RX400H transaxle

AC, DC, amps, volts and kilowatt. It's all discussed in here
Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

RX400H transaxle

Post by Tritium_James » Wed, 04 Jul 2012, 18:26

Which is exactly why we're buying off-the-shelf SEW motors with the lowest available winding voltage option :)

User avatar
woody
Senior Member
Posts: 1711
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

RX400H transaxle

Post by woody » Wed, 04 Jul 2012, 20:02

Rewind is a good deal!
Buying wound for low voltage is a good deal. (this is what I managed)
Finding a controller with 4 times the current capacity at the same voltage is hard / expensive work.

Tesla & AC-Propulsion T-Zero have both gone this way.

But for converters: gearing this high speed motor down in a car designed for 4-6000 rpm is an added expense/weight - differentials don't go much outside of 2:1-5:1, so either keeping the gearbox (possibly not designed for high speed/torque input) or something more specialised (e.g. Lenco).

Crazy Ideas:
When "replacing the existing wire with wire of four times the cross section", instead replace it with 4 wires of the same cross section and bring the wires out separately. Then you could have 4 controllers or 4 sets of gates (mosfet/IGBT) driving into the same motor which could/should be cheaper/smaller than 1 controller with 4 times bigger gates.
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
Richo
Senior Member
Posts: 3567
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

RX400H transaxle

Post by Richo » Wed, 04 Jul 2012, 20:51

T2 wrote: So I want all those bitching and complaining about spending $800 on a rewind ... It's one of the best deals going.


I agree.

The other point is to compare a cheap Industrial AC motor with a rewind to that of an off the shelf DC motor - The AC motor is still better value ($/kW) without even taking into account the added benefits of efficiency, regen and lack of brushes.
Help prevent road rage - get outta my way! Blasphemy is a swear word. Magnetic North is a south Pole.

User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Wed, 04 Jul 2012, 22:08

T2 wrote: TTAM The Truth About Motors.

... To recap, since the motor underwent its rewind the current has been four times greater as previously noted but now with the original 400Vac across its terminals it is now absorbing four times the electrical power and producing quadruple the power.

But what about iron loss ? In the event of full disclosure that does modify the continuous rating by exponential 0.7 but this is empirically determined for only one motor. In this case 4^0.7 =2.6

You're preaching to the converted (ahem) here, T2. Well laid out argument.

What bothers me about this though is when you rewind for 1/nth the turns with n times the wire area, you end up with a motor the same size that produces the n times the power and torque (at the same speed).

[Edit: this is wrong. You also end up with 1/nth of the field strength (because of the n times fewer turns, hence n times fewer ampere-turns), so you have to increase the current n times (compared to original) to bring the field strength back to original, and hence the torque back to original. So the motor size is still proportional to the torque (they are unchanged). ]

So why is the received wisdom that motor size is proportional to torque? Maybe this is only true if the motor voltage is constant. So this might apply in the DOL world, where the voltage is set by the local mains voltage. So maybe this is just saying that for a given voltage and frequency, changing to double the number of poles (roughly doubling the torque) roughly doubles the size of the motor.

With all the variables we have with VFDs for vehicles, I think that the "motor size is proportional to torque" rule of thumb simply doesn't apply. After all, T2's argument of rewinding for lower turns but same area is a counter example.

Maybe in practice you can't fit 1/n turns of wire with n times the cross sectional area into the same slot, but it seems to me you could get pretty close (maybe you have to increase the size of the motor by 10%, but not 400%).
Last edited by coulomb on Fri, 06 Jul 2012, 16:56, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

RX400H transaxle

Post by Johny » Wed, 04 Jul 2012, 22:17

coulomb wrote:With all the variables we have with VFDs for vehicles, I think that the "motor size is proportional to torque" rule of thumb simply doesn't apply. After all, T2's argument of rewinding for lower turns but same area is a counter example.
Rewinding for a lower voltage does not increase the nominal torque though - just the speed at which nominal torque is still available.
So motor size is roughly proportional to torque for the same number of poles.

