Sunday, April 17, 2011

This week we return to Slingblade's Dynomometer for another test. I'm a little weary of test graphs this week as I've been making a number of them. But this is probably more interesting to you.

High Performance Electric Vehicle Systems http://www.hpevs.com has kind of cracked the AC drive market with the only system we're aware of for less than $5000. They produce a sturdy little 50kw motor with a face mount matching the popular Netgain Warp 9 which it makes it very easy to install with available adapter plates.

They don't do the controller at all. They use a Curtis model 1238 3-phase AC controller. Curtis controllers are widely dismissed by the cognicetti in EV land in favor of more exotic offerings such as our just reviewed EVnetics Soliton1 and the Cafe Electric Zilla.

But if we use EVALBUM's 3218 conversions as a representative sample, we find over 50 different controllers in use but 27% of ALL listings are Curtis powered - 878 vehicles. Since probably a third of the vehicles are bicycles, motor cycles, and lawn mowers, this is probably more like 50% of the market. Number two would be Alltrax at 276 and the long time darling of EVland, the Cafe Electric is a distant third with 155 installations. Our recent find, the Soliton1 has 22 vehicles listed.




Curtis makes controllers for industrial electric vehicles, floor washers, fork lifts, etc. We've talked to them and they profess to be oblivious to the Electric Car end of the world and claim to want nothing to do with it. But their products are migrating in that direction with brake light signals, regenerative braking (on a fork lift?) and more.

HPEVS has done an outstanding job of integrating the Curtis with their AC line of motors. The 50 kw AC-50 is the most powerful and most popular of these. And they've quickly gained a foothold. We used the HPEVS system in Speedster Duh to considerable effect and Duane Ball used it on his Spyder 550 as well.

Just this past month, Curtis began shipments of a new model 1238-7601 controller which is a bit of an upgrade from the 1238-7501 we had been using. The 7501 was nominally a 96 volt controller with a current limit of 550 amperes. We routinely run this controller at 120v and understand that it can withstand voltages up to 130 before shutting down to protect itself.

The new 7601 ups the current limitation by a full 100 amps, to 650 amps total. If you could put 120v in at 650 amps, that would render 78 kw. If we are low enough to get it to start, we actually sag to a little over 100 volts at those current levels and ergo, perhaps 65-70kw are truly available. That's potentially 94 eHP input to the motor.

ALL the other AC systems available that we are aware of START at about $5000 or $6000 for the controller and perhaps the same or slightly less for the motor - typically with a 30kw motor rating and 100-125kw on the controller. To get to 150-200kw in the Tesla/AC Propulsion range, you are in the $15,000-$35,000 range buying a really untested undeveloped one off prototype dream of some crackpot and it just makes no economic sense in an electric vehicle.

The advantages of AC are widely touted as regenerative braking, efficiency, and durability.

We have tested regenerative braking on both HPEVS and the MES-DEA products and were heroically underwhelmed. The oft cited 10% or 15% gains are simply nonsense - entirely myth based on observations of output and input. Real world tests of routes in the same car with it ON or OFF provide results of 5-7% AT BEST and this is very much a function of the driver and terrain. I actually generally get negative numbers when I'm driving.

As we see in this set of tests, the efficiency theory seems to be blown as well. If you take the motor and controller as a unit, they appear no more efficient than our series-DC and chopper PWM controllers - indeed the numbers would indicate LESS efficiency.

As to durability, it again makes sense. But we have no data. There are no brushes to change. Certainly.

But a funny thing happens in AC induction land. ALmost all the high end systems from Siemens and MES-DEA and UQM etc seem to be high voltage, comparatively low current devices. You rarely see anything beyond 300 or 350 amps through the windings of these motors. Instead, voltages are normally used to achieve higher powers- usually STARTING at 300v and going as high as 1000v in some systems.

This poses some problems for an electric car. To get such high voltages, you wind up with an extremely large number of cells. Our Mini Cooper has 112 cells in series, each of 100Ah. Further, all the peripheral equipment has to work at that 375 volts as well, our air conditioning compressor, our DC-DC converter, our water heater, and the list goes on.

