Prius NHW20 transaxle

Technical discussion on converting internal combustion to electric
Post Reply
User avatar
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Sun, 03 Oct 2010, 00:16

I was intrigued by this DIY electric car topic:

Old Prius electric engine?

It seems that NHW11 transaxles (the earlist models officially exported outside Japan) are becoming available cheaply. Even more interestingly, it is claimed that
"Toyota will sell a brand new genII Prius transaxle with the 67kw BLDC motors for $3,300 retail."
Presumably, that's US$ and sourced in the US; we could expect to pay maybe AU$5000 (I have no idea). In fact, as I point out, it's probably 67 horse power for MG2 alone, or 50 kW, with another 30 kW normally input through MG1. Only about 20 kW is designed to come from the small (6.5 Ah) hybrid battery, about 100 A @ 200 V. This is boosted to around 500 VDC by the DC/DC boost converter. The limited power availability of the hybrid battery and limited power of the boost converter are one of the main reasons that a Prius as is doesn't make a very good all-electric vehicle.

However, there are You Tube movies of a Prius with an after-market pack being driven all-electric at 70 mph (112 km/h). Maybe 20 kW is enough to move a Prius at that speed; I guess so.

It would be easiest if it wasn't necessary to modify the Prius transaxle. Let's bypass the 20 kW boost converter, and run it directly off a ~500 VDC pack. That's a lower pack voltage than some AC conversions I could name. 50 kW from MG2 + 30 kW from MG1 gives a theoretical peak output of 80 kW, enough for many conversions. But what about the motors overspeeding?

From this nomograph (adapted and extrapolated to 10,000 RPM from http://prius.ecrostech.com/original/Und ... Device.htm:

Image

it seems that the Prius transaxle, with no rotation of the ICE input shaft, is capable of about 110 km/h if MG1 is allowed to spin at 10,000 RPM. Indeed it is, according to the PDF available from this page:

http://www.osti.gov/bridge/purl.cover.j ... 29-WIfqPO/

Edit: however, it seems to be software limited to 6500 RPM if the ICE is not spinning.

So with a lot of assumptions, one might be able to get a new or second hand transaxle including diff and fixed ratio gearbox capable of some 80 kW that is capable of 110 km/h. You could possibly adapt the existing Prius inverter, which also has an inverter for driving electric air conditioning (but it needs 200 VDC) and a DC/DC (200 V to 12 V). Perhaps the boost converter could be reconfigured to run as a buck converter; after all, when a Prius does high speed regen, the power goes through that converter in that direction.

You could possibly use that DC/DC boost-converter-in-reverse to run a smallish DC motor as well, connected to the input shaft where the ICE normally connects, possibly through a smallish DC controller. That would lift the brown line at the right hand end, allowing a little more top end speed if desired. But that seems to me to be unnecessary.

From the linked PDF, MG2's continuous power rating is 30kW, which is quite handy. If MG1 has a proportionate continuous rating, that would make it about 18 kW continuous, for a total of 48 kW continuous. The Prius comes with a separate cooling system for the inverter and motors, which might be available from the same source as the transaxle.

Earlier models of the Prius have considerably lower power motors, but also lower voltage, as shown in the Wikipedia Prius comparison of models page. Smaller vehicles might be able to use the 33 kW from an NHW11 transaxle, plus whatever power is available from its MG1 (perhaps as low as 15 kW peak). That's still 48 kW peak, and the pack voltage would be around 273 VDC.

Of course, it will be necessary to figure out how to drive the Prius inverters without the ECU, but presumably it just has an input for RPM, as suggested by Figure 4.3 of the linked PDF. It might still be necessary to figure out some CAN addresses and formats, but it doesn't sound too drastic. Or of course, a separate pair of inverters could be used, but presumably used Prius inverters would be much more cost effective. The cooling system might need some reverse engineering as well.

I note that the linked PDF mentions field weakening (by which I assume they mean flux weakening; the Prius motors are Internal Permanent Magnet).

