Changing an induction motor voltage

AC, DC, amps, volts and kilowatt. It's all discussed in here
User avatar
coulomb
Site Admin
Posts: 3327
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by coulomb » Tue, 14 Jul 2009, 18:50

acmotor wrote: Perhaps peak power needs to be something like 10 times average in a vehicle. So peak power needs to be 10 times the cruise at 100kmph power. That leaves 7.5kW motor out ? 10 to 15 kW seems more likely.

In an extreme sports high power vehicle perhaps; say 15 kW @ 100 km/h and 150 kW @ peak acceleration.

For a Barina class vehicle, it's more like 3:1; 16.5 kW @ 100 km/h and 50 kW to match the ICE acceleration. Conveniently, 3:1 is about the Tmax/Tnom ratio for induction motors. So if you size the motor for continuous power, it'll be about right for peak power. That's ignoring overvoltaging.

With overvoltaging, to take account of the higher iron losses, some derating is presumably needed of the continuous power. So that makes the ratio between maximum and continuous power higher, so the continuous power becomes more critical for a low performance vehicle.

For a high performance vehicle, you want a higher ratio anyway, so the peak power may be more important, depending on the ratio you want and the derating that is necessary.

With a 4x overvoltaging, if the new continuous power is 2.6 times the nominal, then a 7.5 kW cont / 25 kW peak becomes 19.5 kW continuous and 100 kW peak, enough for a quite sporty car. But while a non-overvoltaged 22 kW motor might be able to do a 1 minute burst of 100 kW with 4 minute rests (total guess), the overvoltaged 7.5 kW motor might need 5 or 8 minute rests between such burts, since the power ratio is a little higher for it (5.1:1 vs 4.5:1 peak verses continuous).

I think a lot of this will have to come from experience. Unfortunatley, I've decided I don't have the time and space at present to convert White Suzi; I'll have to get my electric kilometers hauling an Atkins cycle ICE around with me. So someone else will likely find out what the limits are.
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Changing an induction motor voltage

Post by Tritium_James » Tue, 14 Jul 2009, 18:53

acmotor wrote:Don't tell me TJ uses a disk brake ?
Yep, we've only got one $30000 solarcar motor, so no option to run two of them back-to-back!

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 18:56

acmotor wrote:I think you will find all posts, where temperature was the aim at least, contain the ambient temperature number, as did the last experiment !
My humble apologies. So it did. 16°C. I don't know how I missed it. Let's call it a senior moment. Image I should have just got the computer to search on "ambient".
Big blower was a skip job yesterday so sorry if it didn't meet your time scale !
Again, hy humble apologies. I assumed from your comment that "This took all evening", that measurements without the blower were done on the same evening as the one with the blower.
I like your use of kelvin ! Image Its not that cold in Perth !
I was trained as a physicist (among other things) and it's standard to give temperature differences in kelvin. It helps avoid mistaking them for actual Celsius temperatures. But you knew that.
Sorry, data as recorded. You are welcome to explain the results ! But please don't blame me if you can't follow what is happening. I have enough trouble myself. Image
I certainly don't blame you. I'm very grateful for what you've done. But I am completely stumped. It's totally inconclusive in regard to continuous power.

Here's the frequency, power loss and temp rise data from your last experiment (without big blower), normalised so that the lowest frequency results are all 1.0.

normalised     normalised     normalised     normalised
frequency     power loss     temp rise       temp rise
                                winding to case     winding to ambient
1.00             1.00             1.00             1.00
1.90             1.86             1.00             1.01
2.81             2.64             1.00             1.07
3.71             3.57             0.91             1.16

1. If you assume the Danfoss kW and calc shaft kW are correct and the torque is constant then the losses go up pretty much in proportion to the frequency, which is way worse than I thought, and means the continuous (or one hour) rating would barely increase above nominal. (Except due to increased cooling, which could have been applied for much the same effect (on continuous power) without any rewind.) But this totally disgrees with what we see in the temperature data.

2. If you assume the winding-to-ambient temperature rise data is correct and thermal resistance winding-to-ambient was constant, then the losses barely went up at all with frequency, which is way better than I thought, but that implies that if Danfoss kW are correct, the calc shaft kW are way out and the real kW must be higher than calculated at the higher frequencies.

What's frustrating is that I know if I was there, and we could work together on this setup, we'd soon figure out what was going on. You supply the coffee (grounds), I'll bring the chocky bickies. Image

Trouble is, for the price of the plane ticket (and nasty fossil fuel burning) I could build my own dyno.

