Hacker News new | past | comments | ask | show | jobs | submit login
In Mahle's Contact-Free Electric Motor, Power Reaches the Rotor Wirelessly (ieee.org)
137 points by hsnewman 32 days ago | hide | past | favorite | 67 comments

That's a common trick in large generators. The field winding rotates, and the armature winding produces the output. So the heavy power doesn't flow through brushes.

So how is the field winding energized? There's a second smaller generator on the same shaft, with the armature winding rotating and the field winding stationary. This generates AC for the field. You want DC for the main field winding, so there are also diodes in the rotating system to convert the AC to DC.

The shaft floats on a self-pumping oil film. So there's no moving metal to metal contact in the machine. That's a good thing in something that's expected to run 24/7 for decades.

So now this power-station sized technology has been downsized to car size. Unclear if it's worth it, since the lifespan of cars is only a few thousand operating hours, not the hundreds of thousands of hours of a power station generator.

I don't think that's the concept used. It's more likely to be a salient pole synchronous machine with the field winding supplied by an rotating transformer and diodes (the second illustration in the linked article seems to corroborate that). This concept has been proposed at least since the 1960s, to the best of my knowledge, but is only viable with modern power semiconductors. The hyped-up style is typical of american "technical" articles about electrical machinery.

Thank you for the comment. As someone ignorant of the field, the entire thing sounds like “turbo encabulator” to me! This gives me lots to Wikipedia and learn.

Both you and and animats are describing the same thing, brushless synchronous machine, which exists for long time.

I expect at least 10,000 hours out of an electric car, and maybe up to 30,000. They should be more reliable than an ICE.

Cars are commonly engineered to be driven for 8000h which corresponds to 15 years lifetime. Different story for commercial vehicles of course.

Fun fact: cars are commonly designed to be operated for 8h in reverse gear over lifetime.

Another fun fact: Electric drills are used for an _average_ of 8 minutes over their lifetime.

> Electric drills are used for an _average_ of 8 minutes over their lifetime.

I did see that somewhere, and I think it was in the context that every home has an electric drill, but most might only get used for 1 or 2 projects a year.

You would hope though that this might be only a design guide for one of those $30 stocking stuffers.

I'm an owner builder and I am pretty sure my drills and impact drivers have racked up many tens of hours.

Also "used" in this context clearly just means "spinning". I did a modest project the other day involving my electric drill. It involved 20 holes, each hole drilled in two steps (pilot with a jig in place, full depth with the jig removed), for 40 spins.

I'll guess that each spin took about 3 seconds (because that makes the math easy), so 2 minutes total. I couldn't exactly share the drill with a neighbor in the approximate hour that I spent measuring and repositioning the jig between uses.

In the context of duty cycle lifetime, maybe 8 minutes isn't that low -- but I am definitely using it for more than 8 minutes.

Using a hole saw and cutting into metal, I'm pretty sure I could spend 8 minutes per hole. Of course I prefer to use the drill press for this, but some projects can't fit and must be done with a hand drill.

Then there's the grit wheel or flap disc, which might have me axially loading the drill for 30 minutes at a time (such as for stripping paint or sanding rust off an iron engine block).

I think they must be including a lot of people who buy or receive a drill as a gift and never use it at all, to get to 8 minutes average. I do find it believable that the cheaper ones might have a design life of 8 minutes, but that's a sad state of affairs.

Yeah, that was just the most recent project. When I had to drill through tile back in November, it was definitely several minutes per hole (on the slow speed dictated by the special tile bit).

You're probably right about the stat including people who never use the drill at all, or only ever use it to tighten screws.

What wears out first in an electric car is the battery, I would assume.

I would probably assume it would be ball joints, bushings, bearings, AC components, seals, body corrosion, paint, or something similar. An ICE vehicle with 300,000 miles on it is usually only still on the road because the motor works. Everything else is usually completely shot

Their failure mode is continuous though and the bigger the battery, the smaller the absolute impact of wear. What people fail to grasp is that yes right now batteries may only go up to 200 miles for an affordable EV but if EVs had the same range as ICE cars then their useful lifespan would be even longer than it is right now.

It is unlikely that the battery would wear out faster than other key components in an EV or hybrid.