So for someone who was interested in a AC or BLDC conversion, maybe this is an option?
Last edited by coulomb on Sun, 24 Oct 2010, 03:57, 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.

bga
Senior Member
Posts: 492
Joined: Mon, 01 Sep 2008, 19:27
Real Name: Bruce Armstrong
Location: Perth WA

Prius NHW20 transaxle

Post by bga » Sun, 03 Oct 2010, 17:10

Hi Coulomb,

Yuk! that's a horrible page for the power split device.

Here'a a much better one that makes sense:
http://eahart.com/prius/psd/
It also has some levers to play with.

70 MPH....
Wait a moment, are these the guys that welded the gears up so MG1 and MG2 turn at the same speed?

I haven't looked at it in detail, but it changing the sun and planet gears to get a lower MG1 speed (bigger sun gear) may do the trick and allow better utilisation of MG1.
Last edited by bga on Sun, 03 Oct 2010, 06:10, edited 1 time in total.
It's not the end of the world, but I can see it from here.

User avatar
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Sun, 03 Oct 2010, 19:02

bga wrote: 70 MPH....
Wait a moment, are these the guys that welded the gears up so MG1 and MG2 turn at the same speed?

I don't believe so. The 70 mph apparently comes exclusively from http://pluginconversions.com/ (the web site is useless: no details, not even battery capacity, at least that I could find). They apparently somehow use MG1 and MG2 at the same time, and allow the ICE to spin with no fuel.

I suppose that's another possibility: put a huge governor or the like onto the input shaft, so that it can spin but at a limited speed. But that seems silly as well, not to mention space consuming and dangerous.

The only project I know of that has welded the gears is EV Insight with a Prius Heart. There aren't many details there about how the car performed, and the site doesn't seem to have been updated in years. A pity, really.

According to the gadget on the page you mention above, just under -10,000 RPM on MG1 with 0 on the ICE input results in 65 mph, which is only 104 km/h. That's a bit on the low side; I think a practical car needs 110 km/h capability for those occasional trips down highway one.

Edit: fixed formatting.
Last edited by coulomb on Sun, 03 Oct 2010, 09:01, 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.

bga
Senior Member
Posts: 492
Joined: Mon, 01 Sep 2008, 19:27
Real Name: Bruce Armstrong
Location: Perth WA

Prius NHW20 transaxle

Post by bga » Tue, 05 Oct 2010, 02:23

Yes, that's the one I was thinking of.

I was having some more thoughts on the gears.

MG2 is about 50kw, MG1 is about 15, so ignoring MG2 only makes it a bit more feeble. Perhaps, this could be done by removing the planet gears, which decouples MG1 but reduces the maximum power to 50kW. It also gets around the unconstrained ICE shaft, which may be removed. I saw some comments sbout the oil pump that may complicate things slightly.

Perhaps two prius drives can be grafted so that the MG2 from the second is put in place of the ICE. That'd make 100kW, so it should be able to go fast.

It even sounds "almost possible".

some Prius drive cutaways are here

[edit: fixed 'plant gears']
Last edited by bga on Mon, 04 Oct 2010, 15:24, edited 1 time in total.
It's not the end of the world, but I can see it from here.

User avatar
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Tue, 05 Oct 2010, 06:08

bga wrote: MG2 is about 50kw, MG1 is about 15,
Well, I'm claiming 30 kW for MG1, based on the PDF link above.

It may well be 15 kW for the NHW10 and NHW11, but they have smaller MG2 as well (30 and 33 kW).

For both series, therefore, ignoring MG1 is a bit much. Assuming I am right.
so ignoring MG2 only makes it a bit more feeble.
I assume you mean MG1 here.
Perhaps, this could be done by removing the planet gears, which decouples MG1 but reduces the maximum power to 50kW.
An interesting idea.
I saw some comments about the oil pump that may complicate things slightly.
I saw somewhere that suggested that the oil pump was connected to MG2, i.e. to the output (to the wheels via the diff). So if true that would not be a problem. [ Edit: I now believe that the oil pump is connected to the ICE shaft, so it would be a small problem. ]
Perhaps two prius drives can be grafted so that the MG2 from the second is put in place of the ICE. That'd make 100kW, so it should be able to go fast.
Interesting. You'd need to remove the diff on one of them. But you'd end up with a CVT on the other (assuming you kept MG1 on that one).
In the series 1 cutaways, MG1 sure looks small compared to MG2. Not so much with the series 2.
Last edited by coulomb on Sun, 18 Dec 2011, 03:51, 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
woody
Senior Member
Posts: 1711
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

