Rewinding Toyota Prius MG2 stator

AC, DC, amps, volts and kilowatt. It's all discussed in here
Post Reply
User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Rewinding Toyota Prius MG2 stator

Post by BigMouse »

This thread is an offshoot of the "AC electric motor rewinding for EV use" thread HERE.

I have posted the discussion so far below. Please note I have changed the order or replies slightly to improve readability.
mizlplix wrote: One thought to ponder:

The Prius only uses the electric drive up to a point (not sure), when the I.C.E. motor takes over.

So even when you got it working OK alone, you will have at best a 30-35 MPH car.

If you can over freq. the motor, you might pick up some RPMs there too.

If rewound to 6 or 4 pole to pick up torque, it would worsen the problem.

(Not trying to put a dampner on it)

I would like to try one myself.

Miz

BigMouse wrote: The ICE never "takes over" in the Prius. It only comes on above certain speeds, and always works with the electric motors. The big motor (MG2) is coupled to the drive wheels by a fixed gear ratio, so rotational speed isn't an issue. The motor is capable of spinning (and commutating) at any speed the car is capable of. Otherwise it wouldn't be able to regen at freeway speeds.

The purpose of the rewind would be to lower the voltage in order to maintain constant torque to a higher RPM. It's not really over freq'ing the motor because it commutates at these speeds anyway during normal use, just not with as much current. The rewind would just push the torque fall-off knee to a higher RPM.

I would suspect that the motor can't be rewound to a different number of poles because the rotor has 8 poles with permanent magnets.

mizlplix wrote: Cool, I am far from knowledgeable on these.

I DO know, there are several converters in California doing the plug in thing by removing the ICE and related stuff and adding another pack to the car. So, I would assume the electric motor will RPM just fine.

The first group to do this is a bunch of engineering students. I guess the hardest part was a new VCL for the controller, which they wrote.

(And I suck at)

After reading that PDF on the operation, it seems that a planetary is instrumental in keeping the electric motor within it's RPM range no matter what the ICE does.

It acts as a priority power flow divider and each power trane can operate at it's own speed without interference from the other.

Like the ICE can shut off locking it's input source allowing the Electric motor to provide 100% of the motive force. OR both can operate allowing the power flow to combine to one output source.

If this is correct, then the electric motor will not throttle up and down with the ICE, but run within it's own range. I wonder what that is?

Miz
BigMouse wrote:
The Prius transaxle has two electric motors. One (the big one and the one of interest for EV conversions) is coupled to the differential via a fix gear ratio. It always spins at a constant multiple of the wheel RPM.

The smaller electric motor isn't much use for an EV conversion. It's just there to start the ICE, and transfer the ICE's torque to the wheels when required. It is coupled to the differential via the planetary gears and its RPM is a ratio between the ICE RPM and the wheel RPM (determined by the planetary ratio). It's what gives the transmission its "continuously variable" nature, and serves a role similar to a brake band in a normal automatic transmission, but does its job magnetically rather than through friction.

In a PEV conversion of a Prius, the ICE would be removed, and the smaller electric motor would either be removed as well or the planetary would be welded to allow the two motors to work together. This seems like way more work than it's worth as you'd need a second controller for it and it wouldn't add much power.

Johny wrote: Just in case you haven't seen this, it may be a helpful visualisation aid.
Scroll down the page a bit - then play with the controls.
http://eahart.com/prius/psd/

PlanB wrote: Tnks for the rewind info offer Miz. The tritium controller I'd use is 450VDC and 368ADC input and 320VAC and 300Arms output. The IPM is currently 8 poles & I'm not sure that's alterable because of the permanent magnets?

coulomb wrote:
BigMouse wrote: In a PEV conversion of a Prius, the ICE would be removed, and the smaller electric motor would either be removed as well or the planetary would be welded to allow the two motors to work together. This seems like way more work than it's worth as you'd need a second controller for it and it wouldn't add much power.

I don't know why you'd want to ignore a motor that is 60% of the power of the main motor (30 kW verses 50 kW, though it's hard to find definite numbers for MG1). 50 kW peak, with much less continuous, is pretty light on for a Prius sized car with an EV sized pack. 80 kW peak is much more reasonable.

It's a pain having to drive the two motors with two controllers, but the car comes with two water cooled inverter back ends already matched to the motors (before rewinding, of course). So if you can deal with a 500 VDC pack (you'd need 650 VDC on later models), you get two inverter back ends for free. You do need to provide the inverter front ends, but that's "just software" and a microcontroller.