I like your load-cell idea for torque measurement. Should be able to just bolt it between the "down" foot and the base. Could at least use it to keep torque constant even if it wasn't calibrated. But calibration should be possible too.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 19:02

TJ, point taken.
3PIMs for load are cheap though ?
iMiEV MY12     102,980km in pure Electric and loving it !

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 19:25

weber, glad we are talking again ! Image You're right, we would be dangerous together. I'd get mad with you always being right ! Image

coulomb, given my experience with red suzi, there is not much more can be done with standard 50Hz motor. The v/f wall is very final. Image
You can't overanything if there's nothing left to push with !

Perhaps if you factor in direct drive and the fact that without changeable gears you cannot achieve peak kW at lower speeds (as required with smaller motor), you may be more inclined to my 10 times peak kW target. I still feel tesla has it right and prius has it wrong.
Fit enough emotor kw in the first place and the tooth box (however hi tech) is not required.

weber, I don't think it is valid to normalise temperature to ambient ?
You know the temperature change equations. I'd have to google. The energy transfer difference is not linear with delta t. Can you check that out please ?
iMiEV MY12     102,980km in pure Electric and loving it !

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 19:35

coulomb, just to clarify.
A non overvoltaged 22kW motor can't even imagine 100kW.
up around 66kW on a good day would be the limit.
Live data from red suzi suggests 37kW VFD so around 32kW shaft is the peak on 11kW motor at standard voltage.

The other point with the voltage change idea is improving the power to weight after all. Image It works in a Tesla !
iMiEV MY12     102,980km in pure Electric and loving it !

djsharpe
Groupie
Posts: 90
Joined: Thu, 21 Feb 2008, 23:43
Real Name: David Sharpe
Location: Melb

Changing an induction motor voltage

Post by djsharpe » Tue, 14 Jul 2009, 19:36

Here is an idea for a cheap dyno. Find a 3 ph IM of rating close to what you are testing. Set the stator on a balanced cradle with a spring balance on a 1 metre arm. Connect rotor to subject motor. Gradually apply DC to stator either in delta or star. This will cause a short in the rotor bars. Measure torque & revs & you have kWs. This same principle can be used to brake (& break) an EV at low speeds when regen ceases. DS

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 19:40

acmotor wrote: The idea of seeking input on the forum is also for the scrutiny it offers. You should be honoured, not complaining ! Image

Sounds like "egos at ten paces", where an imaginary weber would reply: "You should be honoured I'm scrutinising, not complaining about the time it takes to get it right." Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 19:49

djsharpe wrote: Here is an idea for a cheap dyno. Find a 3 ph IM of rating close to what you are testing. Set the stator on a balanced cradle with a spring balance on a 1 metre arm. Connect rotor to subject motor. Gradually apply DC to stator either in delta or star. This will cause a short in the rotor bars. Measure torque & revs & you have kWs. This same principle can be used to brake (& break) an EV at low speeds when regen ceases. DS

Yes. Good idea. The balanced cradle thing is the hard part, because of course the shafts must stay aligned as the stator rotates against the spring balance on the torque arm. That's why ACMotor's idea of a load-cell is so good. Negligible movement.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 19:52

Quite right up to a point David. DC injection braking is fine and both these VFDs can do it with full control over current and time.

2 problems.

Pullout applies to braking torque in DC as well and once your slip speed passes motor normal slip RPM e.g. >170RPM on this little motor, the braking torque drops off. Think of DC injection as 0Hz drive to motor.

2. The problem of heat.
All the braking energy goes into heat in the rotor. There is only so much heat the rotor can handle. Definitely not kWs as with this motor.
Regen OUT of the motor is the best way to go.

(The stator also warms as normal due to the IR loss.)
iMiEV MY12     102,980km in pure Electric and loving it !

User avatar
Johny
Senior Member
Posts: 3681
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Changing an induction motor voltage

Post by Johny » Tue, 14 Jul 2009, 19:55

djsharpe wrote:...Connect rotor to subject motor. Gradually apply DC to stator either in delta or star. This will cause a short in the rotor bars. Measure torque & revs & you have kWs. This same principle can be used to brake (& break) an EV at low speeds when regen ceases. DS
Hmm. You have described DC braking which comes as standard on most VFDs. Trouble with acmotor's existing setup is keeping the shaft coupling happy. Good point though.

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 20:06

weber, that raises the question of rotor heat in the OV OC equations. Is that a factor confusing our understanding ? (speaks nicely after coulomb humour !) Image

edit: keyboard error
Last edited by acmotor on Tue, 14 Jul 2009, 17:47, edited 1 time in total.
iMiEV MY12     102,980km in pure Electric and loving it !