There is plenty of real world data from tesla model S owners showing that the batteries easily last more than 200k miles, which is about how long cars in the US last before getting scrapped. https://maartensteinbuch.com/2015/01/24/tesla-model-s-batter...

The same is more or less true for Priuses, their batteries last around 150-200k miles.

What I take from that graph is that there is not enough data to say what happens beyond 50k miles.

Really? The data in the second graph appears to corroborate with the testing graphs that Tesla themselves have released, which is what the first graph is.

I don't think they would have used brushes in this application. I'm quite certain that the most popular choice in this space is a squirrel cage induction motor [1].

[1]: https://en.wikipedia.org/wiki/Squirrel-cage_rotor

Tesla used induction motors and even got funding from Toyota to develop them as a rare earth free design. But permanent magnet motors can be more efficient, and newer Tesla motors use permanent magnets. (Unless I’m very much mistaken, Tesla’s PM motor has no windings in the rotor.)

There are plenty of motor designs with neither slip rings nor brushes. I don’t see what’s special about this supposedly new design.

There seems to be plenty of R&D around these days for developing rare earth free motors which are better than the standard induction motor (which, per se, isn't bad, but if you can do better, why not).

Here's an EU project that designed and compared induction, synchronous reluctance, and synchronous reluctance with ferrite magnets in the rotor: http://www.refreedrive.eu

Brushless alternators are (or were) also fairly common on motorcycles.


I'm sure there's gotta be something new here, but what the title and intro describe could apply to induction motors, which date back to the 19th century. That makes the whole thing sound rather silly.

What is new is that it has a high efficiency similar to the permanent magnet motors, i.e. a higher efficiency than induction motors.

The reason for the high efficiency is because the motor seems to be some kind of synchronous motor, where the rotor is not supplied by wearable brushes but by a contactless transformer and a rectifier.

Besides the permanent-magnet motors and the induction motors, there are 2 other kinds of electrical motors without supply contacts on the rotor: variable-reluctance motors and hysteresis motors. However these other 2 share the disadvantage of the induction motors of having a lower efficiency.

The highest efficiency with no contacts to the rotor can be reached with either permanent-magnet motors or with designs similar to this Mahle motor.

That's what I eventually figured. But the angle really made the writer sound like they had no fucking clue about electric motors in the first place.

There is a video here that explains the concept: https://www.mahle.com/en/news-and-press/press-releases/mahle...

If you're like me, and thought this was basically a squirrel-cage motor, then they key point is that this is a synchronous motor which replaces the typical permanent magnets with electromagnets.

But I wonder whether the poles in the rotor (which would ordinarily be permanent magnets) are DC or alternating field. If DC, then perhaps the rotor has a built-in rectifier? If alternating, then the stator coils must be doing some interesting compensation.

The experience curve for electric motors, power electronics, and batteries is going to be extremely interesting to watch over the next decade or two. The change over the last fifteen years has been incredible already.

any good sources for further research on electric motors evolution ?

This showed up on HN a few months ago with good discussion on that https://news.ycombinator.com/item?id=26224709

I don't really understand what problem they're trying to solve. Contacts don't really wear out if they are tightened correctly and there was definitely no need for this on the 3-phase side of the driveline. But you add induction coils like this and you're nearly going to need a medium as big as the motor itself to transfer that energy over an air gap.

It's also dumb to say 'no wear, we removed the last interface!!'. The motor is full of bearings. It actually seems to have twice as many bearings as a normal motor. So yes, you will need to replace all those bearings and it will be a right pain compared to just tightening a bolt for each of the phases.

The article states another benefit of the motor is that it doesn’t require rare earth elements, so presumably they’re trying to solve the ecological and supply problems presented by rare earth mining.


Yeah, it sounds like they're saying they get the efficiency of a permanent magnet motor (which can be around 95% or so) without using magnets. Induction motors I think tend to be in the high 80's.

I don't entirely buy the rare earths thing. You can buy permanent magnet EV motors without rare earths, like the Netgain Hyper9 [1] (which I'm using in a conversion I've been working on for awhile). The hyper9 is kind of heavy for its power density, though, so maybe that's the advantage of rare earth magnets. If this company can get the efficiency of a permanent magnet motor but without rare earths and at a high power density, maybe they've got something new and interesting.

