EMW Blog
Blog home
The First EV Conversion by the EMW Team - Fiat Spider 124 Convertible
- by Valery Miftakhov, May 2011
I myself contracted the 'EV bug'
long, long time ago - maybe because of my High Energy Physics background or maybe
because my father started teaching me all things electric (and later, electronics) at the
tender age of SIX. ;-)
I remember though that my first encounter with the soldering iron
was not very fortunate - as I grabbed its hot business end with my bare hands when I was about
six. My mother reacted in a non-traditional way and insisted that the best way to avoid accidents
like this in the future is for my father to teach me how to use that stuff. So there it went...
Anyway, I've started thinking about electric cars a few years ago when I first heard of
Tesla motors.
Obviously, being right out of my grad school I could only look
at the $120K sports beauty but the idea
got planted. In fact, now, 7 years later, Julia (my wife) and I are
in line for Tesla Model S delivery sometime in
...2012 or 2013. But after thinking about it a bit more we
have decided that we can't wait that long and be at the
mercy of someone else's plans and timelines. So we were going to just do it ourselves.
Soon after (late 2010) we bought our donor car - a 1977 Fiat 124 Spider. Always had a soft spot
for Italian cars and the right kind of California rust-free, bright-red beauty just came along...
Our target EV performance parameters were:
- Range: 100+ miles in city, 70+ miles on the freeway (for comparison, Tesla Roadster has freeway range of 140-150 miles - per Tesla Efficiency blog)
- top speed: 80+ mph
- 0-60mph: less than 6s (compared to the original Fiat's 11s...)
- recharge time - 1.5 hours with minor modifications to the house electric panel (PG&E doesn't even need to know); 3 hours from a standard electric dryer outlet
|
|
The rest of the Team EMW (Henry, Andy) have joined our pursuit not too long after.
After doing a TON of online research on forums,
and reading a "Build Your Own Electric Vehicle" book,
we have tentatively decided on the following "Big3":
- Kostov 11" brushed DC Series Wound motor.
Some other people (most notably a 'crazy' Croatian EV enthusiast
CroDriver - his DIYElectricCar.com username)
have been able to show up to 560 shaft horsepower (!!!)
on this motor with the right type of battery pack and controller. Good enough for my tiny 2200lbs car.
More info on this motor here.
- Soliton1 340V/1000A controller. Not too shabby either. Made by
EVNetics.
This is the most advanced EV controller at
this point- with 1000A continuous current rating and ~300kW top motor power (450hp) - if your
battery can deliver, that is ;-)
- LiFePo4 192V nominal 100Ah pack from CALB battery company. Rated at 8C discharge for 10sec.
Will probably take 10C for a few sec no problem. 2,000-3,000 cycles if no abuse.This is
to my knowledge the most advanced large-cell battery type available today and CALB has a
very good reputation for quality, manufacturing process control, and performance ratings of all
the other large-cell ('prismatics'). For more details, see
their site.
That said, the battery will be the weakest link in the whole system. This has been true
since the dawn of times so no surprise there. After the initial period of testing, we
will plan to add 20-30 additional cells to the pack to match the capability of the remaining
components. For the next conversion (a beta test conversion of a 3-series BMW) we will use ~90-100
cell pack to maximize the power. That level will also be used for the production conversions
we are planning to roll out later (for the total wheel horsepower of 300hp - to exceed
the power of any gas 3-series BMW except the latest
M3 and 335 models).
I know that most of the EV conversion enthusiasts would consider this setup to be an overkill
for such a small car (2200 lbs) but I really wanted to avoid building a glorified golf cart...
My other car is a tuned BMW 335xi (400+ whp) so I do place some premium on performance... ;-)
With the battery being the weakest link, the total max electrical power that we can get from
this setup comes to ~170kW. Taking into account the drivetrain & motor efficiencies,
this means ~170 wheel horsepower - which is ~double the original car.
What's more important, however, is that the torque at 1000A is ~300 ft*lbs and the sub-6s
0-60mph is virtually guaranteed with at least one gear shift (from 2nd to 4th, for example).
After the most daunting choices have been made, we have spent a lot of time taking a TON
of courses on various maker's subjects like Strength of Materials,
CNC design & manufacturing, Plasma Cutting, CNC Mill operation, Welding,
Hardware design, high-power electronics, switching power supplies, etc, etc.
