Electric Vehicle Design

[bill541]

To be honest Jacque, I don't fully understand it either. See the post at the bottom with the 3-Phase AC drive schematic for some more thoughts. Maybe somebody who understands 3-phase motor controllers with regeneration can explain what is going on and how it works. -Bill-

[bill541]

Hi Guys, I understand how a 3-Phase motor gets up and running with a PWM input signal applied to the upper FET to pull the phase leg high with the other phase leg(s) being pulled low with a PWM input on their respective lower FETS.

I can also visualize that if all FETs were off and the motor was still turning, that the body diodes and freewheeling diodes form a bridge rectifier and as long as the motors peak-peak voltage was above the supply rails, current would flow back into the battery source.

The problem I am having difficulty understanding is how reducing the drive frequency to the FETs to a point below the free run frequency of the motor would cause regenerative braking. Say for example the motor was turning at a rate to generate a 60 Hz frequency and the controller was switching at a 50 Hz rate. Does the phase angle difference between the controller drive and the motors frequency cause an increase in motor winding current to the point that when the winding field collapses, the peak to peak voltage goes above the battery voltage, thus charging the battery?

Is this how they work or am I missing something?

On another note, couldn't this "active rectification" be used on 3-phase induction generators to make them produce power across an adjustable rpm range?

-Bill-

[thunderhead]

Technically, financially and legally this has problems: roughly equivalent to trading your engine every time you fill up.  You don't know how the new engine has been driven; you're not sure you want to be near equipment that whizzes around that much hot, heavy, dangerous payload; and the industry probably doesn't want standardised engines anyway.

I say we dump filling stations and install electric points in roadside cafes.  A 230V  32A three-phase plug (say using a CEEform connector) provides about 25kW - which is a kWh every 2 1/2 minutes.  If you have lunch, and take 90 minutes over it, you should 80% recharge your 50kWh car, and (according to the simulation results of my car project) get another 170 miles range at 65mph, or another 150 miles at 75mph.

So if I set out to visit my brother in Edinburgh, in my lithium-powered car, I might set out at 6AM, to avoid the worst of the M25, and get past Birmingham around 8AM.  We could break for breakfast at Sandbach Services near Congleton, by which stage the batteries would have about 10% capacity left in them.  During breakfast it would be drawing about 25kW off the grid, which would mean that it would reach 80% capacity about 10:10, and start trickle charging.  By the time we've eaten, and fed our son, changed his nappy (diaper) and used the restroom, we'd come out about 10:30, and the batteries are 90% charged.  We rejoin the M6.

Around lunchtime (say 12:00) we'd arrive in Scotland.  Maybe 12:30 we'd break for lunch at Tebay Services, north of Kendal, by which stage the batteries are back down to 10% again.  Another meal, another leg-stretch, another diaper, another 90 minutes on the charger and we're ready to go again.  

Since the last part of the journey, avoiding Glasgow and through the outskirts of Edinburgh is not motorway driving, by the time we arrive at my brother's house, maybe 4pm or so, the batteries still have 20% left.  Now we need to plug into his house current, which will take all night to re-charge the car, because it'll only do so at 2.2kW.  But that's all right, we've covered a days driving and gone 420-odd miles.

Now I know that the US is bigger than the UK, but do people really drive 400 miles without bothering to eat or even use the restroom?

[thunderhead]

A separate controller for each motor is exactly what I had in mind.  A friend of mine who races suggested the steering input to the controllers, so you can drive around corners with the traction control helping. He seemed to think it would be competitive at rallies. He seemed to think my mother could be competitive at rallies, especially if you entered the pace notes into the computer.  Unfortunately FIA/WRC regulations are heavily biased towards ICE. :-(

Going back to those DAF and Volvo vehicles, they used independent CVT for the two back wheels.  They tended to scrabble a bit on full lock at slow speeds, when both transmissions reached their lowest "gear", but other than that they were fine for cornering, even though they had a separate transmission for each back wheel.

Most 4WD vehicles use a "limited slip differential" which does something similar: with a normal "diff" if one wheel spins there is no torque applied to the other.  They also seem to get by OK - most WRC cars use limited slip diffs and 4WD.