I'm a little surprised that no-one seems to have done this yet. Perhaps it's the pack voltage.
Johny wrote:
coulomb wrote:I'm a little surprised that no-one seems to have done this yet. Perhaps it's the pack voltage.
If I rememeber correctly the later Prius' pack is around 200 VDC and they use an up-convertor to 650 VDC. So a DIYer would not have to fiddle with the 650 VDC side. I agree - it's a little surprising that given all the reverse engineering that's taken place on the Prius that a BEV hasn't been done this way.

Edit: looked up and changed battery pack voltage

BigMouse wrote: Remember that 50kW number is before the re-wind. Rewinding for even 1/3rd the voltage could push that up to 150kW. Seems hardly necessary to go through the added trouble to integrate MG1 in to it when you have a 150kW 400Nm beast of a motor already connected to the differential.
Johny wrote: ...I think I'm starting to understand (again) why no-one bothers to play DIY BEV with the Prius.
Bigmouse, that 150kW is only valid if you can use the extra Revs.

coulomb wrote:But the boost converter has a limit of ~ 20-25 kW (2004 and 2012 models respectively), as provided. I suppose you could beef up the size of the boost converter; that might be less trouble than dealing with 500 or 650 V packs (for the 2004 and 2012 models respectively).

Johny wrote:
coulomb wrote:But the boost converter has a limit of ~ 20-25 kW (2004 and 2012 models respectively), as provided. I suppose you could beef up the size of the boost converter; that might be less trouble than dealing with 500 or 650 V packs (for the 2004 and 2012 models respectively).
It's getting too hard isn't it. So much reverse engineering when you might as well go from scratch in a car you actually like driving. (IMHO)

BigMouse wrote:
Johny wrote: ...I think I'm starting to understand (again) why no-one bothers to play DIY BEV with the Prius.
Bigmouse, that 150kW is only valid if you can use the extra Revs.
The gear reduction between the wheel and MG2 is 4.113:1 for the 2004 model (3.905:1 for the previous model).
With the standard configuation and the high voltage, the rated torque of 400Nm is available until 1540rpm. Assuming a 0.3m radius on the wheel, that corresponds to 42km/hr.

A re-wind which would allow constant torque to 4000 motor rpm on a reasonable battery voltage would take you to 110km/hr and produce about 168kW.

An 8-pole motor at 1540rpm is operating at 100Hz. For 500V supply, that's 5Vdc/Hz. (note: using DC values for simplicity)

To maintain that Vdc/Hz to 4000rpm (267Hz), we'd need to supply 1335V. Therefore, if we want to accomplish this with a battery pack of say 365V, we'd need to re-wind the stator for about 1/4 the voltage. This would be 1/4 the number of turns per coil and 4x the number of wires in hand, right?

mizlplix wrote: "To maintain that Vdc/Hz to 4000rpm (267Hz), we'd need to supply 1335V. Therefore, if we want to accomplish this with a battery pack of say 365V, we'd need to re-wind the stator for about 1/4 the voltage. This would be 1/4 the number of turns per coil and 4x the number of wires in hand, right?"

Yes, but with an increase in wire gauge also. That thin wire is not up to the higher currents I am afraid. There again, 18 Ga seems the best trade off.

You can simply unwind it and count the coil turns and the in hand bundle. Divide the turns by 4 and mulitply the in hand bundle by 4.

Figure that in hand amount in circular mills and convert it to the 18 gauge wire equivalent. (Use an online chart)

Length (Turns) stays the same at 1/4 the original, inhand count gets smaller, (but really stays the same in circular mills.)

Miz

BigMouse wrote:The slot openenings through which the wire is inserted is very narrow on the prius stator. 18 guage wire may not fit.


Just a note, I'd like to concentrate on the first generation Prius if possible. Their battery packs are reaching end of life at this point (after nearly 20 years!) and so I expect the availability of used, working transaxles to improve and the cost to go down. Also, this generation had a problem with the windings that caused them to overheat as wound, meaning there are also a number of transaxles out there which are otherwise fine but have a stator in need of a rewind already. These should be the cheapest of all to obtain.
Last edited by BigMouse on Wed, 24 Jul 2013, 07:23, edited 1 time in total.
mizlplix
Groupie
Posts: 41
Joined: Sun, 06 May 2012, 05:06
Location: United States
AIM: 17719 E. San Tan Blvd

Rewinding Toyota Prius MG2 stator

Post by mizlplix »

Can anyone measure the stator slot width?
The AC Motor propulsion study group.     www.ivanbennett.com/forum
User avatar
coulomb
Site Admin
Posts: 4324
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Rewinding Toyota Prius MG2 stator

Post by coulomb »

BigMouse wrote: I'd like to concentrate on the first generation Prius if possible. Their battery packs are reaching end of life at this point (after nearly 20 years!) and so I expect the availability of used, working transaxles to improve and the cost to go down.