TechShop Menlo Park
is awesome for that kind of stuff!
And with 1 CS and 2 Physics PhDs on the team, it was reasonably easy sailing ;-)
|
|
The conversion itself split into the following 7 major types of tasks
|
- Procuring the components
- Building a high-performance EV charger from scratch (more on 'why' below)
- Designing & making a motor-transmission coupling
- Designing & making a Motor mounting system
- Designing & making an Under-hood mounting sub-frame and boxes for battery & controller
- Designing & making an In-trunk mounting sub-frame for batteries and charger
- Replicating various accessory systems (e.g., alternator, brake vacuum booster, etc.)
1. Procuring the components
Not as easy as it sounds. While there are a lot of companies selling those, getting a
compatible set of 30+ types of components one needs to convert a car is not particularly easy.
Some of the companies do offer kits that are supposed to work together but (1) they are
generally underpowered for what we wanted to achieve and (2) are overpriced as they always
include stuff that you don't really really need. So we spent some time researching and shopping
for just the right connectors, just the right cable, etc, etc. Majority of misc components were
bought from EVSource,
motor and controller from RebirthAuto, and batteries direct from
CALB office in LA. Generally, the experience with the above vendors was great, with the
exception of having to wait for over a month while our controller was being assembled by
EVnetics team...
Total component cost at this point - ~$15,000:
- $8K for the battery pack
- $3K for the controller
- $2.5K for the motor
- and ~$1.5K for the assortment of 30+ misc parts
like vacuum pump for brakes, liquid cooling system for the controller, extra-thick cables to
carry 1000Amps, and a bunch of other stuff
2. Building a high-performance EV charger from scratch
Why would we do such a thing, one might ask. Turns out that this is an area of particularly
fat margins if you are really after high performance (i.e. short charge time in case of a
charger).
And we were definitely after high performance. The typical US household has a 240V dryer hookup
with a 30A breaker. A smart charger capable of fully utilizing such a line would set you back for
at least $4-5K which was definitely not acceptable when the total cost base of all other parts
combined is $15K. Also, thinking about the future conversion services, we did not want our
customers to have to spend that kind of money on a decent charger. We felt that we would be
better off doing some R&D in-house and passing the savings to our customers later.
Therefore, we have decided to do it ourselves.
And we made it fully open source!
All the fun details are available on another page of this site but the summary is as follows:
- 10kW output (50A) charging for our 192V pack
- Total component cost of ~$500 - including the DC-DC converter
(the latter basically is a replacement for an alternator)
- Size and weight lower than the commercially available solutions of similar power level
So 10x savings relative to available
commercial options! Not too bad...
3. Designing & making a motor-transmission coupling
The main task here is to build a small (something like a 3" diameter, 3" long) part that
accepts motor shaft on one side and connects to the flywheel on another.
This is probably the most complicated part of any EV conversion - for 3 reasons: (1) very tight
tolerances for the 6,000-rpm part to avoid vibration and resulting drivetrain destruction,
(2) it needs to be custom-made for the
[normally irregular] shape and design of the stock flywheel units, and (3) it needs to be
made of extra-hard material (such as stressproof carbon steel) to withstand the torque loads
from a big electric motor.
After unsuccessfully trying to find a machine shop that would design & build this for us, we
have decided to do everything ourselves. For some reason, everybody wants to charge you a few
thousand bucks on anything custom. So we said screw it, especially given that we want to do
that again and again with our future conversions. So we sat down,
learned a bunch of CAD design tools, done a lot of measurements and a couple of
versions of the final design before we got it right. Took us probably 30 person-hours total.
We have CNC machined the first copy of the finished design out of aluminum - fine for
initial motor break-in (<300A) but likely not good enough for full power. For the latter, we have
also machined a copy out of stressproof 1144 high-carbon steel
(the same material the motor shaft is made of).
We wanted to keep the clutch system for 2 reasons: (1) the clutch disk has stress-soaking
heavy-duty springs that will help absorb torque spikes from the huge electric motor when
we floor the gas ;-), and (2) you need to shift at least once if you want faster 0-60.