[1] https://www.evwest.com/catalog/product_info.php?products_id=...

edit: another advantage of avoiding magnets is that they tend to demagnetize if they get too hot. Not having magnets means you might be able to run the motor hotter without damage. Though, a 95% efficient motor should be pretty easy to keep cool.

the contacts they are referring to are brushes to tranfer energy to the rotating shaft not bolted connections for stator leads

This is a common tech, with many variations of such induction used.

What is important is that the wireless power transfer is lossy. It's not very efficient.

Designs that do not require permanent magnets are good, but I believe that using good quality carbon brushes reduces the complexity of power transfer and is reliable enough for use case of cars.

Carbon brushes also have isolation issues. If your motor is to be air cooled, you will be blowing outdoor air through it, complete with fog/condensation. In brushed motors, that fog will cause a leakage current from the motor brushes to ground. The leakage current may be tiny, but in a system where you want 10's of megaohms between the high voltage system and the chassis for human safety, it becomes a problem.

Or you could try to hermetically seal your brushes in to stop moisture getting in, but now you have the difficulty of a spinning moisture seal and build up of carbon dust which is probably going to start to get warm/hot.

> What is important is that the wireless power transfer is lossy. It's not very efficient.

And that is why their reported efficiency of 95% is significant.

We have DC-Motors with brushes on work (50kW-500kW). Brushes really need a bit care all the time. And if something happened, you have to rework the whole thing on a turning machine.

Permanent magnets are handy for spinning up a power grid from zero, are they not? Many generators need power to produce power, so they’re useless if the grid is down, until they get fed some power by something else.

Starting up power grids from zero is called black start[1].

Synchronous generators can have external exciter systems, or be self+excited[2].

[1] https://en.wikipedia.org/wiki/Black_start

[2] https://en.wikipedia.org/wiki/Excitation_(magnetic)

Many induction generators are 'self starting' - even a tiny bit of the earth's magnetic field is enough to get them generating something, which in turn makes them generate more, etc.

Would it really be the earth's magnetic field? Residual magnetization of the steel laminations seems more likely.

They speak about 95+% efficiency, I'd hope they included this wireless connection in the calculation.

Dumb question: Couldn't you also reduce wear-and-tear considerably by placing the contacts at both ends of the axle, so they lie in the center of rotation?

This would mean you have to transmit mechanical energy from the motor using a gear or a belt - but that has to be done anyway if there is a transmission, as I understand it.

Nice to see that Mahle, which at least to me is known mostly as a supplier of pistons (and possibly other ICE parts), is preparing for a post-ICE future.

Haha, I like the wording in the URL

Soo.... They invented an induction motor?

They are at least 100 years too late.

This was my first thought too. And then I remembered reading so many entries at the USPTO's website where they simply wrote about a 100-year old technology, obfuscated the heck out of it with convoluted language, and got the patent through. The odds that this company is doing the same thing are pretty good.

Patents are a joke.

One of my old jobs holds (software) patents on my work. I built some pretty clever software, but I wouldn't have said that I "invented" anything.

The actual content of the patents is absolutely ridiculous too. Here's an excerpt from one of the patents on my work:

> There is also provided a system for identifying software capable of capturing personally identifiable information, comprising a module arranged to connect to a remote server via a local device, the module further being arranged to send at least one request for data and retrieve at least one packet of data in response to the request, and an identification module arranged to receive the retrieved at least one packet of data and determine whether the received data originated from a third party server other than the remote server, and if so, determine whether the received data received contains evidence of the presence of personally identifiable information

It's almost indistinguishable from nonsense claims generated by GPT-2 [1].

Also, half the claims in the patents were for features that, in the end, were never even implemented.

But worst of all, I'm not even credited as one of the inventors. I don't care one iota about the intellectual property, they can keep all claims to that, I just want the bragging rights.

[1] https://aipatent.wordpress.com/1000-samples/

It's actually not an induction motor. It's likely synchronous, with the wound rotor actively energized.

Oh, bingo


That's a rotary transformer excited synchronous motor. A bit better, they are now only 90-80 years too late.

This isn't a physics discovery. It's an engineering application. Engineers rarely invent new things, they improve upon existing things.

Find me a brush-less, magnet-less electric motor like this with a 95% transfer efficiency.

Simple induction motors easily get to 95%


> Find me a brush-less, magnet-less electric motor like this with a 95% transfer efficiency.