Yes, if you can get them, they should be dirt cheap. Toyota did offer new replacement packs to New Zealand customers only (one of the three places that they concentrate: Australia, UK, and New Zealand). But no-one seems to know where these packs came from, so we can't seem to do the same in Australia or elsewhere.

So yes, these are ripe for full electric conversions, or even hybrid electric conversions when replacing the NiMH pack with lithium. (The engines seem very reliable, except for the MAF sensor, which often has to be carefully cleaned).

But the issues with the NHW10 and NHW11 models (pre-2004) are

* They are old cars now, some 16 years old. Colour choice is limited.
* The motors are less powerful than in the later models. However, this may be a lack of "overvoltaging"; they don't have a boost converter, and ran with a 288 V nominal hybrid pack (more like 312 V in normal use, up to 360 V on charge).
* The air conditioning ran off the ICE, so you'd need a separate motor to run the A/C compressor, or do without, or replace with an NHW20 (2004+) version.
* The multi-function screens are in Japanese, and are invariably clapped out.
* Few creature comforts (e.g. no Sat Nav).
* The NHW10 model were never imported into Australia, so parts may be hard to find. However, the NHW11 model was, so most parts excluding the hybrid battery may be interchangeable with the NHW11.

On the plus side:

* Many owners are having trouble with the hybrid battery, and will get rid of them cheap.
* Since the motors were designed to run with a ~ 320 V pack, it may not even be necessary to rewind them. Or possibly a star-delta rewire may be possible.

Good resource here:
http://autos.groups.yahoo.com/group/Mk1_Prius/ .
Free registration required; from memory it takes a day or two to get approved.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 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
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Rewinding Toyota Prius MG2 stator

Post by BigMouse »

According to the ORNL report for the 2010 prius, the slot width is 1.88-1.93mm. 18 gauge wire is just over 1mm. So not as tight as I thought. I'd read somewhere that it was an exceptionally small gap and got the impression only fine wire would fit.

From the same source, the wire gauge on the 2004 prius was 19.

Other data for the 2004 Prius:
Slots: 48
Turns per coil: 9
Paraller circuits per phase: 0
coils in series per phase: 8
Wires in parallel: 13

I can't find any winding data for the earlier Prius other than the following statement:

"In dissecting the motor, ORNL researchers assessed that the windings of the 2004 and 2003 models have the same gauge wires, same number of turns per coil, same winding distribution, and the same stator punching. The only difference between the stators is that the windings are connected in series instead of in parallel in the 2004 model."
User avatar
coulomb
Site Admin
Posts: 4324
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Rewinding Toyota Prius MG2 stator

Post by coulomb »

BigMouse wrote: I can't find any winding data for the earlier Prius other than the following statement:

"In dissecting the motor, ORNL researchers assessed that the windings of the 2004 and 2003 models have the same gauge wires, same number of turns per coil, same winding distribution, and the same stator punching. The only difference between the stators is that the windings are connected in series instead of in parallel in the 2004 model."

I've been trying to parse that statement for years. Presumably they are comparing NHW11 and NHW20 models. Are they talking about a delta to star change, with the NHW20 motors in star? Perhaps the other way around?

I note that 500 / 1.7321 = 288.7, so a star winding change would make sense when going from a 288 V pack and DC bus to an up-to-500 V DC bus.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 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
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Rewinding Toyota Prius MG2 stator

Post by BigMouse »

I think it's just a matter of having poles on the same phase in parallel rather than series. They use the terms "double" and "half" a few times in the same section talking about voltages and currents. It's not clear whether they're just making an example, or whether they're referring specifically to the motor. I wouldn't be surprised if each phase is wound as two halves. For example, all the "north" poles and all the "south" poles wound together, then paralleled when terminated at the motor lead junction. This would make sense as they point out that the leads are much thinner on the later model, meaning there are fewer wires brought out to that point. Supposedly wiring poles in parallel is a common method of reducing motor voltage ratings.

From all the photos I can find of stators for the gen 1 and gen 2 prius, they appear to be very obviously wired in star configuration. Converting to a delta winding would be another way to bring the voltage down.
mizlplix
Groupie
Posts: 41
Joined: Sun, 06 May 2012, 05:06
Location: United States
AIM: 17719 E. San Tan Blvd

Rewinding Toyota Prius MG2 stator

Post by mizlplix »

When going from Star to Delta on an industrial dual voltage motor, the change was figured into the motor design and the current change was allowed for.

The same can not be said for a single voltage motor when changed from star to delta. The current change will be double and place it in the danger zone for overheating or outright failure.

