How to Build a $15k Electric Car—Think in a Bigger Box, Not Outside One
What you’ll learn:
- Andy Turudic is the new editor of Electronic Design’s biweekly Automotive Electronics newsletter.
- As engineers and designers, what we know and our experience, puts us in a solutions box and that bias blinds us to either solving problems or dismissing them.
- Automotive heat pumps and vehicle spec sheets are keeping affordable electric vehicles away from the masses—how to change the thinking process.
It is just a box
That we design our stuff in...
Make your box bigger
About a month or so ago, I was “warned” that I’d be getting assigned a newsletter (emails with links to applicable content that are sent to readers who opt in to receiving them on a regular basis)—the automotive one, at that. Of course, being the Mostly-analog Editor, I have to admit that I got a bit whiny about why I couldn’t do an Analog one, since Power & Analog are a combined newsletter and splitting it out seemed logical given my experience and interests.
However, that was not possible, so I begrudgingly volunteared[sic], adding it to the already immense pile of work I give myself in writing these blogs, articles, soliciting and editing contributed author content (Stephen Dyer’s SPICE series is the latest example), and all the other “editor” stuff.
Now, this really was not as horrific as it might appear—my management are well aware of my Electric Vehicle advocacy on my personal LinkedIn account (I do post my new Electronic Design article notifications there as well). I have a hard-earned network of automotive execs and journalists in my circle, including Car & Driver’s contributing editor on EVs, where we exchange ideas, discoveries, counterpoints, the occasional friendly jab, and opinions as peers.
Every Idiot Can Count to One…
You see, a year prior to joining Electronic Design this past April, I was delivering my peer-reviewed engineering paper at SAE’s giant WCX conference in Detroit, asking the analog-guy question of what happens when you intentionally do not switch a traction inverter’s power stage fully on, or fully off? To quote one of our analog engineering greats, Bob Widlar, “Every idiot can count to one.”
So, you’ve got this three-phase motor-winding-current selector bridge that’s typically running somewhere between 50 and 100 A at a nominal 400 V DC on the highway on average. Every automotive traction-inverter designer is doing their darndest to full-switch the half bridges on and off to minimize power losses, including using silicon-carbide MOSFETs to reduce on-resistance, and Andy is standing in front of an audience of automotive engineers, at their major conference—engineers from a who’s who of leading automotive manufacturers saying—“So...what if you don’t do this”?
About a year before the conference, I posed the same question to the USPTO’s patent examiner, and they agreed that my solution was worthy of calling me an inventor, yet again. This was great, because I had spent a lot of my personal time and money on the design concept.
MOSFET Mess
Since many of our readers are analog-savvy, they very well know that what happens when you intentionally partially turn on a MOSFET in a high-power application is a laughable disaster and you send the idiot, proposing to do so, back to his village.
For our “digital” and coder readers, what happens is the MOSFET Rds, the apparent “resistance” of the conduction channel, rises, and you get a lot of heat being generated—it’s that I2R stuff you learn in the first week or two of a basic electronics course going on at 100 A. Or when you blow up your four audio power amp MOSFETs at $20 apiece as a teen and you’re pulling in $7 a week on your paper route—don’t ask how I know this.
With a “dumb” enough design, several kilowatts of intended heat (heating can be turned on and off with the gate driver) can be produced in the output-bridge MOSFETs while producing the current that makes the mag field, that makes the torque, that gets multiplied by the gearbox, that goes to the wheels and tires, that produce the force, that pushes the car down the road at highway speed. See? I know how cars work, lol. I have had a restomodding/hotrodding hobby, by the way, for about 55 years...
The Heat(er) is On
Not only are cars required to go down the road, or push your eyeballs into the back of your skull from a standing start, they’re also taken for granted in providing comfort for their occupants—air conditioning in summer and heating in winter. For electric cars, dead-dino juice is not being burned to heat a circulating coolant that’s looped through a cabin heat exchanger (the “heater”).
Most of the earlier electric vehicles, including Chevy Bolt, Tesla Model X, and Tesla Model S, among others, used PTC (positive temperature coefficient) element-based heaters, with some suppliers electing to use metal resistors. Many OEMs heated coolant to avoid the dangers of having 400-V power in the vehicle cabin, though Tesla used a PTC that heated cabin air directly. Tesla decided to abandon the PTC, a 100% efficient resistance-based heater, for an automotive heat pump that was around 300% electrically efficient in producing heat. Cool, eh?
Yes, being a heat pump, the same refrigeration pump and motor could be used for cooling and heating. Winter range of these electric vehicles was improved to where it’s becoming the norm in vehicles to come, including the second (third?) generation Chevy Bolt(ium).