Prius NHW20 transaxle

Post by woody » Tue, 05 Oct 2010, 06:53

coulomb wrote:
I saw some comments sbout the oil pump that may complicate things slightly.
I saw somewhere that suggested that the oil pump was connected to MG2, i.e. to the output (to the wheels via the diff). So if true that would not be a problem.
Err, oil pump for what?

ICE oil - who cares, the ICE is gone.

Gearbox (i.e. planetary + diff oil) - is probably not connected to MG2, as you probably need oil pressure when not moving (e.g. charging). If it's common oil for both motor and diff, you may not need it for the diff / planetary anyway - depends on how it's designed.

Most manual gearboxes and diffs just drag/splash the oil all over everything by the motion of the gears through a wet sump.

Ice motors usually pump the oil directly into the bearings... Maybe Toyota did this for the prius gearbox too.
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack

User avatar
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Prius NHW20 transaxle

Post by 7circle » Tue, 05 Oct 2010, 06:55

I have been wondering how the Prius EV mode copes or mantains a zero ICE RPM. As the MG2[edit:MG1) that starts the Motor is a acting in a See-Saw balance with the ICE, like a differential, where you spin one wheel and the other spins the other way. To move the output drive shaft the ICE needs to at least oppose the torque from MG1.

Then after reading the PHV Prius arcticle
which desrcibes EV speeds of 65MPH I thought does the ICE shaft have like a oneway (ratchet/clutch) lockup like on a bike hub.

This would protect the motor against reversal and give a stationary planet-carrier so the MG1 can spin out to neg 6500 RPM.

I hadn't heard of a ratchet action in the Prius transmission can anyone confirm?

I read the oil pump was connected to the engine planet gear and when the planet gears were welded to the planet frame the pump would operate allowing EV mode with no ICE. but I thinks someting else needs to be operated than just the auxiliary oil pump.

Now this makes me wonder about the PHV Prius as it can run without the ICE shaft turning.
Do the gears and the motors need the oil pump?
The transmission oil may be a different oil than the engine and oil pump.

[Edit: MG2 is fixed ratio to the wheels, meant to say MG1 as at end of 1st paragraph]
[Edit: The Prius only needs less than 20KW to over come wind drag at 65 MPH, so only the MG2 motor is needed, but performance would be very slack]
Last edited by 7circle on Thu, 07 Oct 2010, 05:45, edited 1 time in total.

User avatar
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Tue, 05 Oct 2010, 14:46

7circle wrote: I have been wondering how the Prius EV mode copes with or maintains a zero ICE RPM. As the MG2 that starts the Motor
It's MG1 that starts the motor. MG2 is connected to the output.
is a acting in a See-Saw balance with the ICE, like a differential, where you spin one wheel and the other spins the other way. To move the output drive shaft the ICE needs to at least oppose the torque from MG1.
Here is my take on how it works.

There seems to be a defined torque split from the ICE to MG1 and MG2/output.

Edit: I now believe that this is wrong; see below.

It's something like 0.72 ICE to output, and 0.28 ICE to MG1. So when the ICE has no fuel and therefore produces no torque, zero torque is transmitted to MG1 or to MG2. Well, actually compression (which is lower in the Prius engine because of the Atkinson cycle) transmits a slight drag to MG1 and the output. It takes 30 seconds or so for the ICE spindle to spin down to actually zero RPM; you can actually feel a slight shudder when it does spin down to very low revs. MG1 may actually keep the ICE at a low RPM that is ready to provide power if needed in an instant, like 1500 RPM or so. So when the ICE is off, MG1 cannot (in normal hybrid mode) contribute any torque to the output.