So I guess that in practice separate drive will not be a problem.

As for gearing, I missed a step out of those calculations.  56Nm at 400A is 2200N at the two front wheels, if you use a 6.5:1 ratio.  That leads to a top speed of 78mph, a 0-60 time of 14.5 seconds and a standing quarter in 21 seconds.  That may not be particularly quick, but it's no Trabant.

Replacing the batteries with 42kWh of "Thunder Sky" lithiums changes that 0-60 time to 11.9 seconds and that standing quarter to 19.2.  (As well as giving nearly 200 miles range at 65mph.)

Then adding a double motor to each back wheel and dropping the ratio to 5.5:1 then gives a top speed of 92mph, drops the 0-60 to 4.6 and the quarter to 13.4.  That will see off almost every four-footed beast on the roads.

It doesn't have to be a Trabant. :-)

[jacquesm]

That would work.

btw I only meant swapping out the batteries, not the drivetrain, but I see your point.

And yes, people really do drive 400 miles without stopping (there being next to nothing to stop for, interstates are like long straight lines with gas stations being the only distraction from the usually gorgeous scenery. Just the parts that I've looked at).

In fact, I'm guilty of that myself. I live about 650 Km north-west of Toronto and I have made the trip at least once without stopping at all (needed to get someone to the airport in time and we left late).

Are there any ev's out there right now using Lithium batteries ? The honda had a pack of NiMh batteries, but since it's a hybrid that hardly counts.

What are the powerdensity / lifecycle / deep discharge characteristics on Lithium compared to good old lead acid ?

[jacquesm]

Looks like you've got something there. I just threw in the Trabant for comparisions sake, if people were happy driving that thing (in the absence of alternatives) that means it was able to hold its own in todays traffic. And it still does, there are still plenty of them around. Also the 'trabbi' has some of the trimmings of ev's, such as an all plastic body !

Now we're getting on topic, my mom (yes, I know) had a series of Daf's when I was a little kid, a 33 ('daffodil'), and two '55's. We had great fun in the 55 because it outpulled many a sportscar in those days from a standstill. No gears and a fair sized engine (1600 CC).

She drove them for so long - and they drove so well - that she got hooked on CVT, nowadays the nearest thing she can get is an automatic which is what she's got now.

Back in NL Daf's are known as 'elderly lady's cars'. Volvo bought out DAF (does that mean Ford now owns DAF ??), and produced a 343 for a while that had the CVT in it, but it was discontinued.

The only CVT cars on the road at this time are japanese ones. They use an advanced version with much higher RPM, a downgearing on the far side so it's much more compact than what DAF used and a steel belt instead of a rubber lined belt.

Pretty impressive technology, and very compact.

Smart guy that mr. Van Doorne.

[thunderhead]

The 200Ah battery that Thunder Sky make stores 0.6kWh in a unit that weighs 5.5kg, is a shade over 2 litres in volume, and costs $250. Best battery life is obtained by discharging them to 50%, which gets you 1500 cycles.  Discharge to 80% (the maximum rated) gives 500 cycles.  At 200 miles maximum range, that's 150,000 miles at 50% discharge per cycle or 50,000 miles at 80% discharge per cycle.  The Peukert number is tiny.  Charging is similar to lead-acid: that is, CI then CV, but carefully.

The only production car I know with lithium batteries is the T-Zero, but there are other prototypes and kitcars out there.  Many others.  Most of them use the TS batteries, because right now they are the best.  But lots of others are announcing lithium automotive batteries about now.

I guess in big countries like Canada you want a generator unit to support long range.  I guess you don't drive 400 miles every day.

http://www.acpropulsion.com/Products/Range_extending_trailers.htm

[thunderhead]

Think of the motor as a generator delivering a voltage that is less than the batteries require.  So to charge the battery you need six boost convertors, which you time to coincide with the movement of the armature.  That means you need to know the position of the armature, but you need a rotation sensor anyway to run the speedo.