The in hand count is 13.

What is the wire diameter ?

Then we can at least guess if it will survive.

Miz

Last edited by mizlplix on Wed, 24 Jul 2013, 18:19, edited 1 time in total.
The AC Motor propulsion study group.     www.ivanbennett.com/forum
User avatar
coulomb
Site Admin
Posts: 4324
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Rewinding Toyota Prius MG2 stator

Post by coulomb »

mizlplix wrote: The same can not be said for a single voltage motor when changed from star to delta. The current change will be double and place it in the danger zone for overheating or outright failure.

I don't believe so. Whatever current you push into the motor, it will be split such that 1/sqrt(3) = 0.577 will go into the two windings connected to the terminal. Kirchoff's current law.

It's similar to arranging resistors or windings in parallel instead of in series; it rearranges the currents and voltages in such a way as to change the effective impedance. In the case of parallel to series the share of the current changes by 2:1; in the case of star to delta, it's sqrt(3) = 1.7321 to 1. In both cases, for the same overall power, the current through the impedances (here, windings) is the same (as indeed is the voltage).

Edit: the above paragraph is wrong, sorry. For the same power, the current remains the same. You are just changing the overall current and voltage, by either 2:1 or 1.7321:1.

For the same torque and hence the same power (at the same speed), the delta motor will require 1.7321 times less voltage, and 1.7321 times as much current, but each winding will see the same voltage and current. To get more power, you can apply more voltage and speed with the same current (which is 1.7321 times as much as you needed when wired in star), and you get more power from the same motor. (The usual overclocking / over-frequency / over-voltage thing; the iron losses will of course go up while the copper losses remain the same.)

Oh, I also meant to say: if the NHW20 motors really do have two windings in series (edit 2019: was parallel), then they can presumably be rewired (as opposed to rewound) with minimal effort and cost so that they can operate effectively at about 250 VDC bus, and with a bit more power at 320 or 360 VDC. These latter numbers, while high by EV conversion standards, are a lot more practical than 500 VDC.

I'd really like to see this attempted.
Last edited by coulomb on Wed, 24 Jul 2013, 19:01, edited 1 time in total.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 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: 4324
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Rewinding Toyota Prius MG2 stator

Post by coulomb »

coulomb wrote: ... if the NHW20 motors really do have two windings in parallel, then they can presumably be rewired (as opposed to rewound) with minimal effort and cost ...

And perhaps rewired to parallel AND delta, for a total factor of 3.46, or nominally 144 VDC instead of 500 VDC... interesting! (And 188 VDC for the later 650 VDC nominal motors).

But star to delta may invite circulating currents if the windings are not perfectly matched. I believe that nearly all AC generators are wired in star for that reason. But maybe the loss of efficiency that these circulating currents would induce would be worth the convenience of the much lower required pack voltage (and/or higher ability to overclock).
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 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: 4324
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Rewinding Toyota Prius MG2 stator

Post by coulomb »

mizlplix wrote: When going from Star to Delta on an industrial dual voltage motor, the change was figured into the motor design and the current change was allowed for.

Ah. With the industrial motors, presumably these are designed for a fixed line voltage, and hence the current will increase a lot when the back EMF is reduced by a whopping 1.73x. So your comment is presumably valid in that context, but when a variable motor drive voltage is available, I don't believe that it holds.
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 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
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Rewinding Toyota Prius MG2 stator

Post by BigMouse »

coulomb wrote:I'd really like to see this attempted.


I'd really like to attempt this! I've been curious about the cost and logistics of shipping KDRYAN's two prius transaxles (2nd gen I assume, they're from taxis) down to Wollongong to play with. Too many projects at the moment though.
User avatar
carnut1100
Groupie
Posts: 334
Joined: Tue, 24 Feb 2009, 16:39
Real Name: Greg Milligan
Location: Brisbane
Contact:

Rewinding Toyota Prius MG2 stator

Post by carnut1100 »

I have read of a guy in America who has used a prius transaxle in a conversion, but he had to do his own controller for the prius inverter.
gtyler54
Noobie
Posts: 26
Joined: Mon, 13 Aug 2012, 00:54
Real Name: George Tyler
Location: Hamilton NZ

Rewinding Toyota Prius MG2 stator

Post by gtyler54 »

the NHW10 uses "field weakening" to get to higher RPM than base RPM, it does this by using buried magnets in the rotor and FOC. One component of the current fights off the PM field, increasing the RPM per Volt rating. not sure what the base RPM is, and the only way I can see to use this is to used a FOC inverter. A pity, but I don't know how to begin to access the prius inverter to that extent.
gtyler
Post Reply