and Android tablet for display. If anyone knows any other systems available on the market today, please let me know.

Comments / suggestions?

Let the fun begin! (again...;-)

Valery.
 
See less See more
#2 ·
Hey! be careful.

You don't have enough battery power because your voltage is too low!

800 x 0.0009 = 0.72 - 3.2 = 2.5v x 90 = 225v.
800A x 225v = 180Kw from cells and roughly 200 hp at motor shaft....... so, not enought!

You probably need at least 250 hp....
 
#3 ·
Thanks Yabert! I was thinking of going to 100 cells which would take care of that. But then again,
1. pure hp rating matters more for top speed than acceleration
2. My latest shipment of CALB cells lists 0.3-0.5 mOhm IR (each cell is tested separately). So the sag should be lower (at least on new cells).

V
 
#5 ·
sounds like a fun build, for the power level you are looking for you might want to look into higher power density cells to shed some weight unless you need it all for the range you are looking for. Headway/A123 etc.

I really like your display ideas, nothing like taking a step towards a Tesla model S and away from an obvious garage creation.

I kind of like the idea of a higher end automatic, I know the transmission in my G35 shifts faster than a stick, and it doesn't have the typical automatic lag. The only thing I'd complain about is it's always in the wrong gear, but with the wider power band of an electric it would be much easier to keep it in the right gear.

Using the idle feature in the Soliton1 has been very helpful while setting up my belt driven A/C even though the car is a manual.
 
#6 ·
#9 ·
Kostov was great on my Fiat. Plenty of power. Couple of notes though:
1. You need a high-voltage pack for it. Seems like a very obvious need but for some reason did not get through to me initially so I was running a 250V motor with a 192V pack and it didn't work very well. 256V nominal is getting there.
2. Pros say that you can't overload Kostovs for as long as NetGains (lower thermal mass) but that is likely to be relevant only for racing.

I still can't get over the fact that you can get dual Kostovs for less than 2 motors bought separately!! ;-) Direct drive with dual 11" in series on a 340V pack and 1000A controller, anyone? 1800lbs (0.5G) push at the wheels from 0 to 70-80mph? mmm... ;-)

V
 
#13 ·
Hi Guys,

still musing over the details of this conversion.

Narrowing down the DC motor choice to Warp11HV mated with 340V CALB pack. It just have to have a big DC motor to provide the power we need.

However, we are going to do some long-range demo drives in this car and every bit of efficiency counts so I was thinking of adding something like an AC50 system to the big DC motor. AC50 would then soak up braking / downhill and provide most of the 'baseline' load on freeway etc. Being AC and therefore more efficient, it should further add to the economy.

What do you guys think? Has anyone done it before? Any links / pointers?

Thanks,
Valery.
 
#16 ·
Do you only compare the coplete efficiency of the AC50 vs WarpHV????

DC = 99% controller efficiency x 87-88% motor = 87% efficient than seem good to me....

What about AC50?.... 94% x 92% = 87%???????

My final advice: Put a bit more cells. They costing less and they weighting less than an AC50 systems!!
 
#20 ·
ok guys! FINALLY some good stuff to look at ;-)

$1K in the friendly shop gives you an EV-ready beemer! Engine, radiator, gas tank and exhaust systems are out

design work is starting today. Everything in CAD to be replicable for our E46 kits.

V
 

Attachments

#22 ·
didn't weigh it but according to the published specs on parts, about 600lbs. Also, 16 gallons of gas will not be there, or another 100lbs. Finally, will be removing spare wheel and a 12V battery, or another 100lbs. So total -800lbs, give or take.

The original empty weight was ~3300lbs. So we are looking at ~2600lbs starting weight for conversion.

EV stuff:
1. Motor+controller+mounts: 300lbs
2. 100 100AH cells with boxes and mounts: 750lbs
3. Accessory drives etc: 50lbs.

So I am expecting to gain ~300lbs or ~10% of the original weight.

V
 
#27 ·
Hi Damien - nice to see you here ;-) check out http://www.diyelectriccar.com/forums/showthread.php?p=249779#post249779 for the charger update - got a 10kW PFC stage working. Produces 390V output so good for any battery voltage. In fact, If one uses 1200V IGBT part, one can boost to pretty much any voltage likely ever encountered in an EV.

The buck charger we all built then connects to this PFC stage (with all the AC input stuff (bridges / caps, etc) removed.

Total parts cost is ~$130. The toughest part was to find a PFC IC that would work at IGBT-compatible frequency (using 22kHz now - seems to work with a 200A half-bridge module I got from ebay - $20 in quantities of 10+).

V
 
#35 ·
Thanks Bowser!