The torque split described above is something I read from a web site, and it seemed like a guess from the author. There is definitely a simple subtractive relationship between the 3 shafts' speeds, but when I try to figure out how the torque distributes, my brain freezes.
Then after reading the PHV Prius arcticle
which desrcibes EV speeds of 65MPH I thought does the ICE shaft have like a oneway (ratchet/clutch) lockup like on a bike hub.
That would be awfully inefficient, I think. Besides, there is no need for that; MG1 can spin to various speeds with various torques to make sure that the torque and power goes where it is needed.
Now this makes me wonder about the PHV Prius as it can run without the ICE shaft turning.
Do the gears and the motors need the oil pump?
The transmission oil may be a different oil than the engine and oil pump.

Yes, I wish I could get some details on how the Toyota PHEV works, though I haven't looked in a while.

I'm 99% certain that the transaxle oil is separate from the engine oil. I believe that the oil in the transaxle is mainly for the diff and the two idler gears that connect MG2/epicyclic output to the crown wheel of the diff. It may also assist with motor cooling, but the motors have a separate cooling system (shared with the inverter, but separate from the engine's cooling system).
Last edited by coulomb on Tue, 05 Oct 2010, 06:13, 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
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Tue, 05 Oct 2010, 15:10

coulomb wrote: Here is my take on how it works...

The torque split described above is something I read from a web site, and it seemed like a guess from the author.

Thinking about it after that last post, it seems obvious how it should work now. It is the speed ratios that are fixed. The only variables apart from the ICE speed and power are MG1s speed and power. MG2's speed is set by the speed of the vehicle, tyre diameter, and differential ratio. The torque required by MG2 is whatever is demanded by the throttle position.

So the main thing is MG1 can precisely set the speed of the ICE shaft. If you want a certain torque and speed from the ICE, you arrange MG1 to bring the ICE shaft to that speed, and arrange the fuel as needed. If the torque from the ICE is a bit low, you increase fuel; if high, you reduce it. Maybe they don't bother measuring the ICE output torque, and just run that part of it "open loop". MG1 just provides whatever torque (often negative) is needed to maintain the speed it needs. When MG1's torque is negative, that means it's generating, and its power plus about 20 kW from the battery via the boost converter is what is available to MG2.

When it's operating like that, the torque probably does split according to the ratio that author describes.

Edit: I now believe that this is wrong; see below.

But that doesn't happen automatically; it happens that way because MG1 is spinning with just the right speed and torque to make it happen. So the torque splitting is an effect; the speed splitting is the cause and is forced by the gear ratios.
Last edited by coulomb on Tue, 05 Oct 2010, 05:22, 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
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Tue, 05 Oct 2010, 16:58

Ok, I think I know how the power and torque splits now; sorry for all the false information above. Some of it comes from another site, where I now believe the author was wrong.

The planetary gearbox ensures a ration of speeds; that's it. In fact, the ratio happens to be

S = 3.6 * C - 2.6 * R

where S is the speed of the Sun gear, connected to MG1
C is the speed of the carrier, connected to the ICE
R is the speed of the rin gear, connected to the output (input of diff).

Rearranging a little,

speed of MG2 = speed of output = 1.38 * speed of ICE - 0.38 * speed of MG1.

Energy must be conserved, so at any instant of time, power (rate of moving energy) must also be conserved. So MG1 has to supply the difference in power between the output of the ICE and the power from MG2.

The Prius transaxle's main purpose is as a continuously variable transmission. The ICE can maintain its comfort zone of 1000-4500 RPM, while the power to the wheels remains roughly constant. In other words, it arranges for more torque at low speed, and less torque at higher speed (at the diff/wheels).

First consider low vehicle speed. Maximum torque is required to accelerate quickly. In a conventional vehicle, this is done with a high gear ratio (not tall). It amplifies the torque from the ICE, at the expense of vehicle speed. In the Prius, MG1 takes away a lot of the power from the ICE, spinning quickly, so that the little torque it takes from the motor is converted to a lot (~ 30 kW) of power, which is basically fed to MG2 to provide maximum torque. So MG1 effectively "robs speed" from the motor, and "donates torque". Here, MG1 is a generator, taking mechanical power and producing electrical power.