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 20:31

acmotor wrote: weber, glad we are talking again ! Image You're right, we would be dangerous together. I'd get mad with you always being right ! Image
Now that's one of the nicest backhanded compliments I've had in a while. Image I'll make an effort not be be right so often in future, but it will be very difficult for me. Image

Seriously. Thanks for making an attempt to accomodate tests of my theories. Sorry if I've been ungrateful. Just a shame parts of the data seem to be totally contradicting others.
weber, I don't think it is valid to normalise temperature to ambient ?
You know the temperature change equations. I'd have to google. The energy transfer difference is not linear with delta t. Can you check that out please ?

Well about 10 years ago I wrote a textbook chapter on heat flow for a course in Renewable Energy Technology (used in both energy efficient building design and solar thermal energy collection (e.g solar hot water systems). Plus I'm familiar with heatsink theory from my earlier electronics background with Telecom Australia (as it was then) and Ross Pink's Electronic Innovations P.L.

This looks like a good place to start:
http://en.wikipedia.org/wiki/Heat_transfer

The summary: Heat flow in solids, particularly metals, is very linear with temp difference. It's essentially an Ohms law thing, for the very reason that heat transport is occuring via the same mechanism as charge transport, namely average velocity of free electrons.

Radiation is very nonlinear. Natural convection is nonlinear, but forced convection is quite linear.

That's another reason (I hadn't thought of before) to have the forced-convection as high as possible: To ensure that radiation and natural convection are playing a minor part. But I think it is clear that radiation and natural convection are playing a minor part anyway, because we know if we turn off all fans the motor will overheat at a very low power level (although that could be tested).

So I think delta-T should be quite proportional to power loss. But I'm open to arguments to the contrary, given what we're seeing here.

Maybe we've discovered a new heat-loss mechanism. I know what it is. It must be pumping up the zero-point energy of the vacuum so all those damn over-unity devices like the "Lutec Electricity Amplifier" can keep working. Image
See http://www.lutec.com.au/lea.htm
Last edited by weber on Wed, 15 Jul 2009, 12:58, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 20:58

acmotor wrote: weber, that raises the question of rotor heat in the OV OC equations. Is that a factor confusing our understanding ? (speaks nicely after coulomb humour !) Image

Now that's a damn good point! The rotor has those little centrifugal impeller vanes on both ends which blow air through the end-windings where your winding temp probe is (as you reminded us recently) on the outside of the windings, not buried amongst them.

That could totally explain the near-constant delta-T between winding (sensor) and frame, maybe even the drop in that particular delta-T at the highest speed, despite an increase in losses.

I had failed to appreciate that this internal forced-convection heat path, which is in parallel with the purely conductive path in the middle, is actually quite significant in transferring heat from winding to frame, not just from rotor to frame.

So it looks like you're right this time ACmotor. Image It could well be that although the losses increase even faster than I thought they would, it doesn't matter as much as I thought it would because the internal thermal resistance of the motor goes down with speed, due to the internal rotor vanes.

We should still be cautious about this conclusion and it would be really good if we could get some independent confirmation of the drives' calculated torques (and hence shaft powers) by one of the mechanical arrangements that has been suggested.

An electronic kitchen scale (that uses a load-cell), mounted under one foot, should be about right. Then you'd have to pack up the other foot and those of the load motor to match.
[Edit: Formatting]
Last edited by weber on Tue, 14 Jul 2009, 18:17, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 21:05

ROFL ! Image

Now..
Good that you are heat wise !
I'd go for delta t being non linear being a big factor in these tests.
The large surface area of motor is quite radiant and even my forced airflow was rather one sided.
As I say, the measurements may not be fancy but they stand as live data.

Just as long as I don't cook the motor.

A note on motor design... plastic fans are poor thermal conductors and you lose one path for heat extraction from the shaft c.w. say aluminium fan. rotor and bearing temperature could be lower with metal fan ?
iMiEV MY12     102,980km in pure Electric and loving it !

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Tue, 14 Jul 2009, 21:14

Sorry, leap frog post.
Yes to your internal heat flow thinking and thanks Dave for the lead to rotor heat.
The mechanically confirmed torque is still desirable.
Now prof. weber, what is you prediction should we set up to read shaft torque ? Image (I need this in case your wrong !) Image
iMiEV MY12     102,980km in pure Electric and loving it !

User avatar
coulomb
Site Admin
Posts: 3327
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by coulomb » Tue, 14 Jul 2009, 21:26

acmotor wrote: coulomb, just to clarify.
A non overvoltaged 22kW motor can't even imagine 100kW.
up around 66kW on a good day would be the limit.

Yes, of course. Sorry, was thinking of the MX-5, which has a 1.7x overvoltage.

So my point is, even though with overvoltaging you can get a pretty spectacular ratio of peak to continuous power, how short does the duty cycle have to be?
Learning how to patch and repair PIP-4048 inverter-chargers and Elcon chargers.

