Why do EVs have a single speed transmission?
I’ve seen online some people are claiming that you can have a 30% efficiency boost due to a multi-speed transmission, and even a new commercial product by Inmotive claims to add 15% to range. There are gearheads who love the concept and proclaim that it’ll become standard. I respectfully disagree with all of the above.
Let’s think about this before even getting into the details. It takes ~16-20kW to move an electric vehicle on the highway at 100kph (~60mph). If you could have 30% increase in range with a better transmission, that would mean that at least 23% of the energy spent would have to be lost in the drivetrain due to gearing, right?
Without even doing any research, let’s walk that one through, and – Reductio ad absurdum – let’s assume that we can save that entire 23% and bring the losses down to 0% by using a more efficient drivetrain.
That would mean that 3.7-4.6kW of energy is being lost just to the transmission.
Where exactly is all that heat going? Why wouldn’t every EV be capturing that energy and heating the cabin? EVs would be perfect winter vehicles! (Pro-tip: when something seems obvious and you can’t figure out why everyone else doesn’t understand, chances are you are the one who doesn’t)
The other obvious thing to consider – if it really was that much better, surely everyone would be doing it already. We have multiple companies who have spent the past two decades working on this, trying to squeeze out maximum range, and are all producing vehicles with single speed transmission, with the exception of a few niche high performance vehicles. Tesla, Ford, GM, VW – all using single speed transmissions.
So the answer here is simple – the premise must be false. Even with Inmotive’s 15% claims, this just doesn’t even pass the smell test. While perhaps we could contrive a worst case scenario where 15% savings could be had (I will cite one below), this absolutely won’t hold up for normal usage.
How much energy does it take to drive?
To answer this and get to the truth, let’s first let’s try to figure out exactly what energy is required to drive and the breakdown. There are many articles on this that conclude roughly the same amount, but this one is neat because it talks about all the ways that we actually can save up to 27% energy and increase range (spoiler: multi-speed transmissions aren’t one of them).
Electric and Conventional Vehicle Performance over EcoDriving Cycles: Energy Benefits and Component Loss
TL;DR – Out of a standard driving complete cycle (across city, highway, etc), it takes 147Wh/km to move your car. Of this: 12Wh of the energy required to drive the vehicle is from the drivetrain (8%). Where does the majority of the energy go? Aero 30%, Rolling Resistance 25%, and Motor/Inverter 27%. Note that this is for a total average overall driving. Rolling and motor losses don’t change that much with speed, but aerodynamic losses are proportional to the square of the speed. Around 55mph they’re ~50% of total and by 70mph they are ~65% of the total energy spent. The other losses are proportionally much less at those speeds – and that’s really the only time we care about range. When we do, drivetrain losses will be less than 8%. Also, air is more dense when it’s cold – up to 20%. This is a good chunk of why winter range is worse. But that’s for another article.
So already, right off the bat, even if we reduce the drivetrain inefficiency to zero, we would gain about 8% only for average city driving and less for highway driving. This can be backed up by many articles, but here’s another look:
Fact #882: July 20, 2015 Hybrid Vehicle Energy Use: Where Does the Energy Go?
This is for gas/hybrid – but it shows that drivetrain losses are 3% for gas driving. Using their numbers, if we scale this to EV driving it’s 7% – and these numbers independently line up with the study above. Note that this increase is not due to increased inefficiency with EV driving – it’s because the engine losses go from 66% down to <10%, so all other losses are larger.
So how efficient is the drivetrain?
Note that the general age-old “15% rule” of transmission power loss in a gas vehicle is outdated and also based on automatic transmission vehicles with all wheel or rear wheel drive. Here’s a decent primer on that. So right off the bat, worst case full multi speed transmission and rear wheel drive we’re at 15% loss.
Digging into the details of the drivetrain efficiency some more: Drivetrain Losses (efficiency)
They calculate the drivetrain:
- A full 6 speed manual transmission is about 94% efficient.
- Driveshaft is 98% efficient.
- We don’t have a rear differential or propeller shaft which would add 9% to the losses (this is where the majority of that “15% rule” is from)
Obviously a single-speed transmission will be more efficient than a full 6-speed. But even with that, we’re under our 8% from above.
So we’ve determined through multiple sources that drivetrain losses are no more than 8%. But what about the motor?
Why do gas vehicles need a multi-speed transmission?
A gas engine can have less than half the efficiency and power outside of its sweet spot. So to accelerate you want to shoot the engine RPM up (at the cost of efficiency), then bring it back down again for steady speed driving.