Now consider high vehicle speed. We can't provide full torque to MG2, we just don't have the power for that. With a conventional transmission, a low ratio (tall gear, probably overdrive) is used, which does not amplify, indeed reduces the torque from the ICE. In the Prius, some of the torque from the ICE delivered to the output is taken by MG2, now acting as a generator, to supply power to MG1. MG1 is now adding speed to the motor, and MG2 is taking away torque from the output to supply power for MG1.

So at low speed, MG1 is taking mechanical power and MG2 is supplying mechanical power.
At high speed, MG1 is producing mechanical power, and MG2 is taking mechanical power.

In both cases, power is coming out of the ICE, except for a fraction of a second when it needs to be started. So the torque is obviously not split in a constant ratio between MG1 and MG2; indeed the direction of torque in both motors changes direction depending on vehicle speed and demanded acceleration.

Of course, it gets a little more complex when you add in the power to/from the hybrid battery pack, and regeneration. But the fact remains; the power and torque can and do change directions, mainly for the purpose of matching the ICE output to the requirements of the vehicle.

Final summary: the planetary gearbox sets the speed ratio only. The torque and hence the power split is determined dynamically by the hybrid system, mainly to match the power output of the ICE to the speed and torque requirements of the vehicle.

For the purposes of using the Prius transaxle as a motor and differential, the planetary gearbox is essentially a nuisance. We can force the ICE shaft to be zero RPM (welding it or otherwise immobilising it), but then MG1 has to spin very fast (10,000 RPM at 104 km/h). The other way is to weld the planetary gears to the carrier, so the whole planetary gearbox spins at the same speed, and we just ignore the ICE input shaft (or use it to drive air conditioning or power steering). Then both MG1 and MG2 are spinning at a reasonable 6000 RPM at a vehicle speed of 160 km/h, or ~ 3500 RPM at 100 km/h. At least MG2 and almost certainly also MG1 uses flux weakening, so a separate gearbox is not needed; the motors are constant torque to base speed, and approximately constant power above base speed (with some power droop at higher speeds). The flux weakening is automatically handled by the Prius inverters. In EV mode, the inverters would be driven with their speed inputs (i.e. you request a certain speed, and the inverters do their best to achieve that speed), probably using current sensors as feedback.
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.

bga
Senior Member
Posts: 492
Joined: Mon, 01 Sep 2008, 19:27
Real Name: Bruce Armstrong
Location: Perth WA

Prius NHW20 transaxle

Post by bga » Wed, 06 Oct 2010, 04:46

I think the above more or less has got it.

It's easy to confuse MG1 and MG2. I'm assuming that MG2 is the big motor, attached to the wheels (indirectly) and MG1 is the small motor off the PSD pinion.

MG == Motor-Generator?

Good point here about the relation between MG1 and the ICE. It is refrred to as the 'Starter' and the 'Generator' in various documents.

ICE's don't like turning backwards or slowly, so the choices for it are stopped or somewhere in its range of 1000 to 4000 RPM approximately.

When the ICE is stopped, the torque from MG1 can only be zero or positive (starting) and when the ICE is running, the torque can only be zero or negative (generating). If it's zero, the ICE should be stopped.
I have discounted negative (starting) torque on the running motor, since it will only speed the idling motor up and waste energy. At higher road speeds, gearing and motor limitations mean the mimimum ICE speed will be more than it's idling speed (so much for that 'rule').

This means that, MG1 (small) is almost always generating then the ICE is running. Its output either goes to the battery or MG2, allowing it to charge the battery or propel the vehicle.

The torque being absorbed by MG1 is a fixed ratio to the torque being output by the ICE, which is also a fixed ratio to that being transferred to MG2 and the wheels. This must always be the case, but the shaft speeds may vary, according to the ratios mentioned in coulomb's post.

On to the equations of state...

It's not the end of the world, but I can see it from here.