HPEVS has carved out a niche as the ONLY sub $5000 AC system out there and it does it at LOW voltages and HIGH currents. Very unusual and a characteristic of their motor. If they can get a controller to do 650 amps and a BIT higher voltage, say 200v, we could in fact see 130kw.

Rumors exist noting that Curtis is CONSIDERING (not really working on) a 144 v model of this controller. That would be very useful.

In the interim, this 96v 650amp model tests out very well as you'll see in the accompanying graphs. We did a full 650 amps and upped our HP on the rollers to 74 from about 63 just by doing the controller swap.


















As you can see, the -7501 model offers improvements of 15 to 25% depending on what you're looking at but torque jumps as much as 25%. If nothing else, this considerably widens the size of vehicle this system would work on - probably very effective at up to 3000 lbs. It's relatively low price under $5000 (we don't actually have pricing yet on the new model), relatively easy connection and configuration, make this drive system a contender for most electric car conversions we've seen. You will not get the torque or horsepower of an 11 inch series DC out of this package, but it is very sturdy and offers the advantages of AC if driveability is your thing instead of racing. It would still be under a little strain in our 3500 lb mini cooper. But at 3000 lb and below, with a transmission for leverage, this is a good package that got better.

Enjoy

Jack Rickard

35 comments:

  1. Very clear on the graphs how the voltage sag from the higher current gives a narrower, though higher, torque curve. Here's hoping they actually produce that 144V controller to stretch those peaks out a little more.

    ReplyDelete
  2. For sale:
    One Curtiss 1238-7501 550Amp controller. As new. Teehee.

    A good video. Shows the advantages of DC over AC are not borne out on the road. Yet every man'fr of EV's all seem to use AC induction motors. I guess for fear to appear second player to all the others. Not unlike why they have regenerative braking.

    Funny that... "gas" car's are positively oozing with parts that need replacing at service times. Never mind replacing brushes possibly once in a lifetime!

    Could it be RFI issues forcing the game to AC?

    ReplyDelete
  3. I am very impressed with the low voltage high amp system. I am impressed with the vehicle and impressed with the setup that is an off the shelf AC system that is low voltage and uses common parts found in DC systems. It is now in the realm of many to do an AC system and have a good motor and controller that is more or less affordable and you don't have to have hundreds of volts or hundreds of cells installed. DC like but AC. Nice.

    ******Could it be RFI issues forcing the game to AC?******
    No, AC is being touted for regen and that is still a selling point and it also allows for a higher cost system. There is no need for such a high voltage system anymore and you still get the regen if you really want it.

    ReplyDelete
  4. Not to start the regen debate all over again the fact remains that for average drivers in most conditions regen gives greater benefits than Jack sees. Could those differences be minimized by better driving techniques? Sure, but the general public isn't that interested in trying to change their driving habits. Notice that when other people besides Jack drove the car with and without regen they showed a real benefit from regen. Add in steeper hills and more stop and go driving and the benefits will also increase. Since AC has regen built in for free there is no reason not to use it. Also since AC can operate at higher RPM's running a single speed gear reduction instead of a transmission is easier. AC can be water cooled, series DC, not so much. Finally, since AC can easily run at higher voltages than DC, wiring can be a lighter gauge. The higher voltages are not the same problem for OEM's to deal with that it is for us DIY types. Oh yes, and no brushes to deal with. I'd say there are plenty of good reasons for OEM's to use AC or BLDC motors over series DC.

    ReplyDelete
  5. We've been through all this. Regen isn't what everyone thinks. The gains just aren't there. But it IS kind of a perception issue.

    All OEMs use AC Induction motors. And one of the OEM engineers did patiently explain it to me. It is entirely about public perception. You can't be in a side by side magazine review with a DC series motor and no regen when everyone else has AC and regen. You just LOOK to out of it. BUt they are well aware that the sales claims for regen are simply bogus.

    That said, the HPEVS is a nice little AC system and now working into useful power levels.

    jack Rickard

    ReplyDelete
  6. Bottom line, what OEM wants to be compared to a forklift?

    They don't care what motor, you are the one that will pay for it.