You can see that for steady driving, lower speed higher torque (higher gear) is more efficient. This, incidentally, is why having a small motor to generate electricity which is used to drive the wheels can be more efficient than driving directly – because you can always run the engine at it’s peak efficiency, something that is very hard to do when driving normally. Note that the absolute peak efficiency of a gasoline engine is typically 40%, but as you can see above, normal driving is in the 20-30% range. It’s commonly cited that a gas engine is about 15-20% efficient overall.
The purpose of this article isn’t to get into those specifics – I’m only mentioning it so that we can see why gas engines need a multi speed transmission.
What about a modern electric motor?
Let’s look at the Chevrolet Bolt’s published motor efficiency map:
This shows at ~10% torque and 70% motor speed (aka steady highway driving) we’re at 95.25% motor efficiency. Note that there’s very little difference across the entire usage – the range is mostly 94-96.75% efficient. There’s basically nothing to gain here from a lower RPM. Halving the RPM on the highway would go from 95.25% to 96.25% efficient.
Now we need to point out that a 1% absolute increase in efficiency of the motor is proportionally large due to how efficient the motor is already. That would take the motor from having 4.75% losses to 3.75% losses. That’s relatively speaking a large 21% decrease in energy lost, which at ~27% of total energy used, would translate to 6% less overall energy used. But to do this, we have to increase the complexity (and losses) in our mutli-speed gearbox. We’ll definitely lose a chunk of that.
Putting it all together
So here’s what we can prove – drivetrain is ~95% efficient. Motor is ~95% efficient. Inverter is more efficient at higher loads (more on this below). There’s very little variation in those efficiencies based on speed – maybe 2% combined. Even cutting down total energy used by the motor by 21% saves 6% to our bottom line, but the increased complexity of the transmission will eat some of that. So we might see 2-3% overall savings. Remember – eco-mode can get up to 27% if we really want to save energy.
What about actual research?
There is a lot of actual research on this. If you still don’t believe me, here’s an actual research paper that actually looked at this exact scenario.
Improving electric vehicle energy efficiency with two-speed gearbox
This is a great paper – because it directly talks about how much has changed recently. It discusses previous research showing that with more inefficient motors, a 2-speed gearbox could save up to 10-11%, but how that is not applicable anymore with the vastly improved motors that we have. A lot has changed in the past 10 years.
The net results are fascinating. The overall total efficiency gains of just adding a 2-speed transmission to the Leaf is at most 0.93% more efficient and at worse 0.06% less efficient depending on the drive cycle.
Tweaking the gear ratios specifically to the type of drive cycle (obviously useless but is interesting academically) gives us 0.75 to 1.49% improvement. Even assuming that we can further optimize the motor and inverter and find an optimal pair of gear ratios, the best they can do is 1.8% to 3.7% improvement. That’s it.
Why is this? Look at the motor/inverter efficiency map either in the article or what I posted above. It’s already extremely efficient and more efficient at moderately higher RPMs. Drivetrain losses are already very low. While the motor may have a 1% loss in efficiency at higher RPM, the inverter has a 2% efficiency GAIN. So the motor+inverter is 1% more efficient at higher speeds, offsetting some of the drivetrain gains.
Note that this study is done with a Leaf and a 80kW motor – our modern motors are at least twice the power. Having said that, the efficiency overall of the Leaf motor and inverter is similar to modern vehicles, and as I showed above, similar to the Bolt. It might even be a bit higher. Nissan did a good job.
Here’s another research paper showing where we can get up to 20% losses – if the motor is applying low torque at 20,000rpm. At the ~5,000 maximum RPM that we normally use, efficiency is around 95% (yet again backing up everything we’ve shown above).
Lots of other research – but be careful
Also note that if you look hard, you can find some research papers that claim up to 10-12% improvement. But, as always, the devil is in the details. They’re either old, purely theoretical and based on really weak low-speed motors, or very inefficient motors that do have different efficiencies at different speeds – I’ve seen a lot where they assume motor is 80% efficient. But note how modern motors are incredibly efficient across their entire speed band. Both the Leaf and Bolt motors graphs are available above.
We can see the highest performing vehicles today mostly have single speed transmissions
It’s also useful to look at the Tesla Roadster or Model S Plaid – 0-60 in 1.9 seconds and no multi-speed transmission. But that’s also going to run you a cool quarter mil. Yes, multiple motors and gear ratios, but point being – no multi-speed transmission.
With all this, we can see that there is no reason to do this for every-day performance and maximum efficiency gains are negligible. If you want up to 27% more range, eco-mode optimizations can give you that.