User avatar
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Prius NHW20 transaxle

Post by 7circle » Thu, 07 Oct 2010, 07:40

(edit - bit of a ramble below so here are some points)
(1 - do the motors have different nominal voltages eg say 6500 RPM ?)
(2 - how long would the planetary gear last if it was under load all the time? edit Spinning the planet gears transferring power.)
(3 - Would having a third motor to replace the ICE be beneficial? To efficiency.)
(4 - Can over revving the motors over volt the DC bus?)

It's now hit me in the head that when the Prius is in EV mode and the ICE is a no load or off, all the Drive power goes through bigger MG2 and is coming from the NiMH battery via the boost converter. (Edit thanks to the comments and issues raised in this post)

So MG1 is just idling along, maybe with a trickle of power to keep it at the correct RPM holding the ICE at zero RPM.

Getting ahead of things with the new Toyota Plugin that was described about on Prius Chat. I thought that the BOOST inverter may have been upgraded to allow triple the power from the three lithium packs. This would have givien the EV mode plenty more power to keep up with traffic with no ICE running. But this system still using the one motor for all power.

It seems such a waste to have two motors and controller but only able to use one. So that's when I thought if the transmission (PSD) had a way like new auto transmisions do, of locking the planetary ring to the housing then both motors could be powered to push or regen-brake the car.

But its likely not a reality after reading plenty more on the PSD.

So standard EV mode was 20KW, then with triple boost converters and battery packs it went to 60KW still going to one motor (MG).

Now if the full power from two motors is not enough and the ICE was replaced by a third motor (MG3)

So I was wondering what efficiency benifits could be had by only powering one motor at a time until the motor was operating at an RPM and power level that was more than 88% efficient.

http://techno-fandom.org/~hobbit/cars/ginv/VH.html
http://techno-fandom.org/~hobbit/cars/ginv/nomv.gif
Image

{I think this eg below I started to write is a bit dopey, keeping all motors operating together would be better. But it made me think about voltages needed.)
So MG2 with 50KW runs first up to 20MPH (1170RPM), with MG3(ICE) held at zero by and MG1 at -3042RPM.
Then for more power MG1 is driven harder to rev ou to -6500RPM MG at 0RPM and MG2 now running at 2470RPM or 42MPH .....


The diagram above from Smokey's website, adds lines for peak and RMS voltages for the motors for the range of RPM.

This suggest that the MG1 smaller motor has half the turns as MG2 the larger motor. So the choice of modifying the PSD to be just an EV drive (no ICE) may make locking the planetary ring to zero/frame will allow the ratio
"
speed of MG2 = speed of output = 1.38 * speed of ICE - 0.38 * speed of MG1. " becomes MG2 = speed of output = - 0.38 * speed of MG1.
to work effectively with the DC-Bus voltage and Kv of the motors.

If the planetary gears were locked the MG1 and MG2 would spin at the same RPM and MG1 would only need about half the BUS voltage and the FETS would only be at half duty. But thinking again if the motors are in torque control (ie Current control) then the current for MG1 would be half MG2.

Its a bit more complicated due to the double size of MG2 to MG1 so the
windings are longer but may be paralleled.

mmm... MG1 at 25kW and MG2 at 50kW. and RPM at 65 MPH what are the voltages and currents.. more home work.

Well the other issue is if the PM Sync 3Phase AC motor is being run over base speed with flux-weakening ( by using the motor inductance like a step up converter) what would happen if the motor is running at 10,000 RPM from a 600V dc bus and the motor needs to go into regen.

I'm concerned that the higher RPM may drive the DC-BUS over the ratings of the Capacitors on the Bus and Capooff. If the controller was to loose it (stop driving the gates) and the the diodes just conducted the current to the lower bus voltage, it would n't take much to fully charge them.

More to think about.
(Edit: I need to stick to one issue at a time .... and find some poles not just hurdles)
Last edited by 7circle on Thu, 07 Oct 2010, 05:52, edited 1 time in total.

User avatar
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Sat, 09 Oct 2010, 00:05

coulomb wrote: I saw somewhere that suggested that the oil pump was connected to MG2, i.e. to the output (to the wheels via the diff).