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 21:53

acmotor wrote: Image weber, you need to lighten up !!! Image

And here at last, hot from the lab, is the experimental evidence:

Image

(Photo of lightened up weber) Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
woody
Senior Member
Posts: 1703
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

Changing an induction motor voltage

Post by woody » Tue, 14 Jul 2009, 21:56

Gah! You don't have a beard or braces! What kind of engineer are you?
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 22:17

woody wrote: Gah! You don't have a beard or braces! What kind of engineer are you?

Just be glad I spared ya the glare off me bald strip. Arrgh. Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
Johny
Senior Member
Posts: 3681
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Changing an induction motor voltage

Post by Johny » Tue, 14 Jul 2009, 23:23

I don't understand - where is the rotor? Why does the power inlet have those long line of white cable clamps?
I knew acmotor's experiments into over unity would result in some kind of space-time collapse/monster thingy effect!!!

User avatar
Hemonster
Groupie
Posts: 90
Joined: Mon, 01 Dec 2008, 00:28
Location: Christchurch NZ

Changing an induction motor voltage

Post by Hemonster » Tue, 14 Jul 2009, 23:39

I was following this thread for technical interest, now its just plain entertaining ... Image

Thanks guys ... keep it up ...

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Tue, 14 Jul 2009, 23:47

Johny wrote: I don't understand - where is the rotor? Why does the power inlet have those long line of white cable clamps?
I knew acmotor's experiments into over unity would result in some kind of space-time collapse/monster thingy effect!!!

You don't know the half of it. All is about to be revealed. Hold yourselves in readiness.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2429
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Changing an induction motor voltage

Post by weber » Wed, 15 Jul 2009, 00:13

Johny wrote: Yeah acmotor, and while we are at it, I am supremely disappointed that coupling the DC buses together didn't result over-unity and then in some plans being sold on the internet for powering ones house from 2 VFDs and 2 ACIMs. I'm sure that if I reworked some of your photos and changed the captions I could make a heap of cash Image

Funny you should mention that, Johny. You know, you don't even need the other motor and VF drive! Coulomb and I have made an amazing breakthrough! Well OK, we have to give some credit to a couple of pesky 9 year old kids that fiddled with our setup when we weren't looking.

You know how every industrial VF drive has a 3-phase input rectifier that doesn't normally get used in an EV, right? Well we had our VFD running a motor on the bench, from some batteries, and I said to Coulomb, "The VFD is producing 3 phase AC from its output, to drive the motor, right? What if we make use of that unused input rectifier by feeding the VFD's output back to its input, and generate our own power, so we never need to charge the batteries again!".

And Coulomb replied. "Weber, you're a bloody genius. What don't you know?".

I replied modestly, "Oh several things Coulomb".

So we hooked it up and waited. But for some mysterious reason the batteries just kept running down, same as before. So we took a break and had a cuppa and did a bit of Google research on "over-unity devices", not realising that while we did so, two 9 year olds, who we'd earlier told to p*** off, were tampering with our setup.

During our Googling we learned that there are these rules called the 3 Laws of Thermal Underpants or something, which go a bit like this:
1. You can't win.
2. You can't break even.
3. You can't get out of the game.

We learned that over-unity devices don't claim to break these laws, but they claim to be tapping into a new source of energy, sometimes called "the zero-point energy of the vacuum" and sometimes "quantum gravity". "So why doesn't ours do that?" we wondered.

So we thought we'd give it another try, and back to the shed we went. Powered it up. Motor running. While watching the battery voltage we noticed that it was only dropping about a third as fast as before! Somehow we were amplifying the electricity coming out of the battery by a factor of 3 or more!

What was different this time, I hear you ask? Well here's the photo so you can see for yourselves what we found when we took the cover off.

Image

The kids had stuffed a toilet brush inside the VF drive and its wire loop was apparently acting as some kind of "antenna" for the zero-point energy! We had invented

The Loo Tech Electricity Amplifier

This is not to be confused with that tawdry ripoff of an Adams motor from those guys in Cairns. Even if it did win the Bent Spoon Award in 2001


And it should not be confused with this Loo Tech Motion Control Device either.

Image

[Edit: Fixed broken link to Bent Spoon Awards]
Last edited by weber on Thu, 11 Feb 2010, 04:14, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
acmotor
Senior Member
Posts: 3591
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Changing an induction motor voltage

Post by acmotor » Wed, 15 Jul 2009, 00:51

Hmmm, I'll copy that post and paste it whenever you get too serious ! Image
iMiEV MY12     102,980km in pure Electric and loving it !

Post Reply