    Jack Rickard

    ReplyDelete
  7. First, drivers DO learn to drive differently with regen. They can learn differently without it as well.

    The brush issue is real. The liquid cooling is as well, although we don't have it with the HPEVS system.

    I think AC in general is over rated. By quite a bit actually. I guess the advantages would merit some level of premium, but not what it costs. The entire Warp9 soliton cost less than the Rinehart Motion Systems controller alone in the Mini for example.

    I just don't believe regen is what is perceived under any driving conditions, and the oft touted efficiency gains for AC is likewise suspect.

    As to efficiency, across the UNIVERSE of ALL products, 80% is easy, 85% is hard, and every percent above that is almost like giving birth and certainly not worth the cost for the additional efficiency.

    ReplyDelete
  8. Jack,

    In DC motors, too high of a voltage causes arcing. What does too high of a voltage do to an AC motor? And have you thought of over voltage testing a HPEVS AC50 on the bench like you mentioned you planned with the Warp's?

    ReplyDelete
  9. It looks like you may be able to purchase EV components from Nissan to make a very nice 80kW system like in the Leaf. Rumor has it that the motor is $2700, and the 3.3kW charger is $1700. No price disclosed yet for the inverter, but I imagine that it would be reasonably priced. All of this is second hand, I've not called the Nissan parts dealer to verify any of this.

    Follow along here for updates... http://www.mynissanleaf.com/viewtopic.php?f=8&t=3408

    ReplyDelete
  10. I mentioned it once before but I will say it again - I want to see how the Vantage Green Van goes with that spare 1238-7501! Matt will be plastered over the back window when he puts his foot in it.

    I am particularly interested in the AC50 installation and tests you are doing as I am working out a spec for a conversion that I would like to offer for sale. The Blade EV (hyundai getz with Azure dynamics system) is my inspiration but I don't agree with their regen on the throttle pedal, especially as you have to press a floor button to disengage regen while you change gear. I much prefer the brake transducer idea you are running with and I intend to plagarise it (as you did). Agreed, regen is possibly not all it is cracked up to be, but I think the market for "off the lot" EV's is attuned to the AC/regen option.

    As you have mentioned, the AC50/Curtis combo is an elegant little package and offers pretty much everything you could want, for peanuts in EV dollars. If Curtis do come out with a higher voltage offering - 120V/650A or even 144v/650A, that would be great. Maybe HPEVS should start looking into an "AC75" or "AC100" motor to go with it?

    By the way - one thing I noticed about the spyder is that the motor cables are crossed over from the controller. I this for a reason or can you reposition them parallel (I know, I AM anal) and "swap" phases in the controller software to keep it all turning in the right direction?

    ReplyDelete
  11. I purposely crossed my motor cables with the idea of canceling out noise compared to running them parallel. Don't know if it makes any difference, but what is the harm in doing so?

    Thought of two more benefits of AC, electronic reversing without contactors and a controller failure mode that would not put full power to the motor the way series DC controllers can.

    ReplyDelete
  12. Nick:

    I agree on all points. The AC-50 I assure you is capable of much higher power. We run a 30kw MES-DEA motor at 100kw and with the TIMS600 we were regularly doing 125kw. It did put out some heat. And it was water cooled so we could deal with it. The only drawback on the AC-50 is that it is air cooled. But it is running very cool at the current 68 kw and I think it could come very close to 100kw with no ill effects.

    The cables are crossed largely because that's how Duane did it. You are quite correct, we could straighten it out and deal with it in the controller. Perhaps we will..

    Jack Rickard

    ReplyDelete
  13. JP:

    Quite so. I had forgotten that as well. Electronic reverse is just hugely easy with AC motors and quite a mess with series DC. In fact, with the brush advance, it's not really a good thing with series DC at all.

    As to the failure mode, I'm not sure I follow you. But I guess shorting an IGBT would not make the motor turn as it would in a series DC. So that's a possible safety advantage I suppose.

    Jack Rickard

    ReplyDelete
  14. Too high a voltage does cause increased arcing on the commutator in a seriees DC motor. There is no such analogue in an AC motor. The motor itself can sustain unlimited voltage, and is normally current limited by the size of the wire in the windings.

