As Electric Motors Impove, More Things Are Being Electrified
economist.com
economist.com
Tesla and Steinmetz figured out the theory of AC machines in the early 20th century. This was the first industrial technology where you needed advanced math to get anything to work right. Complex numbers. Calculus. Laplace transforms. That bothered some people, such as Edison. By the 1930s it was figured out, and generations of EEs struggled through rotating machine theory in college.
Then all the cool kids went off to radio and transistors and computers. Rotating electrical machinery was a mature field. Maybe a few people at GE in Schenectady worked on it.
Then came power semiconductors, and chopper-type motor control, where power was being turned on and off at moderately high frequencies. At first this was just applied to existing motor designs. But AC motors were designed for sine wave power. Choppers didn't produce clean sine waves. Much effort was put into making variable-speed controls that produced the nice sine waves motors needed. Classical AC theory was built around sine waves, and engineers knew how to do that kind of analysis. This worked, but it made AC motors buzz at the chopper frequency, and as chopper frequencies went up, whine. When you ride on BART, that's what you're hearing from the motors. The waveform mismatch also led to unwanted heating in the windings and inefficiency.
Variable-frequency 3-phase AC motors went from exotic to normal. Today, everything from a Tesla to a drone to a Diesel-electric locomotive uses such motors. The big power semiconductors required aren't that big. Here's one for a locomotive.[1]
In recent years, motors have started to be designed for the non sinusoidal waveforms that come out of chopper power supplies. This required new theory and much simulation of magnetic fields. There's plenty of compute power available and commercial packages for that kind of analysis. Now we're seeing more advanced motor designs that match well with their control electronics.
After most of a century, rotating electrical machinery design is cool again.
[1] http://www.ametrade.com/eng/electronics/products/IGBT_IGCT.s...
There are some trains on the London Underground that use gate turn-off thyristors, which make a sort of Star Trek klaxon sound when they start up:
Closely followed by dynamic braking, or as it is more accurately and much more awesomely called in the UK, rheostatic braking.
https://www.reddit.com/r/electricians/comments/6vpg8n/the_me...
Although the Northern line trains have an almost identical body design, they use a more modern IGBT traction system which, although presumably more efficient, has a little less character!
The locomotive doing the towing is a 2006-model Electro-Motive Diesel-electric locomotive. It's towing the old steam locomotive at a very slow speed, smoothly, with no wheel slip. Nothing special about it; Union Pacific has hundreds of them. Six axles, each powered by its own motor. AC synchronous, IGBT power control. The motors are run as servomotors, all locked together. No wheel can slip relative to the others. This is a huge win when you need to get a few thousand tons moving.
This boring but useful modern locomotive, although much smaller, has more drawbar pull than the Big Boy. Multiple locomotives can be ganged together with wheels locked in sync, if you need more pulling power. Here's Union Pacific overdoing that.[2] They put together a train 3.5 miles long, with 9 locomotives spaced throughout the train, and ran it from Long Beach to Denver. This was just a test; they don't normally run trains that long.
All this is made possible by modern electronic motor controls.
[1] https://youtu.be/JRoIrFgcIjc?t=894 [2] https://www.youtube.com/watch?v=jdIzRFOaTCY
Sounds like there is a steam engine inside (but the 451s are electrical)
http://www.nytimes.com/2009/02/21/nyregion/21about.html https://www.youtube.com/watch?v=WL436Qddigw
From what i remember, square waves are better for a motor. Thinking about it i see no benefit of it being sine (*with many poles, that is, as a square would shake the mechanical parts too much otherwise).
The trams here have 3-phase engines.
Good driver chips (e.g. https://www.trinamic.com/) still synthesize sinusoidal-like waves for these motors. It reduces noise and power consumption.
It's always interesting to observe differences in history from different vantage points.
In Europe the story told goes like this: Some time in the early 19th century people noticed that a thing called induction exists. Then in the later 19th century a chap called Ferraris (also known for his AC energy meter) developed multi-phase AC and the general idea of an induction motor. Some bit later, a Russian guy named Dolivo Dobrowolsky, invented a two-phase induction motor. Meanwhile, some other folks started to build transformers. Not even a handful decades later and virtually everyone in Europe used AC and both synchronous and asynchronous generators and motors were used everywhere for everything in industry.
The reason for these differences is of course that in these few years key advancements were made by different people at the same time.
VFDs are very complex pieces of circuitry- far more than you'd expect. Efficient drivers require a great deal of computation, and 16 or 32 bit processors are not uncommon. That also requires high-power, cheap silicon transistors which are only gradually taking over from simpler control schemes. They make a huge difference in a lot of cases.
The author is also very wrong on Synchronous Reluctance-assisted Permanent Magnet motors, but its hard to fault them on that; it's complicated even for many engineers. The purpose is not to increase the power density, it's to increase efficiency. The magnets act like a "cruise" motor. At low torque, they provide all of the rotor magnetization at a very high efficiency. At higher torque, the stator induces a stronger field into the rotor, causing it to act like a reluctance motor. That allows you to turn on extra power as demanded at the price of lower efficiency (the same as a reluctance motor).
If, instead, you just used a larger PM motor, it would be more expensive and it would also have an efficiency drop at low torque (where the motor spends most of its time operating). The magnets are highly efficient but they "set" the operating torque of the motor somewhat, so there is a loss at low power/high speed to hysteresis. A reluctance motor meanwhile never reaches the peak efficiency of a PM motor.
Anyway the article doesn't really say much convincing and feels mostly like fluff.
Disclaimer: I work at Visedo.
Commutated DC motors could also be used but they are low efficiency.
ive always wondered what it would be like if you used a dual motor system in a car, where one motor is wound and sized for very high torque and the other motor is wound and sized for very high speed. i think it would be great because, as long as they were induction based motors, you could run one and leave the other off with no interference from the one that is turned off. it would be like having a transmission without any of the energy loss or maintenance problems. you could also distribute power however you want among the two motors, and in a way have something like an infinitely variable transmission. that would be really cool.
Traction motors for locomotives and transit vehicles used to be built with multiple windings and switching. The windings were switched from series to parallel as speed increased. Locomotives once had manual "transition controllers" for this.[1] Here's the more automatic mechanism from a PCC streetcar of the late 1940s.[2] The operator just has an accelerator pedal, and all that control gear takes care of the complexity of operating the motors.
[1] http://cs.trains.com/trn/f/741/p/249329/2783262.aspx [2] https://www.youtube.com/watch?v=pgSWuTxQh6s
The dual motor versions (D) of Tesla electric vehicles (models S and X at least) use this principle. As far as I know, they have induction motors. However, I am not sure to which extent one of the motors is used primarily for high torque and the other for high speed.
They shut the rear motor down at highway speeds giving the Dual Motor cars slightly better range.
Or do they have a third kind of motor which is wound to be a better balance?
Single motor, non-P: 1 large motor in the rear.
Dual motor, non-P: 2 small motors in front/rear with front motor geared for cruising, rear for accel.
Dual motor, Performance: 1 small motor in front geared for cruising, one large motor in the rear.
There's really no reason to have a front motor only. It's done in ICE cars to save cost/complexity since the engine is up front. With how the center of gravity shifts during accel you want the drive wheels to be on the rear since they have better traction.
Same reason in reverse you have large disk brakes on the front wheels and drum in the rear.
But you need a powerful motor to act as the brake (you want to avoid friction braking as much as possible). And that needs to be in the front.
You also want friction braking for cases where you've got 100% charge or the battery is cold and you can't dump energy into the pack.
I'll argue front wheel drive is better / safer for most driving situations most people find themselves in, especially in slippery conditions, on wet or dirt roads.
That said the traction control on Tesla is responsive to < 1ms. Lots of throttle inputs that would cause traction to break free on a traditional ICE due to momentum in the drivetrain doesn't on our Tesla. Heck I can floor it in the rain and it doesn't step out at all.
Battery-powered motors almost always operate in the range of 80 to several hundred kHz, which is basically unavoidable. Those make up a minority of motors in your home. Wall-plugged motors use AC, in which case they have no high-frequency harmonics, or they are DC in which case the ultrasonic noise comes from the brushes scraping and arcing on the commutator. This is the source of noise in vacuums and the one you're most likely to hear but it's basically static at a very high frequency. Supposedly this is one of the reasons pets hate vacuums, because it's a loud noise of a rarely-heard pitch. Vacuums will be much louder than almost any other motor in your house.
The difference between cheap an expensive power supplies is often down to the use of fully integrated PSUs. The expensive ones operate in the MHz, which lets them use very small capacitors that are built into the chip. Cheaper supplies operate at closer to audible frequency, but they aren't actually louder despite using larger components. The efficiency is relatively close to a more expensive supply so the amount of energy loss is also similar.
>ive been meaning to buy something that will let me detect ultrasonic sound so that i can smash those devices with a hammer.
It's actually very easy to make one! Cheap (<50 cents) electret microphones can hear up to 100 kHz and are pretty common in DIY bat detecting microphones[1]. Note that you can't use normal ultrasonic sensors like those used in rangefinders (about 1 cm wide, black plastic cylinders). Those are heavily tuned to resonate at 40 kHz and can't pick up sound outside that range at all.
>ive always wondered what it would be like if you used a dual motor system in a car, where one motor is wound and sized for very high torque and the other motor is wound and sized for very high speed.
That's exactly why 4wd Tesla models have a longer EPA rated range, or at least speculated to be why, anyway. The EPA range test is done at a set speed and Tesla optimized on that by having one motor for lower speed and one for higher speed. It's also useful for vectoring torque to the front/back for braking/acceleration. You can't use it like a variable transmission though, since there's very limited benefit to splitting the power up.
The benefit in practice is pretty unclear since you won't be driving the exact speed it's optimized for and the difference in efficiency is only a few percent anyway.
[1]: http://www.wildlife-sound.org/equipment/technote/micdesigns/...
I hope to see the day when projects like this could be cost effective.
Reluctance motors don't use permanent magnets or rotor windings and don't have rotor currents. Universal motors (which until very recently were absolutely ubiquitous in appliances) are commutated rather than inductively coupled and also don't have motors. There are a half-dozen other types but they're mostly novelties or variations on a theme.
Take a note, while SRMs are very cheap, some vendors would bundle a some kind of constant torque motor driver circuit at very little extra https://www.alibaba.com/product-detail/60v-72v-1200w-switche...
Let them study all the things!
Size is where the majority of advancements for BLDC motors will come.
They are essentially an offshoot from the decision made nearly 20 years ago: Lappeenranta University of Technology decided to focus on energy tech, and added environmental tech as a second leg a few years later. Visedo is one of the results.
I also remember, while their offices were still in the tech incubator warehouse, trying out their prototype electric "car". Lightweight, all sheet metal, uncomfortable to sit in and steer - but went from stand-still to ~50km/h in no time at all. Looks like they've managed to scale up and refine their technology quite a bit since then.
In general the magnetic permeability of materials is 1-3 orders of magnitude higher than the electrical permittivity. Materials with high permeability are also 2-3 orders of magnitude cheaper than materials with high permittivity. It makes much more sense to make magnetic motors because of those two facts.
However charge-based motors have existed since the mid 1700s[1]. They're exclusively novelties, but they are exceptional novelties indeed. You can even use a kite to power them from atmospheric electricity[2]! That's the same voltage difference that eventually creates lightning, but without a storm the electricity is quite weak and the motor quite inefficient. The voltage is high enough to be unpleasant though.
> As Electric Motors Impove,
Assuming you have three 60W-equivalent light bulbs on in your appartment, using halogen lamps your lights use as much power as the PS4. If you still use incandescent lamps, that's 180W, about as much as much as the PS4 and a 55" LED TV. With 7W LED bulbs you need 5 hours of lighting to match the PS4, or 8 hours to match PS4+TV.
Of course you might have a computer that uses significantly more power. But the best current-get Intel i7 still has a TDP of only 112W (165W for i9). Add to that a GeForce GTX 1080 TI with a TDP of 250W. Even during gameplay both will on average use less power than their TDP. Without going with dual-GPU configurations you can't build a reasonable current-gen computer that matches the 500Wh that our three 7W LED use over 24h. And I suspect you actually have more than three light bulbs.
1: https://en.wikipedia.org/wiki/PlayStation_4_technical_specif...
I think that's a spelling error. ;)
In general storing mechanical energy at the nanoscale is not efficient. As you get smaller mechanical properties become less relevant due to the increase in relative surface area (aka square-cube law). That makes electrical and chemical properties much more powerful.
Muscles are also excellent at storing energy in spring-like fibres that stretch/relax. But I can't think of any manufactured materials at that scale.