T2
Groupie
Posts: 121
Joined: Sat, 08 May 2010, 04:51
Location: ON,CANADA

RX400H transaxle

Post by T2 » Thu, 05 Jul 2012, 15:26

- johny thanks for your input you fully get this.

- Richo I just put the latest piece of my rubbish on your thread in the members machines section. Somehow I think Richard Gervais has got into my writing style.

- Coulomb I apologise if I have made some of my posts so overly long that you needed to skim through and may have overlooked a relevant point. Incidentally I have a couple of your posts still to answer from last week so I know you are quite involved here and busy with your own projects but I do appreciate that you noticed the care I am taking with my layout. My posts are usually read over at least a hundred times before I press that PostReply button. I like to be certain that I have weeded out any ambiguity in the event that the wrong use of a word or tense left uncorrected may later cause my argument to become derailed. Obviously I still have room for improvement.

As Johny noted, rewinding does not change the nominal torque despite the manyfold increase in current needed to preserve it after having a large number of those turns replaced by fewer turns of thicker wire.
But it's all on page 4 here if anyone cares to look.

Woody notes, as do others, to the effect that we all know what we should be doimg it's just that the conventional manual transmission doesn't give us that elusive 10 : 1 ratio able to accept a 10,000 rpm input. And I think that says it all there.

I am returning to How to convert a Hybrid since the Prius may indeed have that compatible mechanical interface as Coulomb described there sometime last November I believe.   
T2

User avatar
weber
Site Admin
Posts: 2559
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

RX400H transaxle

Post by weber » Thu, 05 Jul 2012, 17:50

I agree with T2, and not TJ. After a rewind the I^2R copper loss at the same torque is essentially unchanged.

Consider the equivalent case of rewiring for one quarter (n=4) of the voltage. Assume a 4 pole motor where each phase has its four pole-windings (pole phase groups or PPGs) in series. We separate them and reconnect them in parallel. If each PPG was 1 ohm, then we've gone from series 4 ohms to parallel 0.25 ohm, a factor of 16 (n^2) reduction. To get the same torque as before, we'll need 4 times the current, so we get the same current as before in each winding. I is multiplied by n, R is divided by n^2, so I^2R stays the same.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
weber
Site Admin
Posts: 2559
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

RX400H transaxle

Post by weber » Fri, 06 Jul 2012, 03:00

Oh! And Coulomb is confused. T2 and Johny are right that rewinding an induction motor, for lower voltage, doesn't change the maximum torque (unless it also changes the number of poles).

I've been off the net for the last few days because I've been out in the bush installing off-grid solar power systems (where lead-acid still rules). Then this forum seems to have been down all today.

Jeez. When folks like TJ and Coulomb can still get this stuff wrong, what hope do newbies have? Image

One thing I'm still confused about: What limits the maximum instantaneous torque from the motor, other than mechanical breakage. In most designs I've seen, it is limited by the inverter's current capability. But what if we can make the inverter as big as we like? Does the torque just keep going up and up in proportion, albeit limited to lower and lower rpm? Surely saturation initially causes diminishing returns and eventually a limit? Is there any relationship between this and the direct-on-line breakdown torque specification for the motor?

Here's another one. We were talking about hill starts. With an inverter-driven induction motor having rpm feedback ("closed loop"), can the motor be made to deliver the full torque possible within the inverters current limit, even at zero rpm?
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
Jeff Owen
Groupie
Posts: 155
Joined: Thu, 13 Nov 2008, 15:53
Real Name: Jeff Owen
Location: Brisbane

RX400H transaxle

Post by Jeff Owen » Fri, 06 Jul 2012, 03:31

Richo wrote:
The other point is to compare a cheap Industrial AC motor with a rewind to that of an off the shelf DC motor - The AC motor is still better value ($/kW) without even taking into account the added benefits of efficiency, regen and lack of brushes.