So where could a multi-speed transmission be useful?
Audi and Porche seem to think so. There’s an article that says “Why the Porche Taycan EV’s two-speed transmission is a big deal” but it offers no substance as to why. So why would we want to?
People aren’t used to the smoothness of constant acceleration
First obvious answer is – people who love to drive, typically love either a manual transmission, or the feel of a multi-speed transmission. One of the really odd things about electric vehicles is while they are fast, you don’t feel it. We’re so used to the jerky push-back-in-my-seat then forward then slammed back again that we have equated that with performance. In the Bolt you floor it and it’s just smooth all the way to 90mph. You’re at 60mph before you even realize it. It just doesn’t feel the same. So that’s one reason.
There’s also something to be said for that synergistic feeling of slamming the clutch down, dropping three gears, and having the perfect balance between applying the gas pedal and releasing the clutch to seamlessly slam you into high acceleration. I’ll admit, I miss that.
Weaker motors for heavier vehicles
There’s an interesting research paper (The influence of multi-speed transmissions on electric vehicles) which concludes that you can get 10-15% more range with a multi speed transmission. But like I said above, the devil is in the details. Looking at their setup, they have a heavy vehicle with a weak motor, and are looking at 0-100kph times of 13.9 seconds with a 9.52:1 gear ratio and 170-185Wh/km. They show that by using a multi-speed gearbox with a 6:1 secondary gear they can drop that to 152-158!
So this seems like it could be a useful place to use them – put in a cheaper motor with a mutli-speed transmission! … So why don’t companies do that?
Notice how not only are modern vehicles are much more efficient already (~150Wh/km for Bolt/Leaf) but also have a much lower gear ratio (7.05:1 for the Bolt). This results in most of the efficiency gains that they’re seeing already present in our single speed vehicles. We have more powerful and efficient motors (0-100 in 7.0 seconds) so there’s no need to use a multi-speed gearbox to gain that performance back. Note that they show the performance of the transmission goes from ~93% to ~91% with a multi-speed transmission, which is a 29% relative increase in energy consumption. There is a cost to using a multi-speed transmission.
So getting back to why don’t they put in weaker motors? Because with the average selling price of a new vehicle around $40k USD, the motor is less than 3% of the sale price. On the Bolt it is estimated by UBS that the motor+gearbox cost was only $1100 with low economies of scale. So why skimp on a weaker motor (saving only a few hundred bucks) to have to add the complexity and cost of a multi-speed transmission (which still doesn’t get you back to the same performance)? It just doesn’t make sense. Higher performance motors are cheap enough with selling prices high enough that they’re worth it.
We could get faster acceleration
It’s true – but as I just said, today’s motors are pretty good already. Keep in mind that the vast, vast majority of people don’t care about 0-60mph times. Considering that for gas cars, 0-60mph sub 7 is considered quite good performance. Sub-6 is usually reserved for “performance” vehicles. A basic electric car can do this already – so we’re already “fast” without needing to add the complexity of a multi-speed transmission.
Also keep in mind that if you want more performance (since we know that adding a rear transaxle costs 9% in efficiency), simply add a second motor to the rear wheels. Then you gain all-wheel drive and more performance without the losses and cheaper and more reliable overall – and without adding a multi-speed transmission.
MotorSports will definitely benefit
Certainly on the track, or drag strip, where you are designing vehicles to go up to 20,000RPM or so, a multi-speed gearbox will absolutely be of benefit. As we showed above, you can get down to 80% efficiency in those cases. Great for acceleration but bad for efficiency.
There will be a niche market
Having said that, I can see a niche market for sports EVs who want 0-60mph in sub-4 seconds but still want to remain cost competitive. Especially as people get used to EV’s higher overall performance and want more. There will be a market for sub-$40k cars that can do sub-4 0-60. But don’t forget, motor design is improving daily as well, so we may still get there without any need for a multi-speed transmission.
Some more useful viewing material:
Why do electric cars only have 1 gear? (only 5 minutes – they get into when multi-gear transmissions are useful; spoiler: racing only)
If you want to see how simple an EV transmission is (4 minutes for the in-depth description):
Professor John Kelly at Weber State University tears down the Bolt motor and drivetrain
As always, I encourage anyone to do their own research. This is only a sliver of what is available. However, most of what I have read and seen lines up with what I’m presenting here. Always be critical and analyze the research before using it to conclude things.
One thought on “Why do EVs have a single speed transmission?”
I am absolutely agreeing with you. A multi-speed transmission will not boost efficiency by 30% (maximum 0.5%).