Turntide gets to about that level with their high-rotor-pole switched reluctance motors, I think:


This pdf of a thesis covers the design.


Here and elsewhere found two issues.

A problem is efficiently driving the rotary transformer due to losses in the drive circuit because air gapped transformers have a lot of leakage inductance. I take that to mean they throw a lot of reactive power back at the drive electronics.

Hidden problem they are trying to solve is these motors are typically/often oil cooled. Oil and slip rings don't play well together.

Thanks for the link, it is interesting.

However the conclusion about the low efficiency of the electronic converter for the rotary transformer is no longer valid.

With modern switches of silicon carbide or gallium nitride and with resonant soft switching techniques, the converter should operate at much higher frequencies and at much lower losses.

This is what has probably enabled today the manufacturing of a high-efficiency synchronous motor with a rotary transformer by Mahle.

We agree that issues with the inverter efficiency were fixable with modern designs and components.

I'm on this train of thought that the service life of electric power trains is 2 to 5 times that of gasoline ones. There is a potential to design cars that last 30-50 years. Which potentially reduces the environmental impact of vehicles. I've noticed though that people in the automotive industry look at you like an moron when you suggest that the service life of cars may double or triple. I think it's so ingrained they can't think around it.

I don't understand this attitude. If a company released an ICE engine with a different configuration than the standard that had been hypothesized and prototyped 100 years ago would you be singing the same tune?

> hypothesized and prototyped 100 years ago

It's not hypothesized and prototyped 100 years ago, it's the most common type of medium power electric motor in the world.

Even if the idea is old you still have to build it in a commercially viable fashion i.e. it has to be competitive with brushless motors.

Not to mention the fact that EVs sparked an interest in rediscovering existing motor technology.

It's not an idea. Externally excited synchronous motors are well widespread.

Seems that a journalist got sucked in by marketing hype.

I had expected better of the IEEE. Actually, now that I look at other articles from IEEE Spectrum, most of them seem full of uncorrobrated and unwarrented hype.

The innovation here is not pure concept but having designed a product with 95% efficiency and low cost, low pollution components. As one of the major automotive part suppliers Mahle presumably has developed that with some consultation to the closely located Porsche and Mercedes engineers and has the manufacturing capabilities to scale this up. This does not seem to be a toy effort and the newsworthy bit is that the broader automotive supply chain seems to be seriously pivoting away from ICE new.

We're talking engineering here not physics. It doesn't have to be a radical new concept for it to be interesting and useful. Across many fields, new applications of existing concepts have proven revolutionary and disruptive.

> Unlike internal-combustion engines, electric motors have practically no moving parts.

WAT. If they have no moving parts, why even have a motor at all?

As other commenters have already said, terrible, terrible article. I expected more from IEEE.

brushless electric motors spin using magnets rather than pistons, so there are less required moving parts to get the one moving part (the shaft) that you actually want.

Electric motors, just like any other motor have moving parts. End of discussion. Their purpose is to move things.

Less moving parts doesn't mean "practically no moving parts", as the idiotic article claims.

Coincidentally, the brushes which wear out on DC engines are not moving, it's the commutators that move and wear out the brushes...

...but cars use BLDC motors (which are induction motors or variants thereof) and not DC motors anyway... Brushed DC motors are only used by forklifts and similar equipment (although more recently they started to move to brushless designs too).

A DC motor and an induction motor have the same number of moving parts. It's not the number of moving parts that matter, but the fact that moving parts don't have to touch each other.

I believe the difference in moving parts between a combustion engine and electric motor is about two orders of magnitude (20 vs 2000).

I would say "practically no moving parts" is an ok term in relative comparison but thats just my opinion.

There isn't really a debate, as far as size, weight, complexity and maintenance electric motors win over combustions engines.

The problem of course is the battery size cost and weight compared to a gas tank.

Bottom line though electric cars have much less maintenance and much of that can be attributes to much less moving parts.

Big difference between electric motors and piston engines is sliding friction. Even the journal bearings in a piston engine are only partially lubricated. Which means they wear wear quickly. Compare with fully lubricated bearings which have a service life that's 10-100 times longer.

The wording in the article is imprecise with regard to the number of working parts, but everybody knows that motors move things so your point is needlessly pedantic.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact