Alternating Current For Peak Performance?

[knightmb]

I disgress, i dont really see ac as being more or less efficient than dc. The same basic rules apply. Ease in stepping voltages is really the only reason the power stations apply ac over dc.

Ease of voltage stepping, but also so the power line doesn't light up and melt.
The advantage of AC is passing more power through the wire with less resistance than DC. Otherwise the wires would have to be a lot thicker (and more expensive) to make up for the added resistance. That's the advantage of AC that most people usually think about when talking about AC vs DC (not the rock band )

The inverter to drive an ac motor really ends up looking like that of a brushless motor controller, the control algorithim is vastly different but the circuit is essentially the same, there wouldnt be any more loss in the inverter than there would be in the dc controller. As far as peak efficiency goes, the best ive seen go to brushless dc motors.
Joe

I see it the same way, but I think since the inverter has to do more switching (and thus produces more heat), the DC brushless controller would win in an overall best efficiently contest. My 12 volt DC to 110 volt AC inverter has a peak power output of 300 watts and gets so hot so fast it needs another fan inside just to cool it. My brushless controller on my e-bike does 48 volts @ 35 amps and doesn't even have a fan on it. Though I'm certain it pushes more than 300 watts of power out of it.

 

[Leeps]

There is the same rules of I^2R loss with ac or dc, there isnt any lower resistance, its ohmic it applies to current and resistance. In fact with high frequency stuff skin effect increases apparant resistance of a wire but thats not relevent. You are right they would need very thick wires to send dc across power lines but thats because they would be forced to send 220 volts across the lines. Today they have around half a million volts on the high tension lines and its a simple job to step it down to house voltage with a transformer so that you can use it.
The inverter that you had has a big transformer in it to give you the ac out. In a motor controller its sending out pwm and the motor acts as an inductive filter to make the sine wave in the motor so to speak.
My main point was that its always more efficient to pass high voltage and low current than low voltage and high current with ac or dc.
Joe

 

[knightmb]

My main point was that its always more efficient to pass high voltage and low current than low voltage and high current with ac or dc.
Joe

100% agreement

Reading back over my other posts, it looks like I was arguing against it, but I certainly agree with your point.

I feel it still boils down to what is driving the AC or DC motor. Good inverter or good controller that may decide which is more efficient.

 

[safe]

My main point was that its always more efficient to pass high voltage and low current than low voltage and high current with ac or dc.

Interesting.

I was under the impression that AC (since it doesn't actually move the "sludge", but just wiggles it) is more efficient than the "forced" DC current approach. I see how high voltage always "trumps" low voltage, but within the same voltage it just seem like there is an advantage to AC.

Given all things being equal AC is better than DC... but if you make things unequal then the higher voltage always "trumps" the lower voltage.

There are two "issues" that are getting "mangled" into one here.

OTHERWISE, why did the worlds electrical grid go AC and not DC? My historical recollection was that given the SAME circumstances you can always get a little better perfomance out of the AC approach... AC always has an "edge"... an advantage...

The super high powered lines I can understand would probably behave differently, but for the wires we use in our homes and in our electric vehicles the physics is roughly the same and the "powers that be" chose AC as the standard for homes.

So if it's "better" for the electrical grid why wouldn't it be "better" for an electric vehicle?

 

[fechter]

OTHERWISE, why did the worlds electrical grid go AC and not DC? My historical recollection was that given the SAME circumstances you can always get a little better perfomance out of the AC approach... AC always has an "edge"... an advantage...

You can't use a transformer with DC. Transformers are used to step up and step down the voltage so the transmission lines can use very high voltages to minimize loss.

For a given voltage, the loss in the copper is the same for AC or DC.

Another reason is 3 phase motors are brushless and can run directly off a 3 phase AC line.

 

[xyster]

Leeps answered that above:

Ease in stepping voltages is really the only reason the power stations apply ac over dc.

Also, if electrocuted, AC is less deadly than DC.

 

[safe]

For a given voltage, the loss in the copper is the same for AC or DC.

Really?

That's kind of "shocking" to use an electrical pun! :lol:

So the whole reason for using AC is to make transformers easier to use and that's it?

No advantage to AC at all?

It's actually HARD for me to believe that that... Hmmmmm....

 

[safe]

From Wikipedia:

"Power is the product of voltage × current (P = VI). For a given amount of power, a low voltage requires a higher current and a higher voltage requires a lower current. Since metal conducting wires have a certain resistance, some power will be wasted as heat in the wires. This power loss is given by P = I2R. Thus, if the overall transmitted power is the same, and given the constraints of practical conductor sizes, low-voltage, high-current transmissions will suffer a much greater power loss than high-voltage, low-current ones. This holds whether DC or AC is used. However, it was very difficult to transform DC power to a high-voltage, low-current form efficiently, whereas with AC this can be done with a simple and efficient transformer. This was the key to the success of the AC system. Modern transmission grids use AC voltages up to 765,000 volts."

This was really the reason... ease of high voltage distribution coupled with ease of voltage step down using AC.

So the bottom line is the DC Brushless motor using proper THICK wires from the battery should be equal to the AC motors if the controller circuitry are of equal efficiencies. So much for the "greatness" of the AC solution?

 

[Leeps]

Thats pretty much what i said in my above posts, no advantatge in ac.
In fact i can give a couple reasons why ac can be more lossy than dc. look up skin effect and inductive impedance
Joe

 

[Patrick]

This is an interesting thread, although there seems to be some confusion about AC and DC - both in the controller and the motor.

Let's start with the motor first - it's easier. There are two types. An induction motor (which I think is what you are calling an AC motor) has no windings or magnets. The field is produced by the relative motion of the rotor and the flux field. Therefore, in order to turn, the rotor always has a different speed than the control freqency, and this is called the slip. Usually less than control frequency, it can actually be higher to produce regen.

Then there is the synchronous motor. The field is produced by magnets or windings, and the rotor turns at the same frequency as the supplied waveform. This is the brushless motor (whether you call it AC or DC).

Now take a look at those supplied waveforms. Fixed, industrial applications use true sine wave voltages both for motors and generators. However, all traction applications use PWM, where the voltage is either on or off. Some of the confusion results around nomeclature - is PWM AC? Since standard brushed DC controllers use PWM, we need to make the distinction here - let's call it 3 phase PWM.

3 phase PWM has several advantages over single phase (brushed motors). Commutation is electronically controlled, and timing can be controlled dynamically. Motors are more efficient when timed for the speed, and direction, that they are running at. This gives a wider powerband to 3 phase motors, and let's them run in reverse at full power (think of a forklift application).

Another advantage to 3 phase PWM is that the modulation technique can be varied (even as the motor runs!) to suit the application. While modulating the PWM to sine wave duty cycle is the most common, there are other methods (3rd harmonic, 60 degree, stateless vector, etc.) that can deliver 15% more power than pure sinewave for a given voltage.

As far as voltage goes - the more the better. At any given wattage, higher voltage always reduces the resistive losses in the system. That's why all the big stuff (cars, buses, boats) typically run at 150-480 volts.

There is still the choice between brushless and induction motors. Induction motors (well, controllers, really) trade complexity for versatility. Wider powerband than brushless, no specific speed limit due to fixed flux density, simpler motor construction. Current hybrids use brushless, while high end systems (EV-1, AC propulsion, Tesla) use induction designs. This is also true of vendors - Siemens, Brusa, MES use induction (Siemens does make a brushless model as well).

If you'd like to know more, a particularly good text is "Electronic Motor Control Handbook", by Richard Valentine, published by McGraw Hill. Even if you skip the (extensive) math, there's a wealth of information there.

 

[Leeps]

Hey patrick could you explain the other pwm algorithims(or link to a site with more information). Id be interested to learn more about the 3rd harmonic, 60 degree, and whatever else exists as algorithims for ac motor control.
Joe

 

[safe]

The Point of Diminishing Returns?

What I've learned about all this is that there is a "Point of Diminishing Returns" where once you get to about 90% efficiency you can't expect to go all that much further no matter what you do. In the "real world" of the electric bike your top speed is almost entirely limited by wind resistance, so a little fiberglass work... maybe even wind tunnel testing... is going to out perform all the efforts you might throw into the motor.

Basically if you focus on good gearing and good aerodynamics and good current limiting then if your motor is "in the ballpark" of efficiency (at least 80% preferably closer to 90%) then you can't expect much better out of it.

You have reached the "The Point of Diminishing Returns".

 

[Patrick]

Leeps:

It's easiest to explain and diagram 3rd harmonic, so I'll give that a shot. The other methods have almost exactly the same improvement.

Let's suppose you are modulating the duty cycle of PWM at 200 Hz (a typical value). Then you add in a smaller signal of the third harmonic (600 Hz), you'll see a wider, flatter waveform. The ratios are a little tricky, I had to look them up. You have to boost the amplitude of the original sinewave beyond what would be 100% in order to compensate for the loss of the 3rd.

Hopefully I can atttach a .jpg (haven't tried it on this forum before) and it will be come clear.

Anyway, the area under the curve is 15.5% more than the original sinewave, on so more power to the motor.

Does that make sense?

 

[Leeps]

It does make perfect sense. If i was at a lower frequency i would have simply boosted the sine wave signal to get more area but this method would let you add area to the curve when you are already at full rail voltage, very interesting. Im not very well read on magnetics and cant speculate on any negative results of mixing in the third harmonic.
Very interesting i will be reading more on this.
Joe

 

[xyster]

That's fascinating -- and the theory is so simple and visual even I can immediately understand it ! Love improvements like that...

 

[fechter]

I haven't seen that 3rd harmonic setup before. Pretty cool. It's very similar to a trapezoidal wave. Usually the idea it to match up the waveform of the drive with the waveform of the back EMF. With a permanent magnet brushless motor, the back EMF is often a trapezoidal looking thing, so the 3rd harmonic drive might be perfect.

Is the 3rd harmonic setup intended to work with induction motors?

 

[Patrick]

Yes, 3rd harmonic is for induction motors, although I believe it could be used on synchronous as well. It is very much like trapezoidal, which is certainly used in the brushless world. I wish I had instrumented that when I had the chance. Unfortunately, Todd sold the Mars/Sevcon test platform we had - damn those sales guys!

60 degree modulation is easier to describe, but harder to diagram (an infinite series expansion). Imagine you have standard sinewave modulation, and it peaks at 100%. Now, increase the amplitude so that 100% is reached at 60 and 120 degrees (and, of course, all of this applies to 240 and 300 degrees on the negative side).

Then you clip the signal at 100% - i.e., you draw a flat line from 60 to 120 degrees. This also looks trapezoidal, and looks pretty much like 3rd harmonic too. The cool thing here is that you are no longer modulating - from 60 to 120 degrees you just leave the power stage on! This eliminates all the switching losses on one third of each cycle. Also for induction, I see no reason this couldn't be used with synchronous. But I'm going to have to think about that one, do some more research.

The vector control algorithms are complex, and not easily described or diagrammed (or really understood by me). One thing I did learn in trying to explain this is that slip can be controlled with precision in an induction motor by varying voltage, current and frequency. As a result, you can run the motor anywhere on the power spectrum, maximizing efficiency, power factor or torque, as you desire. It does require closed loop feedback, and the smartest of controllers - no wonder they're expensive.

 

[Malcolm]

Great thread! Thanks for explaining the differences between the motor types so clearly Patrick.

You mentioned the Mars brushless motor. If I understand right this will only run up to 48V? Is that just because there is no widely available controller that will take it higher, or is it a limit inherent to the motor. Just wondered because it has the same rpm/volt as the Etek and looks very similar in size and construction, so I guess it would be capable of taking more volts.

 

[fechter]

You could run a Mars motor with a Crystalyte controller. The only limitation on voltage would be if the RPM got so high the rotor flies apart.

 

[Patrick]

I think you could run it higher voltage, but I haven't tried it. We ran it at 48v with a 200 amp Sevcon. Performance was decent, but no match for a 48v Etek running 500 amps. I wanted to run it higher - but controllers that can do that are mucho dinero - and about a foot square, not easy to hide on a svelte motorcycle.

Another thing about that combo - we couldn't get the motor above 3000 rpm. Supposedly, it turns 70 rpm/volt - so it should have done 3300 at least. According to Sevcon, it was a limitation of the motor, and they referred me back to the designer. I didn't follow up in time, and now no test platform. Ah, well - I expect we'll have one again someday.

 

[fechter]

As I recall, you could never get it to draw over 90 amps or so at 48v.
To get more power, you would need to either increase the voltage so it could draw more current, or rewind the motor with fewer turns.

I wonder if Mars offers any different winding configurations?

 

[Malcolm]

I wanted to run it higher - but controllers that can do that are mucho dinero - and about a foot square, not easy to hide on a svelte motorcycle.

Just curious – if brushless controllers are relatively simple compared to induction motor controllers, isn't it just a matter of upgrading components for the higher voltage and cranking up the frequency for the Mars motor. Why do the controllers suddenly get bigger and a lot more expensive?

 

[fechter]

Good question. Even the guts that go into an induction controller aren't that much more expensive. The power stage is essentially the same for all 3 phase controllers. It seems to be a supply / demand / volume thing.
Most of the larger units are made for industrial applications, where the pockets are deeper. The RC model market is much closer to where we should be. But they don't run very high voltages.

 

[Patrick]

I misspoke; let me explain.

The Sevcon put out 200 amps; the problem was getting it to put out more. With new firmware, we got it up 210. But that was all.

I wanted to run it higher - but controllers that can do that are mucho dinero - and about a foot square, not easy to hide on a svelte motorcycle.

What I meant was to bring it up to the level of the Alltrax, 500 or 600 amps.

There are two reasons these controllers are expensive and big. One is the sheer power. The models like this also handle higher votlages, and are capable of puting out 30 - 40 kW. Inverters in this range just aren't cheap.

The other is product mix. For these light equipment vendors (Sevcon and Curtis) - these are their top of the line products. So you get all the bells and whistles - CANbus, vector control, super programmability. In fact, the Sevcon espAC is so smart it auto calibrates to your motor. It runs some static tests using pulses, and then spins the motor to compute the rest of the values - you're in perfect tune. How cool is that?

 

[Malcolm]

In fact, the Sevcon espAC is so smart it auto calibrates to your motor

Ubercool! That one definitely goes on my Christmas wish list. I don't suppose they come up on ebay very often do they...

So it looks like brushless motors have great potential for mid-range EVs (I'm thinking motorcycles and lightweight cars). The real attraction for me is the one you mentioned in your Zappy thread, fechter. With a brushless system your vehicle simply stops if the controller fails, rather than taking off and trying to tear your arms out of their sockets.

So all we need is someone to start making a brushless controller rated to, say, 100V and 300A, without all the bells and whistles of the espAC. I'm sure Xyster could fit one of those to his handlebars with some nice silver duck tape.