In fact, it was an Ebay advertisement for a Prius manual on CD-ROM:

Image

In fact, I've bought the CD-ROM, and am combing it for nuggets of new information. Nothing far reaching so far.

Edit: the ebay link is http://cgi.ebay.com.au/ws/eBayISAPI.dll ... 0489445903.
Last edited by coulomb on Fri, 08 Oct 2010, 13:21, 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
coulomb
Site Admin
Posts: 3606
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Prius NHW20 transaxle

Post by coulomb » Sat, 09 Oct 2010, 00:18

7circle wrote: Well the other issue is if the PM Sync 3Phase AC motor is being run over base speed with flux-weakening ( by using the motor inductance like a step up converter) what would happen if the motor is running at 10,000 RPM from a 600V dc bus and the motor needs to go into regen.
Well, you make sure the nominal voltage is a fair bit less than the maximum voltage of the capacitors. In one of the linked articles, it seems to indicate that the voltage is clamped very effectively at 500 VDC. The inverters use flux weakening and all that to make sure that this happens.
I'm concerned that the higher RPM may drive the DC-BUS over the ratings of the Capacitors on the Bus and Capooff. If the controller was to loose it (stop driving the gates) and the the diodes just conducted the current to the lower bus voltage, it wouldn't take much to fully charge them.
I read a paper on that recently. It turns out that if you design the motor for optimal constant power region, then the maximum current from the motor back to the DC supply is very close to rated current. So the IGBTs will survive; if (and it's a big if) the battery cab absorb rated current, then it will protect the capacitors.

So in an EV situation, with a nominal 500 V pack of lithiums (say 156 cells for 156 * 3.2 = 499 V), then if the cells only go to 3.6 V, then the caps will only see 156 * 3.6 = 562 V. The cells could go to 600 / 156 = 3.85 VPC without damaging the capacitors. I believe that the capacitors could withstand a few seconds of overvoltage, at say 156 * 4 VPC = 624 V. You would not want to do this to the pack or the capacitors or the motor very often, but I think there is a good chance of avoiding the "Capoof" in this hopefully very rare situation Image
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
Mesuge
Groupie
Posts: 168
Joined: Wed, 05 Mar 2008, 07:37
Contact:

Prius NHW20 transaxle

Post by Mesuge » Fri, 15 Oct 2010, 13:02

I guess we have already discussed it here:
viewtopic.php?p=27049&t=1661#p27049

It seems the best practical way how to utilize Toyota's (or Ford, Nissan licensed) HSD components, is to use the MG2 only. The actual power depends on the package ~50-120kW (Prius, SUV, ..), the OEM powerstage being controlled via new ECU board ala Tumanako. Or by completely new powerstage and ECU/Tumanako combo, so you are keeping only the motor (MG2). It has been "sort of" done by some aussie school and Miata donor actually if I'm not mistaken, link is burried in some of my related posts..
edit here (.pdf report):
viewtopic.php?p=23589&t=1390#p23589
http://www.eleceng.adelaide.edu.au/rese ... ects/aecv/


I'm still wondering to this day there is no interest in this approach, yanking second hand HSD emotor and mating with quality powerstage (ala Semikron Tritium's supplier) and opensource ECU, or the cheaper option reusing oem HSD powerstage + new driving board. It's likely the least expensive and diy replicable entry into ~50-120kW AC segment there is for years, simply untapped potential.

One would hope that someone like James might launch such solution in "kit form" and fill the void in his <100kW product line. The sales for these entry kits would be quite on different level than the more $hiny/powerfull stuff.
Last edited by Mesuge on Fri, 15 Oct 2010, 03:22, edited 1 time in total.
ImageCzech EV Club | blog, gallery, video | http://old.elektromobily.org

antiscab
Senior Member
Posts: 2327
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Prius NHW20 transaxle

Post by antiscab » Sat, 16 Oct 2010, 03:54

The Camry hybrid drivetrain would also be interesting.

its almost the identical setup, however, the MG1 is rated to ~110kw, and MG2 is ~40kw IIRC.

