Martin Eberhard, Co-founder of Tesla Motors, September 14, 2006

Many of you have asked me about several alternative technologies and why they are not used in the Tesla Roadster. These technologies range from other existing battery technologies (e.g. lithium iron phosphate) to technologies that are on the horizon (e.g. ultracapacitors) to impossible technologies that ultimately boil down to perpetual motion. I will try to address as many of these questions as I can this week.

Batteries A123 to Zebra

The bottom line for us is energy density: how many kilowatt hours can we fit in the given space, for the least weight. The amount of energy we can store onboard directly translates to range. For sorted-out, mass produced batteries today, it’s lithium ion, no doubt about it. That’s why all our handheld goodies use lithium ion.

What’s the difference between energy and power?

The amount of energy available from the batteries is like the size of your gas tank. The more energy available, the longer the driving range – all else being equal.

The amount of power available from the batteries is like the size of the fuel line from the tank to the engine – if the battery’s power output is too low, it will reduce the horsepower (and therefore acceleration) of the car.

You need to design for both.

Other technologies (e.g. A123’s lithium iron phosphate cells) deliver higher power density, but have lower energy density. These are beginning to find applications in power-hungry equipment like cordless drills – and maybe they are the right answer for a shorter-range electric car designed for quick sprints (electric dragsters?).

Some battery chemistries (e.g. Sodium-Nickel Chloride Zebra batteries) have similar energy density to lithium ion batteries, but have much, much lower power density. A 250-mile Tesla Roadster with Zebra batteries would do 0-60 mph in something like 20 seconds. (Zebra batteries also suffer from terrible self-discharge, but that is another subject.) As it turns out, when we designed a battery pack that delivers the desired 250-mile range, it happens also to produce enough power for the kind of acceleration we were after. In this sense, we lucked out with standard lithium ion batteries.


Ultracapacitors are not plug-in replacements for batteries mainly because their energy density is so low. Today’s ultracapacitors store around 1% of the energy of today’s batteries and the resulting car would have around 1% of the range. They do however deliver very high power density and are excellent for applications needing high power and low energy.

Another difficulty with ultracapacitors is that their voltage drops steeply (exponentially in fact) with state of charge. This means that we would need a fairly sophisticated high-power switching power supply to normalize the voltage to the motor, which isn’t impossible, but is definitely tricky and costly.

None the less, I do think that new high-energy ultracapacitors may one day be the right answer for electric cars, just not now. Nobody makes ultracapacitors with anywhere near the energy density or pricing needed. That’s why they are not used in any consumer electronics yet. But when ultracapacitors are ready for prime time, we will be all over them.

Something for Nothing

It is tempting to ridicule the various proposals for perpetual motion, but somehow I just can’t do that. Some proposals show a lack of understanding of the basic laws of physics, but other schemes are devious enough that the logical flaw is not obvious until you really dig into the proposal.

But listen: you can’t get something for nothing. You just can’t. If we put generators on the front wheels to charge the batteries while you drive, they will cause the drive motor on the rear wheels to consume additional energy. This additional energy will be more than the generators produce, in all cases, no matter what the particular arrangement of generators, batteries, and motors

The same thing is true for wind turbines: a wind turbine on the car will always increase the aerodynamic drag, and the energy it produces will never be as much as the additional energy needed by the drive motor to overcome the wind turbine’s drag.

They’re the laws of physics, folks, and there is nothing we can do to overcome them.

I get a couple of free energy proposals every week now – some very simple, some quite involved – and I admit right here that I do not and will not read them. Call me closed-minded, but I’ve come to have a lot of respect for the basic laws of physics. Push them to their limits, bend them to our needs, take every advantage of the corner cases – but don’t count on breaking them. (Oh – by the way – Nikola Tesla never made an electric car that ran for months with no batteries. Somebody made up that story. Trust me.)

Other Ideas

I was all intrigued by the idea of wheel motors. What could be simpler? Mount a motor right in the wheels. Presto, no differential, no drive shafts, no CV joints, instant traction control and stability control. What kills the idea of wheel motors for road cars is unsprung weight.

Unsprung weight is the weight of the car that is below the springs – tires, wheels, brakes, hubs, etc. The higher the ratio of unsprung to sprung weight, the worse the car handles unevenness on the road. Tracking the road is so important that even inexpensive cars now commonly are sold with aluminum alloy wheels, alloy suspension components, etc. that knock a few pounds of unsprung weight off each corner of the car. Adding 30 unsprung pounds to each corner would mean abysmal handling.

We thought a lot about putting solar panels on the upward-facing portions of the body. The trouble is that the square footage available for solar panels would only increase your daily driving range by about 5 miles. While there is a sensible argument for such solar panels topping off your battery when you leave your car in the airport’s long-term parking for a week, it just did not seem worth the complexity, weight, and cost.

Solar panels make so much more sense on the roof of your house, or out in the desert someplace where there is enough acreage available to power all your driving. This becomes obvious when you think of the electric grid as storage: solar panels produce energy during peak consumption times, and your electric car consumes energy at night while charging – when energy is the least expensive.

Solar panels on an electric car seem a bit like fins on a ’59 Caddy.

Balanced Innovation

Our strategy here at Tesla Motors is to use the very best technology that is actually available. We innovate where we can and where it matters, but we avoid frivolous innovation and getting mired in unsolvable problems. I think we’ve done a pretty good job with the Tesla Roadster. We will continue to evaluate new technologies, innovate where sensible, and make the best choices for every future Tesla Motors model.