Generally, the idea is to drive in the city mostly in 2nd gear, on the freeway mostly in 4th
According to my calculations, the above setup will result in in ~5s 0-60 time
(including allowances for all losses in transmission, rear diff, tires, etc). Some
other converters opt for a clutch-less system but that means you need 2-3 seconds to shift gears
as you need to wait until your synchros in the transmission adjust the transmission RPM to the
right value. Too slow for my taste...
The last component of the coupling system is the adapter plate between the motor and transmission
bellhousing. The length of the coupler and thickness of the plate has to be precisely matched to
get the 'critical distance' right (the distance between the edge of the bellhousing and the outer
surface of the flywheel). If you get that wrong by just a couple
of millimiters, you either won't be able to shift or will be slipping the clutch all the time
and hence not have much power delivered to the wheels. The plates were CNC machined on the
TechShop's Tormach mill out of 5/8th inch thick aluminum.
4. Designing & making a Motor mounting system
Apart from designing and building the charger, mounting systems were the most creative part of the
conversion - I felt a bit like an artist making something out of raw metal that would then
have a predetermined function and form. As a result, welding and CNC machining are probably
my favorite fab skills now. Also, the guys from Online Metals and Santa Clara Metal Supermarkets
are my best buddies now ;-)).
Anyway, the task here was to design the mounting system to connect
the motor to the stock engine mounts. The benefit of the stock mounts is that they
(1) isolate any vibration from the motor/transmission assembly, and (2) they absorb at least a
part of the motor torque reaction when you floor the gas. A few converters who used rigid
mounts for their motors had them snap off after a few hundred miles due to stress so using the
right mounting system was rather important.
The construction itself consisted of an
additional back plate attached to the rear of the motor, and welded brakets connecting the
back plate with the front coupling plate. Then the brakets were bolted onto the stock engine
mounts.
So far, everything is holding up great. In idle tests, I can see the motor rotating
about 1 degree or so - or 1cm on the outside circumference - pretty sizeable deflection that
would otherwise have to be absorbed in the rigid components. We will probably have to further
reinforce the mount with some kind of a cross-bar to better withstand the torque once we
lift the motor break-in power limit...
5. Designing & making an under-hood mounting sub-frame and boxes for battery & controller
Again, more art here as we decided that we want to fit in the absolute maximum number
of components under the hood to ensure the stock weight distribution does not change much.
You have probably already guessed from the photo on the top of this page that there is not
a huge amount of space under the hood of that tiny car ;-). An additional complication
was the construction of the vehicle frame - essentially sheet metal rolled into a boxed
structure for rigidity. Unfortunately, thin sheet metal is not so fun to weld things to
(ask me how I know...).
So we had to get creative.
In the end, we welded a separate sub-frame tightly matched to the hood dimensions and bolted it
to a few structural elements found under the hood - such as a cross-bar, front frame element,
and the piece of the frame holding the firewall. Then we mounted everything else
(batteries, controller, etc) to that sub-frame. To maximize the utilization of highly
curvy interior space, we ended up splitting the cells into 4 cell groups, which let us fit
22 battery cells (out of 60) under the hood. And the controller. Most of the batteries and the
controller were mounted in an aluminum semi-closed box that in turn was bolted to the sub-frame.
All in all, we estimated that
we lost 400lbs when we ripped all the gas engine components from under the hood
(engine itself, radiator & cooling system, etc.). Then we put back in: (1) 180-lb motor,
(2) 35-lb controller, (3) 160 lbs of battery cells, and (4) ~60 lbs of a frame and heavy-duty
wiring - for the total of ~440 lbs, or +40 lbs over stock. Success.
6. Designing & making an in-trunk mounting sub-frame for batteries and charger
This was a much simpler task than under-hood mounting as most of the surfaces and shapes are
way more rectangular. We did have to do a small number of modifications, though -
(1) remove the fuel tank, (2) remove the spare wheel and the mounts, (3) remove all
exhaust piping, and (4) board up the holes that appeared.
Once that was done, we
have designed a simple battery mounting rack out of the angle steel stock to hold the remaining
38 cells. Then on top of the batteries we have added some mounting brackets for our
charger (also containing DC-DC converter to get 12V supply for all the car systems and
temperature control circuits & logic, etc.).
I'd estimate that we still have space left for 30 or so cells should we wish to upgrade in the
future.