To compare AC with DC, you should also take into account the cost of controller, batteries and management system. As it appears to be preferable to use a much higher voltage with AC systems, it is unlikely to be practical to use lead acid batteries due to weight limits on the vehicle being converted. This means the most likely solution is a large number of expensive Lithium cells, and a corresponding number of management units and inter-cell connections. How does your $/kW stack up when all relevant costs are included.

Also, could you explain how a box of very expensive electronics is an added benefit compared to some simple brushes.   


User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Fri, 06 Jul 2012, 03:52

weber wrote: Oh! And Coulomb is confused.

Yes, I am. I was somehow thinking "torque is proportional to current, and current is multiplied by 4, so torque is multiplied by 4". But I forgot that when you reduce the volts per hertz by 4, you also reduce the torque per amp by 4. This would presumably be because with one quarter the turns, you need four times the current just to keep the same flux. Now I'm thinking that torque is proportional to flux multiplied by drive current. So when we say "torque is proportional to current", it assumes a constant flux, which is normally the case. But I just don't remember, so I'm back to confused. Hopefully I'll sort it all out and post something coherent later.

I just did a quick search, and the equations describing an induction machine take up many pages. So I don't feel quite so bad about being confused.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Fri, 06 Jul 2012, 04:23

weber wrote: One thing I'm still confused about: What limits the maximum instantaneous torque from the motor, other than mechanical breakage. In most designs I've seen, it is limited by the inverter's current capability. But what if we can make the inverter as big as we like? Does the torque just keep going up and up in proportion, albeit limited to lower and lower rpm? Surely saturation initially causes diminishing returns and eventually a limit? Is there any relationship between this and the direct-on-line breakdown torque specification for the motor?
Yes, that one's got me too. Maybe only the reactive current saturates the core, since only it produces flux, and the real current just produces torque. That would mean that you could keep increasing the torque by increasing the real (sometimes called quadrature) current, with the limits being due to heating and mechanical limits and not saturation.
Here's another one. We were talking about hill starts. With an inverter-driven induction motor having rpm feedback ("closed loop"), can the motor be made to deliver the full torque possible within the inverters current limit, even at zero rpm?

I can't see why not. The rotor still sees the slip frequency inducing current in it, even if the rotor happens not to be moving. That extends even in the case where the electrical frequency is zero, and the rotor speed is the negative of the slip speed. For example, you do a poor hill start and the vehicle rolls backwards; at the instant where the rotor speed equals the slip speed, the electrical speed is zero, and the inverter is effectively outputting DC (still with PWM, but the PWM ratios at that instant aren't changing). In fact, if the vehicle happens to be on an incline of just the right slope, I think you could have the vehicle rolling backwards at constant speed with the inverter outputting DC, and the rotor speed equals the slip speed needed to produce enough torque for the vehicle to balance gravity (so it is rolling backwards at a constant speed, not accelerating).

Oops! But the field would not be getting magnetisation. Well, it would be getting DC magnetisation, which might be just what it needs at that point.

A traditional industrial motor with impeller cooling would quickly overheat in a situation like that, so it is said that induction motors don't like operating at low speeds. But if the cooling is adequate, and/or the motor doesn't stay in this situation for long, I think it might work. It all depends on how the magnetisation works. So if I'm right, you would be able to replace the inverter with say 3 6 V batteries, and get the same effect. Or maybe one shorted output and two 8 V batteries; it would just change the electrical phase, which would have no long term effect on the vehicle behaviour.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Fri, 06 Jul 2012, 05:02

weber wrote: With an inverter-driven induction motor having rpm feedback ("closed loop"), can the motor be made to deliver the full torque possible within the inverters current limit, even at zero rpm?

According to this Wikipedia page, yes:
Inverters can be implemented as either open-loop sensorless or closed-loop FOC, the key limitation of open-loop operation being mimimum speed possible at 100% torque, namely, about 0.8 Hz compared to standstill for closed-loop operation.[9]
(This quote is about 3/4 of the way down the Technical Overview section.)