The six boost convertors are

1. D1, Q2, L1 (+ve)
   
2. D2, Q1, L1 (-ve)
   
3. D3, Q4, L2 (+ve)
   
4. D4, Q3, L2 (-ve)
   
5. D5, Q6, L3 (+ve)
   
6. D6, Q5, L3 (-ve)
   
   

You use them in pairs or threes, so one or two provide a negative current while the other one or two provide positive current.  The currents are set to always sum to zero, of course.

I think a circuit very similar to this was suggested for keeping windmills at their most efficient rotation speed.

[jacquesm]

OK, I'm not an AC whiz, let's see if I got this straight:

The frequency with which you drive the coils is the same as the one with which you drive it when you want to speed it up, you just time it so that the PHASE lags behind what the motor is doing at the moment. Sort of like an electric brake.

And because you have your phase lagging instead of leading you are actually drawing power FROM the motor instead of putting power into it.

The key difference with the way this was described previously is that they frequencey is the same, not lower. At least not right away, but as you draw energy from the motor it will slow down, and then you *will* have to adjust you frequency to match in order to maintain the same phase lock. That would be an ongoing process.

iirc correctly this is how the best stepper motor drivers on the market work.

Is that it ?

[thunderhead]

That's the way I understand it.  The difference in phase between voltage and current is what causes the motor to produce torque: put the current ahead of the voltage and the torque is in the direction of motion, and put it behind the voltage and the torque opposes the motion.  In the first case, the electrical machine is a motor, and in the second it is a generator.

If the motor was travelling quickly enough you'd be able to use the motor voltage without needing to up-convert it: but then you'd be travelling faster than the car was designed to go.  Since you want your batteries to be of a voltage in excess of the peak voltage produced at top speed, to get recharging you need those "buck" convertors to get the charging current to flow.

You see why it might have applications for windmills?

[jacquesm]

It's beginning to dawn on me. This ac stuff is tricky. I think I'll order a book on this and study it for a bit, then I'll be able to make more sense out of it.

My electronics experience is limited to building and designing up to 400 Mhz transistor based transmitters and lots of digital stuff, and most of that was ages ago.

This electromechanical stuff is (apart from the absolute bare minimum) pretty new to me and I must admit quite confusing at times.

I understand basic motor/generator equivalency, and I understand the relationship between phase, voltage and current at a basic level (including power factor), but this is really over my head.

I do see I need to get this under my cap quickly if I want to be able to keep up here. Right now I'm really walking on my toes and that doesn't feel very comfortable, I like to 'know my stuff' and I really would like to participate in a fruitful way in this discussion. I can hardly expect you to provide me with a free education though

To quote the 'terminator', I'll be back (on that), but it could take a while...

Meanwhile if I feel I have something useful to say I'll just butt in.

[thunderhead]

If you've built transmitters then you're familiar with complex loads.  A motor is a complex circuit where ω is the rate of rotation of the shaft in radians per second.  Think of a tuned system but you can see the energy phasor going around. :-)

If you try and drive it above its (present) frequency, then the current leads the voltage in phase - if you try and drive it below its (present) frequency then the current lags the voltage phase.  If you drive it at exactly the "tuned" frequency, then the amount the current leads the voltage depends only on the resistance in the system: the "Q" of the circuit, if you will.  With a perfectly frictionless AC machine driven at its rotational frequency, current and voltage will be perfectly in phase, and the load will be the resistance of the coils.

It's back to all that cos(ωt)+j.sin(ωt) nonsense that we did in college.

Here is a link to some pages by a prof from the college where I did my electrical engineering:-

http://www.ee.surrey.ac.uk/Personal/D.Jefferies/powerac.html

[jacquesm]

> f you've built transmitters then you're familiar with complex loads.  A motor is > a complex circuit where ? is the rate of rotation of the shaft in radians per
> > second.  Think of a tuned system but you can see the energy phasor going  
> >around. :-)

Yes, I can see the similarity there.

> It's back to all that cos(?t)+j.sin(?t) nonsense that we did in college.