    In order to "induce" current in the armature, an AC motor generally has many more windings than a series DC motor in the stator. The smaller the wire, the more windings will physically fit. And this is the limitation from my point of view on AC systems. High voltage and LOW currents. If we tried to run a car series DC on 300 amps it would have to be a pretty small car.

    The voltage limitation on the HPEVS system is purely a controller issue, not a motor issue. This motor I think would take quite a bit more current, and unlimited in higher voltage. But the Curtis controller is MOSFET based.

    MOSFETs are really very interesting devices in that their forward conduction resistance is remarkably low - milliohms in some cases. And so the voltage drop across the device when "on" is extremely low. This lets them run much cooler than IGBT devices and they are more efficient.

    Historically, the problem has been in developing low rds MOSFETS that can withstand higher voltages. They have been limited to peak values of about 200v. And those ARE peak values, so you don't want to run them at 200v. In fact, 130v would be a good limit for a 200v device.

    I'm sure the Curtis simply faults out and shuts down when some voltage is detected - to protect the MOSFETS. And I'm told that is 130v. We may test that a bit.

    Currently, there are some newer MOSFET devices coming onto the market based on silicon carbide formulations that will withstand much higher voltages - as high as 1200v in one case. They are quite expensive, and have some unique gate drive requirements. But I'm confident these components will improve with time. They are currently also quite expensive.

    We're also seeing some quite practical 400v MOSFET modules now. And that is pretty serious for EV's. Operateing at up to 300v would be very practical with these. I've even thought of tackling the OpenRevolt controller using some of those 400v MOSFET modules in the power section. But yes, an AC controller using such MOSFETS would be very cool - quite literally.

    Jack Rickard

    ReplyDelete
  15. Two AC v. DC issues, Jack:

    1. The locked rotor "hillholder" problem when applying throttle (amps) at zero rpm. This is likely to lift comm bars on the series DC motor in short order. Does it do so much damage so quickly to an AC motor as well?

    2. Regen with a push trailer. Its a backburner project, but I ultimately intend to build a push trailer using the front of a FWD car cut off at the firewall, as JB Straubel, Mr. Sharkey and a few others have done. These gas or diesel push trailers provide essentially unlimited trip range, so the car can be driven any distance to a show or elsewhere. With AC, I could regen on the highway with the push trailer powering it, and arrive with a fully charged pack after each long haul leg.

    Sure, its a niche advantage, but a real one nonetheless...

    TomA

    ReplyDelete
  16. The potential for more effective liquid cooling of AC motors is presumably significant at higher powers? To take a hypothetical example (using battery advances as yet unkown), if you wanted to push a Jaguar along at 90 mph for several hours and were burning 70kw to do so, you would need to disipate around 10kw from the motor-controller combination i.e. several times the output of most domestic space heaters.

    ReplyDelete
  17. Hi Jack
    I am surprised at the poor regen figures on the AC motors.
    I have a prius and its regen in the right conditions can be excellent.
    On some roads though it can also be marginal but for any one who lives in Sydney I can practically travel the whole length of Richmond Rd on regen
    I think the big thing about regen though is the braking.
    I have done 126521 Kilometres in my prius and I am still on my first set of brake pads!
    I think the disks will out last the car
    A big contrast to my Ford that had a set of disks every service!
    And a big financial saving as well
    I think regen braking is very nice braking as it is a constant deceleration and the grief coming from the passenger seat supervisor has disappeared since owning the prius

    ReplyDelete
  18. Hi Jack,

    The max 80% efficiency you measured with the AC50 is the same as shown in their own advertised graphs. So, that's good. You guys did good work, and hpevs isn’t bullshitting.

    But that 80% is the lowest figure I’ve found in published AC performance graphs. In contrast what you claim that 85% is a hell of a job, most claim around 90% eff, and some even go as high as 95%. If I remember correctly MES-DEA also claims more than 90% efficiency.

    Have you ever measured the efficiency of the mini with the original TIM600?