I assume that it also works at below standstill, i.e. with negative rotor speeds and near-zero electrical speed.
Last edited by coulomb on Thu, 05 Jul 2012, 19:03, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

PlanB
Senior Member
Posts: 465
Joined: Sat, 16 Jan 2010, 15:24
Real Name: Kris McLean
Location: Freemans reach
Contact:

RX400H transaxle

Post by PlanB » Fri, 06 Jul 2012, 15:14

Reading all this I'm thinkin' if I ever get a direct drive EV built I better not hold it on carpark ramps with the accelerator like an ICE auto? Better to use the brakes maybe?

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

RX400H transaxle

Post by Johny » Fri, 06 Jul 2012, 15:31

weber wrote:.... What limits the maximum instantaneous torque from the motor, other than mechanical breakage.
In complete motor failures that I have seen* with 50kW motors, it's the stator windings that have suffered. While you can cool the motor as a whole, the heat transfer from copper windings to motor casing is rather poor. The windings overheat, insulation breaks down and windings short. The saturation of the stator would not help and would contribute to heat buildup. It may only take a few seconds at 10 times nominal current to completely "fry" an AC motor.

Look at your winding terminations and picture 500 Amps flowing through them.

* Seen pictures and explanation of failure modes - commonly excessive load during DOL startup.

User avatar
woody
Senior Member
Posts: 1711
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

RX400H transaxle

Post by woody » Fri, 06 Jul 2012, 15:46

Jeff Owen wrote:How does your $/kW stack up when all relevant costs are included.

Also, could you explain how a box of very expensive electronics is an added benefit compared to some simple brushes.   
Good points Jeff.

You can run small lead acid batteries: e.g. (Ex-)a4x4kiwi's Hilux ran 50 Greensaver 20Ah.
Or small lithium, e.g. ~200 Headway's I've heard are US$6 ea now

The advantages of AC over DC are useful not indispensible - regenerative braking + efficiency.
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
woody
Senior Member
Posts: 1711
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

RX400H transaxle

Post by woody » Fri, 06 Jul 2012, 16:09

coulomb wrote:
weber wrote: One thing I'm still confused about: What limits the maximum instantaneous torque from the motor, other than mechanical breakage. In most designs I've seen, it is limited by the inverter's current capability. But what if we can make the inverter as big as we like? Does the torque just keep going up and up in proportion, albeit limited to lower and lower rpm? Surely saturation initially causes diminishing returns and eventually a limit? Is there any relationship between this and the direct-on-line breakdown torque specification for the motor?
Yes, that one's got me too. Maybe only the reactive current saturates the core, since only it produces flux, and the real current just produces torque. That would mean that you could keep increasing the torque by increasing the real (sometimes called quadrature) current, with the limits being due to heating and mechanical limits and not saturation.
As you increase the current, you increase the magnetic field.
At some point (~90% of the nominal current) you reach the magnetic saturation - i.e. you are magnetising the stator as much as you can without making it a permanent magnet.
Beyond this point some of the energy is being used to permanently magnetise the stator (in alternating directions).
This is called hysteresis.
This energy is turned to heat (hysteresis losses).

The more current you put in, more is being wasted on this hysteresis losses, as well as I^2R losses, so your efficiency is dropping, the heat is increasing - the time you can "nuke it" is smaller and smaller before you melt the insulation and short the windings.
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Sat, 07 Jul 2012, 03:10

woody wrote: As you increase the current, you increase the magnetic field.
My understanding is that as you increase the reactive (direct, out-of-phase with voltage) current, you increase the field. Certainly if you keep the same power factor and increase the current, you will be increasing the field. But you could increase the active (quadrature, in-phase with voltage) current by setting the voltage phase appropriately (you have to know some motor constants for this, of course). I thought that would be how you would do it, and I thought that would avoid saturation.
At some point (~90% of the nominal current) you reach the magnetic saturation - i.e. you are magnetising the stator as much as you can without making it a permanent magnet. ... The more current you put in, more is being wasted on this hysteresis losses, as well as I^2R losses, so your efficiency is dropping, the heat is increasing...