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

User avatar
Mesuge
Groupie
Posts: 168
Joined: Wed, 05 Mar 2008, 07:37
Contact:

Prius NHW20 transaxle

Post by Mesuge » Sat, 16 Oct 2010, 04:59

Yep, that's exactly the point. I guess the biggest HSD from Toyota/licensed goes upto ~123kW (I listed these donor cars elsewhere). Another possible advantage of these larger perm MGs is that when using tranny retaining application, you can get away with ~50% power output and motor base speed reduced to ~2000rpm, tops at roughly 1.5x that, ~3000rpm, speaking about the lowest treshold ~320VDC on the inverter, for 450VDC WS make it 3000-4500rpm and ~75% output. These OEM hybrid puppies are meant to be operated at ~500-600V, therefore need to apply some derating when we don't have the voltage capable inverters for it. Look up the table for Ben's perm "EVO brand" emotor/inverter they show a table for it. So, that's ~60-90kW peak perm (via tranny) for very little money, quite suitable for some city econoboxes and perhaps even selected slow highway flyers too.

Now the question where to get mid/high voltage PMSM motor control for it, while not paying ridiculous "OEM-only tienda ~ you diy beggars go away prices" which UQM-likes posts around. T_James said something to the effect, HV version of his inverter possible in the future (PMSM already implemented in current ver.), e.g. after 1-2yrs from now..

So, the remaning option is to reuse from that particular donor its stock/OEM power stage matched for the particular PMSM, and debug it on new driving board ala Tumanako project. That HSD/OEM control board seems rather useless and hack protected, since it's also messed up with integrated functions for boost converter, 2x inverter channels (6x) for MG1&MG2 combo, DC-DC and the overall hybdrid drivertrain application with assorted I/O jungle.

Yes, it's possible to buy readymade Semikron HV powerstage and go from there, but there is little or now savings on this route, when several inverter products like by T_James are popping up, using the same hw, already doing the hard dev work.

As shown, the powerstage module hardware and development sw work (and IP overhead) are the 2major $uckers for affordable AC drivetrain, before it gets massmanuf. which is 5-10yrs away, if you are an optimist. Now, you can get the first sorted out by sourcing 2nd hand donor auto parts for "pennies" thanks to toyota&hybdrid friends thus only the development part is in question. All tools here, but not much interest in it so far, strangly..

PS Not to be Semikron exclusive in these debates. There is also the Infineon "HEV kit" good for 80kW (peak) drivetrains. This dev kit is all incl. powerstage and driver&logic boards, DC caps, w. chill plate, sw GUI, etc., sans brake chopper module though. They supposedly sell the kit for ~4grand, you only add the emotor, encoders are usually inside 2nd hand h_emotors, brake chopper, wiring harness, so not much extra needed. It's kind of infuriating watching the pricedrop potential, should they massproduce these damn things, besides this Ifineon puppy is likely the most affordable ~100hp drive there is nowadays, apart from the WS, ..

Link to 96pages .pdf inside:
http://ecomodder.com/forum/showthread.p ... post190973
http://www.infineon.com/cms/de/product/ ... 7cfcb3708e
http://www.ehitex.de/p_info.php?product ... 63e9e9cd2b

--
100kVA PMSM controller for ~$2k, also inverters for ACIMs only double that (link from pedal section)?
http://www.lowcarbon-idea.com/products/ ... erter.html

To explore, Toyota HSD motors might be compatible with these low cost chinese controllers (support up to 120kW peak):
http://www.lowcarbon-idea.com/Controller.html
http://www.diyelectriccar.com/forums/sh ... 49438.html
http://www.diyelectriccar.com/forums/sh ... 48872.html

--
interesting info on HSD boost converters and more:
http://twinkle_toes_engineering.home.comcast.net/~twinkle_toes_engineering/hybrid_car.htm
--

!HSD/PMSM temp limitations at max power, basically go with bigger "SUV" motor:
http://www.insightcentral.net/forums/mo ... post174905
Last edited by Mesuge on Tue, 19 Oct 2010, 23:15, edited 1 time in total.
ImageCzech EV Club | blog, gallery, video | http://old.elektromobily.org

Post Reply