Two speed transmission is used in Formula-E due to motor power limitation (200kW). A gear box is torque amplifier and highest gear ratio allows to use full motor power at low speed.
Anyway, modern electric car has total efficiency close to 72 -74%. And it can be improved to 95%. That means, a total car efficiency can be improved for 30% (battery is included).
About gearbox efficiency. Why in many electric cars is used liquid cooling system? In the rear will drive cars the differential temperature rises 15 – 20C above ambient temperature. 0.6-0.8 kW power is not enough to raise temperature for this difference at 65 mph speed where massive differential is chilled by high-speed air flow. Also, a gearbox of automatic transmission has independent radiator. In some cases, a manual gearbox has radiator too. It means, total loss of drivetrain (differential and reducer) exceeds 8%. Is this energy usable for heating the cabin? – No. 20C difference is not enough for the cabin heating.
In the attached chart Motor plus inverter loss are in the range 20-40Wh/km. It means motor efficiency is in the range 73-81%. I know one motor only, that has similar efficiency. It is DC brushed electric motor that was used in the first electric car by Gustave Trouvé in 1881 year. The efficiency of modern motors with inverter is within 86-96%. In the shown chat, efficiency distribution between the motor and the drivetrain is wrong, but sum is OK (24-34%).
About motor efficiency. Why PMSM is needed rotor liquid cooling if motor efficiency is 96% and main loss is in stator? Neodymium permanent magnet has an AC loss at low magnetic flux density, usually at level 10% (<0.1T) of Bmax. AC of magnetic flux density in this case are high frequency harmonic distortions that converted in permanent magnet directly to loss and torque ripple.
Exists different type of solutions that help to reduce the energy loss in electric car powertrain:
1. Gearboxes and differential manufacturing from super hardest and super polished material like it is released in Mercedes VISION EQXX. In this case, the price of the gearbox and differential is much higher than the price of the electric motor. Mercedes claims it’s efficiency at level 95% W/O battery with energy consumption 100Wh/km or 170Wh/m. I am skeptical about 95% efficiency because it done mainly by drag coefficient 0.17 and lightweight design. Also, it has low power (147kW) and estimated price $160,000, that is not interesting for potential customers.
2. All wheel drive solution with PMSM motors W/O reducer and differential (real direct drive). Below are two examples, where is no middleman (reducer and differential).
– Lightyear One – It has 4 PMSM build in wheel with liquid cooling. Energy consumption 154Wh/m, drag coefficient 0.175, total power 130kW, estimated efficiency 90~92%. Its efficiency looks realistic. Price is $296,000.
– Pininfarina Batista – It has 4 PMSM build in wheel with liquid cooling. Energy consumption 387Wh/m, total power 1417kW, 0-60 acceleration – 1.8s, stated efficiency – 85%. Its efficiency looks realistic. Price is $2.9 million.
In Pininfarina Batista example we can see the permanent magnet issue that is damaging car efficiency. Larger motors use larger permanent magnets that are mostly underloaded in the WLTP test (mostly 55 mph). As a result, motors have high rotor losses due to high harmonic distortion. In addition, the high magnetic flux density of the rotor creates an alternating magnetic field in the stator and, as a result, additional electromagnetic losses and back EMF. It doesn't matter if it's disabled or not.
– Permanent magnet motor is not usable for creating efficient and powerful car at the same time, because it has limited torque capability and reducer is required. In the future, golf cart design will be the main application of permanent magnet motor.
3. What is the best candidate for electric car motor? At a power generation plant usually is used wound rotor synchronous machine that has efficiency 98-99%. Unfortunately, it is brushed. It can be used as synchronous motor with torque capability that exceed 4 times of PMSM torque. In this case reducer is not required if motor is powerful enough. By the design manipulation it can be converted into brushless Double Fed Synchronous Machine that torque is sufficient to power an electric vehicle without a gearbox and differential in an all-wheel drive application with accelerations above 1.1g. Why 1.1g? It is a traction limit for best performed tire. For this power train configuration total efficiency is 95% (battery is included). Mechanical brakes are not needed, because DFSM has full torque at 0 speed. Cooling system is needed for battery, only (at supercharger) and has energy loss significantly less. Energy recuperation is much better. The car body light weight design is not required that can reduce the car price. Rare Earth materials are not used in the motor. Driving range can be increased or battery size can be reduced for the same driving range and total car weight reduction.
As a result, mid-sized passenger car can have energy consumption 150Wh/m at performance of Tesla Plade and 400 miles driving range at the standard Chevrolet Bolt battery capacitance.