So you're saying that the motor is starting to saturate (softly) at 90% of rated load (or at 90% of rated current, which is quite close to the same thing), and as you increase the load (for short periods of time) up to breakdown torque or some other limit, often around 300% of rated load, you are driving the iron over 300% past saturation? My understanding is that the windings become more or less a short circuit only a few tens of percent past the onset of saturation (at least at the peaks of the current waveform). I'd also think that you'd arrange for rated load to be about 90% of saturation, not 111% (100%/0.9) as you suggest, so under rated load, it will not be saturating much at all. (Of course, saturation is a gradual thing, the iron gets more and more lossy, but there is usually a point where saturation is deemed to have started. So 90% of "saturation" will still have higher iron losses than at say 80%.)
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

Canberra32
Groupie
Posts: 238
Joined: Wed, 13 Jun 2012, 04:25
Real Name: Damian butcher
Location: Canberra

RX400H transaxle

Post by Canberra32 » Sat, 07 Jul 2012, 16:04

Ok I have been following this thread and have reached the point of information saturation lol...
While I may be a whizz with fabrication issues the deep electrical technical dose do some baffling on my part.
When I see 90%, 300%, saturation and other such words my brain simply says get a bigger one to close to limits of the product...
Can anyone clear up why this is not the case for me please? And please try explain like the do on star trek where they take a whole heap of complicated stuff and dumb it down for the average.
I find this stuff interesting but once I lose track I chase my tail and really don't learn anything.

User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Sat, 07 Jul 2012, 16:39

Canberra32 wrote: \When I see 90%, 300%, saturation and other such words my brain simply says get a bigger one to close to limits of the product...

Unfortunately, just getting a bigger motor that stays away from magnetic saturation all the time isn't practical. Already, a 132 frame motor as we have in our MX-5 is the biggest we could fit in there; we're already running the transmission higher in the transmission tunnel than is intended. Other vehicles might be able to use a slightly larger motor, but not one 3x the size. So we have to push the motor for short times beyond its continuous ratings, where we know that saturation isn't a problem.

But I'm not yet convinced that the motor will be heavily into saturation at higher than nominal current.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

Canberra32
Groupie
Posts: 238
Joined: Wed, 13 Jun 2012, 04:25
Real Name: Damian butcher
Location: Canberra

RX400H transaxle

Post by Canberra32 » Sat, 07 Jul 2012, 22:36

Ahhh its good to see the EV world still has the need more power curse of the petrol heads lol.
Well if your running 3x the power in a mx5 you will have quite the weapon.
I turboed our 98 mx to just over 300hp and the thing was crazy.
Only thing I had to do was swap the diff out for a skyline one as I kept braking the stock mx diff.
I got lucky with the celica I'm converting as for some reason they make a swap out FWD gearbox with driveshafts that runs 850hp with 600pounds of torque capable... As luck would have it the motor puts out 500hp and 700pounds of torque good luck pulling that power out of a 1.8L for any reasonable budget lol.

User avatar
weber
Site Admin
Posts: 2559
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

RX400H transaxle

Post by weber » Sun, 08 Jul 2012, 04:12

It's important to note that you don't have to know any of this stuff in order to do a conversion -- even an AC conversion. It has all been taken care of by your VF-drive designer -- particularly if you buy a VF drive and motor package, where the drive's motor parameters will have already been set correctly by the supplier.

But yes, in the "electron head's" quest for ever more torque or power in a smaller, lighter or cheaper package, some of us like to try to understand things more deeply. Sometimes it's the blind leading the blind. Other times it's the blind men feeling different parts of the elephant.

Hi PlanB, you wanted to come up with a good conversion plan, but your thread seems to have morphed somewhat, although I know you're not complaining.

I found a paper that, along with the helpful contributions of all you kind folk, has helped answer my recent questions. Although it is heavy going.
"Torque Capability and Control of a Saturated Induction Motor Over a Wide Range of Flux Weakening" by Grotstollen and Wiesing, 1995.