Do you guys have any ideas for the way to construct battery boxes? Of course, we can just weld them from sheet steel and paint but to be sturdy enough (we have up to 300lbs of batteries per box), they would have to be made out of 14 gauge steel or something like that. That would make them weigh 60-80 lbs each! Not exactly optimal...

We thought of building a cage from 1/8" thick, 2"-wide steel strips. Another option is to do custom fiberglass or carbon fiber box. The latter, while cool, has a problem of being less replicable in quantity

Any other ideas?

Thanks,
Valery.
 
#36 ·
You can build your battery box with aluminum sheet and only add steel reinforcement where it's necessary.
Under the pack to support weight and maybe at some other place to add strenght or protection.

And don't forget, aluminum is 1/3 the weight of steel... so 9 Ga aluminum will be two time lighter than 14 Ga steel.
 
#37 ·
ok settled on fiberglass for battery boxes - see the pic for what 12 layers of fiberglass fabric can do (total thickness of 0.2 inches) ;-) Awesome forming capabilities, too.

For motor mount, die-cast aluminum. Will post some pics soon.

the biggest unknown right now is the motor controller. with 288V nominal motor, I'd like to have at least a 400V pack - so when it sags by 0.5-0.7V/cell at 1000A, I still have full 300V at the motor...

V
 

Attachments

#39 ·
Manzanita Micro is selling the Zilla 1K HV model, which in its description says:



Cost = $2,675 USD

So I guess with 3.6V as the fully loaded voltage on a calb battery that would mean 111 batteries in series, = 399.6V
Wow Bowser - this is cool - that really happened (Manzanita selling Zillas)! Z2K EHV available, too. $5K but hey, it's 640kW! ;-0

I think 1000A will not be sufficient for our purposes so the only commercially available option right now is Z2K EHV. Good chunk of change, though. Something to think about...

V
 
#41 ·
Power-wise it's probably fine. What we are also after, though, is 0-60. In fact, some of the events we want to participate in, <5 sec 0-60 is a qualifying requirement...

So, say, if we had a 2000A controller (and a tranny that could handle the torque), we could do the entire 0-60 run in 3rd gear for ~4-sec 0-60...

Shifting gears kills as much as 0.3-0.5 sec.

Our calculations suggest that we need at least 1400A controller to comfortably get sub-5 sec 0-60.
 
#44 · (Edited)
oh yes ;-) I love the fact they have all the driver circuitry in them already. the only concern i have is not being able to fit emitter resistors for load equalization... Have you thought about that at all?

EDIT: just thought up another potential issue: they are 600V modules so I can't really go above ~400V on the battery side. So might be those $500 1200V 1800A module time, after all ;-))

V
 
#43 · (Edited)
Damn i am excited for your 3-series conversion results!

You don't always need a lot of torque for fast 0-60 acceleration, look at the E46 M3, people were doing 0-60mph in 4.8/4.9 sec stock from...

333hp @ 7900rpm
262ftlbs @ 4900rpm

With a Warp11-HV motor, Crodriver was able to apply 1000A to the motor @ a little over 5000rpm...

380hp @ 5100rpm (284V & 1000A = 284kw = 380hp)
300ftlbs @ 0-5100rpm (1000A in an 11" motor)

With the right gearing you should be able to hit 0-60 in sub 5...no?

Have you timed your Fiat conversion from 0-60?
Is it 2200lbs after conversion?
I saw the videos with the burnouts, have you tried to turn down the slew rate and put some tires with more grip on the back? I wonder how fast it could launch 0-60 if more power stuck to the ground?
 
#45 ·
380hp @ 5100rpm (284V & 1000A = 284kw = 380hp)
300ftlbs @ 0-5100rpm (1000A in an 11" motor)

With the right gearing you should be able to hit 0-60 in sub 5...no?

Have you timed your Fiat conversion from 0-60?
Is it 2200lbs after conversion?
I saw the videos with the burnouts, have you tried to turn down the slew rate and put some tires with more grip on the back? I wonder how fast it could launch 0-60 if more power stuck to the ground?
with 284kW electric it will be more like 300hp at the wheels, if that - these motors are not that efficient at these power levels. 300 rwhp would barely get it under 5s. My 2009 335 is a 300hp stock rated at 4.9s...
 
#59 ·
A Harbor Freight Battery tester is about $50. You could run at 3C 4 batteries at a time until you got close, and then 3C on individual batteries. If they are 1/2 full it would take about 10 hours.

You'll want to tip the tester back about 45 degrees and blow a fan at the base to cool it, it'll get hot enough to ruin it otherwise.
 
#60 ·
That is one serious looking beemer! I can't let mine see it or it will get jealous :)

Regards the front end being down. I had this problem when i had 2 agms in a rack between the front chassis rails. Bout 150lbs.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top