So far, the weight reductions in the back were: (1) the fuel tank with full fuel - ~100lbs,
(2) exhaust system - ~50lbs, (3) spare wheel and mounts - ~50lbs, (4) stock lead-acid battery -
~70lbs - for the total of ~270 lbs. Weight increases are: (1) 38 cells - total 270 lbs,
(2) mountining system - ~30lbs, and (3) charger - ~30 lbs.
So net weight increase in the back
~60 lbs. Total weight increase for the car: ~100lbs, with virtually unchanged weight distribution
which is awesome for preserving agile handling of the car. The suspension was changed over to the
stiffer sports shocks & springs to further improve the feel.
7. Replicating various accessory systems
Not as much fun as the previous tasks but needed to be done.
- Without a gas engine to provide vacuum to assist in braking, the brake pedal is very stiff -
not unsafe but annoying. So we need to install an electric 12V vacuum pump, sensors, and
hoses.
- DC-DC converter is required to produce 12V (13.8V, to be exact)
from the main battery pack (192V in our case).
A bunch of companies sell those but they tend to be expensive so we hacked a standard 12V
switching supply from eBay ($70). The resulting 12V is used for all the standard car
electrical systems. Now that we don't have an electric starter to deal with, max current
is quite manageable at 30-40A max. To buffer short-term (few min) high loads, a small
high-discharge LiFePo4 battery (4.2AH, 40C rated) is connected across the 13.8V line.
We are now working on changing over all the lights to LED lights so this current
draw will further go down by a factor of 2-3.
- Cooling system for controller. As you can see from the photo above, Soliton1
enclosure is essentually a humongous aluminum heatsink. As such, it can run for quite some
time without needing liquid cooling - even at 1000A! EVNetics claims that it can supply
max current for ~30 seconds before thermal derating when aircooled. If liquid cooled,
1000A can be supplied indefinitely. I think that on this small car, we will never really need
1000A for more than 30 sec so this is probably an optional item for this particular
project. However, the goal of this conversion is a testing ground for high-performance
conversions of heavier cars (BMW 3 and 5 series) so we will probably install and test
the cooling system anyway.
- Heaters. Now that we don't have a fire-breathing engine with 25% thermal efficiency,
we virtually don't have any waste heat. So we need to create some for those colder days.
Fortunately, not so many of those days in SF Bay Area so we don't have to go overboard with this.
The current design uses a regular dryer heating element, rated for 4kW at 200V - pretty toasty ;-).
The heater will be controlled by the ...charger - after all, the circuitry is all the same and
is good up to 10kW. The charger and heater will never be used at the same time so no
conflicts here. Another cost / component saving idea from team EMW.
|
|
Some more pictures of this conversion below. We are now building a more comprehensive
photo gallery which will be coming online in the next couple of weeks - along with some
other sections of this site. In the meantime, enjoy this limited collection below:
All in all, this was an absolutely awesome experience. We have learned a TON of new skills,
met a bunch of new people, and attracted a huge amount of attention. It feels great
to be able to enjoy a sports car without any reservations for the environment, gas prices, etc.
In fact, with the effective MPG of ~130 (taking into account fossil fuels used to generate the electricity), we are eliminating almost ONE FULL POUND of CO2 per EVERY MILE WE DRIVE
(see this EPA paper
for detailed calculations)! That's ~15,000 lbs of CO2 less per year!
To put this number in perspective for you, this is the amount of CO2 removed from the atmosphere by
AN ACRE AND A HALF OF FULL-GROWN PINE FOREST!!! (Botany.org paper with supporting calculations).
When did you plant 1.5 ACRES of FULL-GROWN PINE TREES last time, huh?!!
Moreover, it's just SO COOL to be able to make something like this. As the most recent
WIRED magazine issue (DIY feature) said, the manufacturing is entering the new stage when
individuals no longer have to just sit and wait until someone builds their dream - we can build
it ourselves!
As I mentioned earlier, we are hoping to make same exhilarating experience accessible to
1000s and 1000s of current owners of gas-guzzling performance luxury cars. So stay tuned. Next
conversion is a beta test conversion of a BMW 3 series! After we prove the ability to make that modern
luxury car faster and more fun than the gas version, we will offer help to progressive BMW
enthusiasts who, just like us, looking to be on the forefront of the electric revolution!
Yours truly,
Valery Miftakhov for the Team Electric (EMW)
May 2011
|
|