Basically, if VF drive current is unlimited, low speed (or stalled) torque is not limited by saturation, but only by thermal considerations as Johny suggested. To make use of very high torques for very short times, a high current VFD probably needs to run a thermal model of the motor windings, which could be as simple as a leaky integration of current squared.

When thinking about induction motors on the grid we tend to think in terms of voltage and frequency. But VFDs typically control current. They control its magnitude and its phase relative to the direction of the flux in the rotor. The VF drive has to do some fancy maths to figure out where the induction motor's rotor-flux is pointing at any given time, based on current, voltage and rpm measurments, but once it has done so, controlling the motor becomes as easy as controlling a DC motor.

Rather than specify the desired current-vector in polar terms (magnitude and phase), the VFD uses rectangular components of current called "direct-axis" (in phase with rotor flux) and "quadrature-axis" (at 90 degrees to rotor flux), Id and Iq. You'll see these mentioned in the Wavesculptor manual. These are analogous to the field-winding current and armature current of a DC motor.

Torque is proportional to Iq times flux

And flux is produced (nonlinearly) in response to Id as Coulomb said. It is pretty much only Id that can cause saturation.

So Torque is proportional to Iq x nonlinear_function_of(Id).

In terms of voltage and frequency, Id is approximately proportional to V/f. For a given motor, there is a constant optimum value of Id that gives maximum torque at all speeds. There is absolutely no point in increasing Id beyond that optimum value. But at high speeds Id (and hence flux) can't be maintained at that optimum value because you run out of voltage, and it's fine for Id to be reduced in that case. Id may also be reduced at other times (when maximum torque is not required) to increase efficiency.

While the actual breakdown phenomenon does not occur with VFDs, the maximum torque available with a VFD, at the motor's nominal speed, is the same as the breakdown torque (also called pull out torque or Tmax), given the same Id as is produced by the nominal V/f, which is usually very close to the optimum Id.

[Edit: Added DC motor analogy to paragraph 6 and split it into two. Corrected last paragraph by adding "at the motor's nominal speed".]
Last edited by weber on Sun, 08 Jul 2012, 11:15, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

User avatar
coulomb
Site Admin
Posts: 3614
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

RX400H transaxle

Post by coulomb » Sun, 08 Jul 2012, 05:08

Canberra32 wrote: Well if your running 3x the power in a mx5 you will have quite the weapon.

Err, we don't want triple the original ICE power (86 kW), we want triple (actually more) the continuous power of the largest motor we can fit in there (22 kW). In fact, with 750 V, we hope to get well over 100 kW from the motor, partly by overclocking it by 2x (100+ Hz or 6000 RPM), and partly by making it draw more than the continuous rated current for short periods (this is basically the tripling part).
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, PIP-4048MS inverter, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.

User avatar
weber
Site Admin
Posts: 2559
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

RX400H transaxle

Post by weber » Sun, 08 Jul 2012, 23:29

I just corrected (I hope) my post above, by adding "at the motor's nominal speed" to the last paragraph. Because as it stood, the last paragraph was contradicting the 5th paragraph.

At speeds below nominal, if the VFD can deliver the required current, it seems the torque can go up in proportion as the speed goes down, producing the same power, but at lower and lower efficiency, until something mechanical breaks or the insulation burns off the windings. So those flat tops we see on instantaneous-torque vs speed curves are not really a property of the motor at all, but of the VFD's current limit.

[NITPICK WARNING = Feel free to skip the following]
However paragraph 5 is not strictly correct either. You may have noticed I used weasel words like "pretty much" and "approximately" in other parts of the post. The "basically" I used in para 5 probably wasn't weasely enough, because equation 6 of the paper shows that, as Woody suggested, there is a contribution towards magnetisation (and hence saturation) that comes from the quadrature-axis current Iq that we normally think of as only torque-producing, as well as from the direct-axis current Id that we normally think of as magnetising. However the contribution from Iq is greatly reduced because it is multiplied by what might be called the flux-leakage-factor, a dimensionless number close to zero. An absence of flux leakage would mean no contribution from Iq. But real motors have a few percent of rotor flux that does not pass through the stator. And there's a positive feedback whereby the leakage factor increases with saturation, so it's far better if Id is a little low, rather than a little high.
[/NITPICK]

You can just keep on going to more and more refined mathematical models of these motors. Another paper will say "That previous model is all well and good, but it ignores X, here's how to include X." and eventually (or already) it gets too complicated to keep in your head.