    Regard Jan

    ReplyDelete
  19. Jan:

    I already said what I said and like much of what I say, it is assumed to be bullshit by almost everyone listening or reading. This is because of the very rich field of knowledge provided by sales brochures and marketing people, along with a hurricane of press releases.

    80% efficiency in a freakin vacuum cleaner is hard. 85% is REAL hard. ANYTHING OVER THAT is generally delusional. It CAN be done. But the marginal cost of each percentage above 85% almost always precludes it being marketable as a product.

    If there are no moving parts, as in our batteries, we can see efficienies in the high 90s.

    The problem is all the inefficiencies are cummulative. The controller counts. THe transmission counts. The tires count. The cabling counts. The connections count. Some are worse than others, but they each add their little bit.

    So if you qualify your statement with sufficient vigor, you can make any claim you like. If you don't turn it on at all it's very nearly 100% efficient in many cases.

    Yes, we found HPEVS claims realistic and in some cases conservative.

    Frankly, they haven't made many regarding this Curtis 1238-7601. They just became available in very small quantities in March. But off the cuff, they said they thought we'd see 18%. Indeed.

    You've answered your own question. If we have one that claims 85% and makes 85%, what are you going to do to a motor to get it to 97%? Special magic sauce? Magnetic bearings?

    Again, get me to -400F and I can show you some shit. On this side of that, there is no 97%. Regen does NOT do what you think. And no, you can't connect a generator to the ass end of the motor and have it run itself.

    Copper foil helmet ENGAGE. Jack Rickard

    ReplyDelete
  20. http://www.hpevs.com/images/stories/jetta/computer2.png

    Hi jack had a look at the hpev site you linked to in your original post and I saw this picture.
    Is that 2 Curtis controllers wired in parallel ?
    Could you wire the 2 you have in parallel even though they are different models?

    ReplyDelete
  21. Re Efficiency (Jan's comment and Jacks reply)...

    ...is the efficiency issue clouded by two things:

    * Peak versus mean efficiency
    * System versus component efficiency?

    On "peak versus mean", this example http://www.metricmind.com/line_art/efficiency.gif shows efficiency varying from 70% at low power and rpm to 88% in a narrow range of rpm and torque higher up the power band. No one is lying if they quote "88% peak" but it is academic in a real-world vehicle.

    As Jack points out, the inefficiencies of individual components is cumulative. A 95% efficient controller driving a 95% efficient motor coupled to a 95% efficient transmission will have an overall batteries-to-driveshaft efficiency of about 85%

    ReplyDelete
  22. This is probably a dumb question. It appears to me that the torque graph shows 3rd and 4th gear producing the same torque at the wheel for some of the same speed at the beginning of the plot. If hp at the wheel is a function of speed and torque at the wheel then why doesn't the hp graph track the same? I must be missing something and once you tell me I will feel stupid.

    I am very interested in this study. I am trying to get a 3phase BLDC motor out of prius running in my volkswagen. The electric transaxle out of the prius has a fixed gear ratio. OEM parts out a junk yard is all my budget will permit right now. I hope my plan works. If so it may prove to be a nice 50kw drive train for under $1k.

    ReplyDelete
  23. Nicholas,
    Those two controllers are hooked up to a special motor, maybe two motors built into one, with dual sets of phase terminals.
    http://www.hpevs.com/images/stories/jetta/motor3.png

    ReplyDelete
  24. Quote
    And no, you can't connect a generator to the ass end of the motor and have it run itself.

    Its funny you should mention that but the Prius does actually have 2 electric motors and while both of them can drive and/or generate one prefers to drive and the other does prefer to generate
    Maybe this is why regen "seems" to work so well on the prius?
    I am surprised that more people have not tried using a prius transaxle as a drive unit
    Good to hear that some one is going to give it a go

    Thanks JP I had not noticed that

    ReplyDelete
  25. Hi Jack,

    OK, I understand you have no such data of the mini. No problem, just asking.

    Regards, Jan

    ReplyDelete
  26. Hi Jack,

    What is your opinion on the future of EV conversions is this AC or DC?