Maybe it wasn't nonsense after all then eh ? The thing is with transmitters that is exactly what you try to avoid. Any transmitter that you want to live for longer than a few seconds will have to have it's impedance matched to the antenna or the standing waves are going to reflect right back from the antenna wire into the guts of the transmitter to be tranformed into heat. (besides giving you a very dirty signal on the air due to self interference). A little bit of mismatch is allowed (1:1.1 or so) but not a whole lot more at any appreciable power levels, and preferably less. One surefire way of smoking any kind of HF equipment is to rip out the antenna wire.

In motor/generator situations that heat just gets added up to the losses, since you are much more able to dissipate it (and you usually have some good airflow going to carry it away)

I never spotted the analogy between those two before, thank you !

[bill541]

I have spent several hours over the last few days researching VFDs (Variable Frequency Drives) and how regeneration works in a 3-phase AC motor/generator. Haven't found much in-depth information yet. But so far, this is what I have found interesting.

The AC motor has two principal currents;

One called magnetizing current (Id) which is almost all inductive and lags the voltage by 90 degrees. This current is typically between 20% and 60% (33% average)of the full rated motor current. If a motor is run with no load, most all of the current is magnetizing current +/- friction and I2R losses.

The other is called the torque current (Iq) which is proportional to the amount of load applied to the motor. This current is in-phase with the voltage.

As far as regeneration is concerned, all explanations still say the the VFD drives at a frequency below the free run frequency of the motor. A speed sensor is sometimes used to ensure the drive frequency is always below the motor frequency for even braking.

In my mind the VFD drive current would only need to be able to satisfy the magnetization current requirements. Current begins to flow backwards in the drive circuts as the motor gives up its mechanical energy. I would think that the torque current is what you are recapturing.

It would be nice to see a "simple" vector analysis of the voltage and current in the motor and VFD during regen, maybe then it would make more sense.

-Bill-

[Tyler883]

Lets put the power lines, which are as extensive as roads are - in cities, at least - lets put them in the roads so vehicles can draw their power without having to recharge or have high cost batteries.

Heck, I'll bet that even large gasoline powered generators is a better solution than millions of idiots riding around in 6000lb guzzlers.

We can then decide if we want to by an electric, or something that is a hybrid that runs 'off-grid'.

Personally, I like the idea of owning a full time electric car with a motor rated for 100,000 hours, and NO BATTERIES! At least one of the 2 vehicles my family owns would be fulltime electric.

cheers

Tyler

[thunderhead]

There has been a suggestion to make the road and the bottom of the vehicle into two halves of a giant transformer, to (say) charge buses up as they wait at bus-stops.

However there is a certain amount of medical evidence suggesting that long-term exposure to non-ionising radiation causes leukemia and various other nasties.  The people who live next to roads (most of the population) would be living in a transformer, and would be constantly exposed to levels of NIR which many would consider unacceptable.

I think for most two-car families having one of their cars run electric is perfectly possible, even with today's technology.  Since that is then the car that can be used for the heaviest commuting, we save over 50% of automotive CO2 emissions at one stroke.  

[Tyler883]

"I think for most two-car families having one of their cars run electric is perfectly possible, even with today's technology.  Since that is then the car that can be used for the heaviest commuting, we save over 50% of automotive CO2 emissions at one stroke. "

I tend to agree with you. However, I get descouraged when I think about my personal situation...

I'm thinking of taking my beat up old Honda ( a 3rd car I didnt mention) and converting it to electric( with batteries). It has some features you often don't see in modern cars like fresh air vents in the dash.

Unfortunately, a project like this is bound to cost me the same amount of money that I have tied up in my 883 H-D motorcycle(a 4th vehicle I didn't mention - LOL), and my harley has a epa rating of 58 miles per US gallon. As long as my options include the purchase and replacement costs of batteries, I think I will always have more incentive to stay with the motorcycle. The irony is that I consider myself to be a good fit for an EV car(I'm a city commuter), yet, I can't seem to make one fit into my collection of vehicles.

still looking,

Tyler

[thunderhead]

You can electrify bikes, too.

[Tyler883]

good god, man! there's nothing wrong with a Harley engine!

<grin>

regards

Tyler

[nack]

Must be one of them newer 883's then. <g>