But again I emphasise that you don't need to know any of this stuff to do a conversion.

For example, the nitpick above is irrelevant because Tritium can put your motor on their dyno and adjust Id to give maximum stall torque (or maximum low-rpm torque) when the Wavesculptor is putting in its maximum total current of 300 A. Then set the WaveSculptor to remember that optimal value of Id, typically around 20 A.

If you were then to put the same motor on a different VFD that had a maximum current of say 400 A you might find that the optimum value of Id is slightly lower because of the contribution to magnetisation from Iq times leakage-coefficient. It might be nice to have an explanation for that, but it doesn't really matter, since you found the new optimum by testing anyway.

BTW, the total current into the motor is sqrt(Iq^2 + Id^2). But Iq and Id are not the same as active and reactive current. The latter are with reference to the voltage waveform, the former are with reference to the rotor flux orientation.

[Edit: Spelling]
Last edited by weber on Sun, 08 Jul 2012, 18:34, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

PlanB
Senior Member
Posts: 465
Joined: Sat, 16 Jan 2010, 15:24
Real Name: Kris McLean
Location: Freemans reach
Contact:

RX400H transaxle

Post by PlanB » Mon, 09 Jul 2012, 01:07

weber wrote:
Hi PlanB, you wanted to come up with a good conversion plan, but your thread seems to have morphed somewhat, although I know you're not complaining.
All grist for the mill Dave, the RX400H transaxle appealed because it was comparatively easy mechanically & I'm only down $1k on it. And there would seem to be options that can find their way up a ramp & out of a multi-story carpark.....
1) Go with the Sew + Wavesculptor & a 5.1:1 diff for 1200Nm $9k
2) Get another RX400H & 2 of TJs controllers for an 1800Nm AWD $11k
3) Try an EVO pancake motor + wavesculptor direct to the diff $12k
4) Ride around in my mates Leaf, 1700Nm & a priceless amount of shame
It's been a steep learning curve, especially that article I read last night on FOC Vs DTC, just ugly. If I can build something a little bit special by just connecting black boxes I surely will.




User avatar
weber
Site Admin
Posts: 2559
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

RX400H transaxle

Post by weber » Mon, 09 Jul 2012, 04:02

PlanB wrote:All grist for the mill Dave, the RX400H transaxle appealed because it was comparatively easy mechanically & I'm only down $1k on it. And there would seem to be options that can find their way up a ramp & out of a multi-story carpark.....
1) Go with the Sew + Wavesculptor & a 5.1:1 diff for 1200Nm $9k
2) Get another RX400H & 2 of TJs controllers for an 1800Nm AWD $11k
3) Try an EVO pancake motor + wavesculptor direct to the diff $12k
...

Here's another one:
1b)Go with the Sew + Wavesculptor & a standard 4.1:1 diff and switch from delta to star at low speeds, to get 1200 * 4.1/5.1 * sqrt(3) = 1670 Nm (at the wheels).

The tricky part might be getting the WaveSculptor to change motor configurations on the fly. The last time Coulomb and I discussed this with Tritium_James, over a year ago, he said it *should* work, but no guarantees as it had not been tested. The current, and hence torque, would need to go to zero briefly, like when the clutch is operated during a manual gear-change.

You would need to write the code for Tritium's Driver Controls Unit (DCU) to decide when to change, and to send the right commands to the WaveSculptor to make it change quickly and smoothly. You could have a large amount of hysteresis in the speeds at which it changed on the way up versus on the way down, to minimise changes. Or you could just have a manual "gear switch" to tell the DCU when to change.

My former employer and mentor Ross Pink did this successfully with a Control Techniques Unidrive VFD in his eVan.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).

Post Reply