    Thank you.
    -Youri

    ReplyDelete
  27. Youri:

    The answer is yes of course.


    The future conversions will be DC, except for those that are AC. And AC will of course predominate, except in areas where DC is more prevalent.


    I don't mean to be specious, but the V8 is a well loved engine, so why do so may cars have V6 or four cyclinder?

    The future I see does not really involve a specific small set of small cars that are electric. It spans everything from 1957 Porsches to 2008 Escalades and from 1964 AC Cobras to 2009 Mini Cooper Clubmans. And each of those cars and trucks are somewhat specialized and in some ways unique. Likewise, different voltages, controllers, motors, and environmental systems will be used and THOSE will change all along the way.

    But for those who have long predicted the death of the series DC motor, the fact remains that for converting current to torque, there is nothing that can exceed it and most designs trade some of that off for other advantages. But they never really approach it.

    So I see both and in plenty. I think some of the most touted advantages of AC are mostly myth, but as JP has pointed out, there are some very real advantages as well. And for right now, AC remains for the most part at a price multiple almost three times higher than series DC systems. You have to have a LOT of advantages to merit a 300% price.

    Ergo our interest in HPEVS. They are producing a competitively priced AC system offering many of the advantages. But as we see in the two speedsters - 68 killowatts AC and 147 killowatts series DC, in the same car and at about the same price.

    And to me, the V8 and 4 cylinder analogy works. The DC Speedster peaks at a little over 3000 rpm, with lots of low end torque - kind of like putting a V8 in it. The AC version of the same car peaks up around 4000-4100 and is really quite good in the 3500-5000 rpm band - just like the original four cylinder Speedster.

    I like the lower powered at higher RPM's AC version. But in our shop, I'm outvoted 2:1 - Brain and Matt both like the power in the series DC version.

    So it's a matter of taste, choice and selection and in the future, it will be more of the same.

    Jack Rickard

    ReplyDelete
  28. Hey Jack!

    One more trivial benefit of higher voltage and lower current (AC) - more smaller cells affords more flexibility in battery placement; they can more easily fit the available shape of the storage spaces.

    One more additional cost of the same configuration - more BMS (sorry, I couldn't resist).

    Mike

    ReplyDelete
  29. Surely Mike. If you have a BMS, no single cell should ever need to be replaced in your car for the remaining time in its life.

    Until around 2am.
    =============================
    Earlier, I enquired about RFI.
    (RFI is Radio Frequency interference.)
    "Could it be RFI issues forcing the game to AC?".

    It became a little sidelined with regen and public perception which, (just look at me!) I have little interest in either.

    Just wanted to know whether anyone could be called to task for being unable to listen to their radio and blah, blah.

    Is there any notable difference Jack, guys?

    ReplyDelete
  30. This comment has been removed by the author.

    ReplyDelete
  31. Jack,
    Thank you for your opinion and I am CANNOT AGREE MORE with you on the DC and AC, where AC would be pushed by OEM and will be more in “Popularity Loop”. Meanwhile cost is the MAJOR factor in the most EV conversion decisions.
    Thank you again for all what you do.
    -Youri

    ReplyDelete
  32. Costs affect profits. Car makers need profits to re-invest. I reckon IF AC is truly the greater expense in manufacturing expense against what people have to shell out for DC series etc. Then there will be a mass market for DC.....

    Providing series DC manufactured vehicles meet with imposed RFI reg's in less Libertarian countries.

    ReplyDelete
  33. RFI in DC motors is addressed in R/C, where it gets into the receiver, with small capacitors from each brush to the motor frame and one from brush to brush. More complex variants include inductors. Commutators and brushes are elegantly simple, efficient, cheap, reliable and will be around for a long time as a result. -Klaus

    ReplyDelete
  34. Jack, Brian & Matt, hope you guys are doing ok after the storms. There's reports of downed power lines, strong winds at the airport and over a hundred customers without electricity in Cape Girardeau.

    JR

    ReplyDelete
  35. Jack won't suffer any downed electric. His cars are full and he has bought a DC to mains inverter a few weeks back. If he has used it then he'll be wearing an EV grin in the house!

    ReplyDelete