Cold, Hard Facts
What's the problem, then? Refrigeration systems are expensive—a high-powered brushless motor drives a sealed pump, pressurized hoses and piping are needed, and refrigerants aren’t cheap. With a heat-pump system, say it has an OEM installed cost of $400, the price of the vehicle has to be margined up (sorry for the marketer-speak) to where the vehicle price could increase by around $1,200. Not a big deal if you’re building a $90,000 electric pickup truck, but a huge percentage of the vehicle price if it’s a $20k car.
So, what if you partially turned the traction-inverter MOSFETs on or off, producing kW levels of heat? Well, you’d need a customized three-level gate driver for starters and another power-supply voltage, say $20 worth of stuff. The SiC MOSFETs are already rated to a Tj of 175°C operating temp, and they’re already in a liquid-cooling loop, so reliably making hot coolant is a piece of cake.
That means $60 on the sticker price of an electric car with good old rollup windows, no air-conditioning, and with a thermally happy battery (still has a liquid-cooling/heating loop) and occupants when the cold weather hits in their 90MPGe electric car, 120MPGe with the heat off. No emissions, almost no maintenance cost, zippy performance, lasts 300,000 miles, and has 1/3 the fuel cost with the convenience for many of “refueling” overnight, at home.
$60 on a $15,000 car? Heating is no longer the long pole in the tent for getting a car built for the masses as battery prices are projected to plummet to 30% of current costs.
Take away the EV subsidies because of petrochemically induced corruption, and ICE (internal combustion engine) vehicles still can’t compete on price and maintenance with a chunk of iron spinning on two bearings, some copper wiring, and a single ratio geartrain that has very little to go wrong with it. Anyone can build a $50k car, but that market is getting crowded. Whoever gets clever and builds millions of $15-$20k cars will rule the world market, in my opinion.
Don’t Get Too Boxed In
My point here is that we, as engineers, think inside the box of what we know, what our experience says is a best practice, and of what technologies are in our bag of tricks. We have biases and prejudices about fastidiously minimizing Rds in an automobile’s traction inverter, just like we had limited insight on how to make an EV battery last more than 30,000 miles a mere couple of decades ago.
We cannot build a $20k car because the cabin heater, as well as the battery heater, is now a $1,200 component of the vehicle price. And that’s due to chasing winter efficiency to get us from 90MPGe on the spec sheet back to 120MPGe, indifferent to ICE owners getting 30MPG on a good day, having the $1k annual repair bills and ICE commands, and demanding an energy product that requires an $865B annual budgets and thousands of lives to secure versus one we all have that starts every morning.
So, to our auto OEMs, we engineers say “no” to the MBA product strategist/planner that tells us they could sell a bazillion electric equivalents of “Chevettes,” “Mustangs,” and “Falcons.” Those were Detroit’s peoples’ cars—cheap enough to where the Mustang and Camaro were touted by the marketers of the day as “secretary’s [salary] cars.”
Why Not Give the Automotive Electronics Newsletter a Spin?
Last week, I took the reins of Electronic Design’s Automotive Electronics newsletters without public fanfare, though those with a keen eye might have noticed some small changes. I’m looking at the newsletter with an automotive designer’s eye. Not as what you use, but what you should consider using or learning, to trigger a new capability, whether it’s a feature, function, or purely an economical benefit.
One “From the Archive” in the newsletter this past week included a column by the late, great, analog legend—Bob Pease—where Bob mused on how he would heat an electric car, and how he would handle the limited range and 30,000-mile battery life, as an engineering exercise. Bob was clearly intrigued; he clearly wanted an EV, but he could not come up with an engineering solution to conquer its ugly economics of the times he lived in.
Bob’s brilliant mind could have solved his EV problems easily if he had even a scintilla of insight on laptop batteries in a bi-weekly newsletter, or if he was inspired to let go of minimizing the heat in his motor controller...
I hope you seriously consider subscribing to the Automotive Newsletter and you enjoy the stuff I select for you to look at as its new editor. Being an engineer, I understand the time value of having someone curate what gets put under your noses. Please feel free to shoot me an email with a link if you spot something on the site that’s an oldie (let’s call it pre-pandemic, though if it’s really good stuff, newer is always a consideration), which you think would be interesting to those in automotive.
In addition to some spiffy analog stuff that they can try to pry away from my cold, dead hands, I also have a heavy focus on automotive battery-management systems (BMS) coming up soon. So, if you, or someone you know, is an expert in BMS design or systems architecture and are interested in contributing a 1,200-ish word article for everyone to read, please pop me a message with the concept and we’ll take it from there.
Have a happy and safe holiday. I’ve phase-shifted my days off to the week this blog posts, so I plan to be doing editor stuff during the last week or so of this month.
All for now,
-AndyT
AndyT's Nonlinearities blog arrives the first and third Monday of every month. To make sure you don't miss the latest edition, new articles, or breaking news coverage, please subscribe to our Electronic Design Today newsletter.
