Toyota MGR Rear Traction Transaxle

[celeron55]

610 must be MGR (driveshaft) rpm? If motor speed, that would be like 610/6.859 = 89 drive shaft rpm or 89 / 795 (rpm@60) = 7mph (for my build).

Yes it's motor speed. Toyota's torque curve begins to decrease at such a low rpm. It's quite a bit lower than what we see in Leaf for example. I wouldn't be worried about that though.

Given that it has a gear ratio of 6.859, it would make about 2.74 Nm / A at the axle.

If the motor makes 130Nm, axle (MGR output) would be 130 * 6.859 = 892 Nm @ axle, not 2.74.

Yes, as you can read I meant torque per motor current. When you multiply 2.74 Nm / A * 326 A (which might be Toyota's maximum current if my guestimate is right), you get 892 Nm.

I think of this totally differently...., but my math of 892Nm aligns with your suggested 1000Nm rating.

1) The Prius MG2 stage is rated for 50kW, a certain internet-famous Irishman tested and confirmed that is "conservative", if not very conservative.
2) The MGR (Q211) is equally rated at 50kW, if we assume that too is even the least conservative, you have a 75HP motor with integrated differential that weighs only 100#. WOW.
3) It's all about vehicle WEIGHT (err, Mass). The Toyota stuff is likely to be under-powered in a modern 4klb vehicle, and to be more power than almost any original air-cooled VW for same of lower finished mass.

So... it depends on what you are building, not if.

I agree, it seems really useful for a non-performance vehicle of any kind. You get to stop worrying about mass and space if you accept lower performance. It can even make conversions possible that weigh less than the ICE version but still have 100 to 200km of range.

[celeron55]

So happy news for the busy, lazy or poor people who don't yet own one of these. Because I now do!

This unit appears as if it was used as a boat anchor for some time. But it seems to be working and the electrical contacts seem fine. It even came with a reasonably serviceable high voltage connector.

I think the most interesting and easy test for this is: How much voltage does it output per rpm? I spun one side by hand at 30 rpm and kept the other side still, resulting in 15 axle rpm, which is about 102 rpm at the motor. The measured AC voltage between two phases was about 5 volts. That's rms voltage of course, so it'll take about 7 VDC at the inverter to do this. That means this motor does require almost 700 VDC at the inverter in order to produce good torque at 10000rpm.

EDIT: Also, I think I miscalculated earlier. What I need for 100km/h is about 790rpm at the wheels, which means this motor will have to spin at 5400rpm, which requires according to this test about 380V. So, with a full 96 cell pack, it'll maintain that speed for a little while and when the battery is almost empty at 300V or so the car might have slowed down, maybe down to 90km/h, who knows. How much does field weakening help here if the required power is about half of the maximum the motor is designed for?

[arber333]

I think you could go 25% into FW without too much loss of torque.
Do you have motor torque - power graph by chance? You have to know where the knee is to be sure. With Leaf at 360Vdc you are in FW range after 3000rpm already! I am not sure weather due to battery limits or motor torque drop due to FW.

[celeron55]

With Leaf at 360Vdc you are in FW range after 3000rpm already! I am not sure weather due to battery limits or motor torque drop due to FW.

The Leaf motor is certainly not in FW after 3000rpm. After 3000rpm the power has increased to 80kW and the inverter starts decreasing torque accordingly to limit power.

[arber333]

Hm... Which is the same as keeping it in constant power region which is what FW does to keep rpm increasing.
Where do you see FW with EM61 then?
I must say I can hear when FW happens in my ACIM motor in Mazda but not with Leaf motor.

[ZooKeeper]

"They" also said that the GS450H would "need" 750v, not apples-to-apples, but....

[rstevens81]

Based on the information in the post I am actually quite hopeful that this little gem could be used in a light weight car (looks like I shall be going shopping soon )

Based on the graph that Pajo_16 found and celeron55's spinning the diff, I have scientifically doctored out with paint (attached) and we can see that the knee of the curve is at I would say 4400 rpm, at that speed we would need:

Voltage req = 4400 x [7 / 102 rpm] = 301 volts which is quite close to the native battery voltage (288v)
Now if we say our system is 95% efficent (probably optimistic, but this will give us the smallest current)...
Current draw = 50000 / (0.95 x 317) = 185 amps
Based on the graph also our Nm/Amp shall be:
Trq Amp = 110/185 = 0.6 Nm/Amp

Now with a full battery accounting for sag etc say we kick out 390volts
We will get 390 x185 x 0.95 = 68500 Watts! So it looks hope full we can get full at least power in 96S config
Motor speed shall be:
MTR RPM. = 390 x (102/7) / 1.05 = 5683 rpm

Now assuming Field weakening can produce a similar constant power region:
based on "49 kW @ 4,610-5,120" (thanks pajo)
we will get maximum power between:
Pmax SPD = 5683/4400 x (4610 & 5120) = 5954 & 6613

Now if we take a smallish tyre say 195/50/15 (Mazda mx5) our tyre circumference shall be:
Tyre Cir = [195 x 2x(50/100) + (15x25.4)] x Pi = [195 + 381] x 3.14 = 1809 mm or 1.8 m
Our Maximum power speed shall be:
Speed = [1.8/1000] x 5594/6.8 * 60 = 89 kph
Speed = [1.8/1000] x 6613/6.8 * 60 = 105 kph
Pmax 49 x 5683/4400 = 63.3 kw

I know there is quite a large margin for error in this but I thought I would share and see what your thoughts were (if i'm off my rocker):

[celeron55]

scientifically

Ok that's the only word I needed to hear.

Now, do we actually know why Toyota even uses a buck/boost converter? Is it really only for speeds above 100km/h?

It's starting to look like even above 100km/h the converter is really not there because the electric motors would need it at the boost converter's rather low power rating. I'm now thinking when the ICE is putting out high power at high rpm, the MGs need to stay out of field weakening in order to handle the power, and that means they naturally operate at elevated voltage as MG1 generates the voltage. The buck/boost converter is there just to continue interfacing the MG rail to the battery in that situation.

[bexander]

I just bought a MGR and started doing some basic testing with a Honda IMA inverter and a "blue pill".
Does anyone know, and like to share, the motor parameters for this motor? Number of polepairs, Resolver polepairs etc.

[Mouse]

I just bought a MGR and started doing some basic testing with a Honda IMA inverter and a "blue pill".

Might be worth starting a new topic for Honda IMA inverter.

I had a few issues with the Bluepill simply because I didn't realise it needed inputs for currents and voltages before it would start and the error codes thrown up were not intuitive for me to work out what was wrong.

I'm quite interested in how this works out as I've considered the Honda IMA inverter for a motorcycle project.

[bexander]

Found the parameters for the MGR here:
viewtopic.php?f=14&t=488&p=11431&hilit= ... ers#p11481

[Martin1775]

Some remarks to the parameter list:

It is for the Prius Gen3 / Yaris Inverter with the Board ver.1b from Damien with his changed software (current sensor at phase1 and phase3), so you have to proof the phases because Johannes Software uses phase1 and phase2 and this don`t work right then in FOC mode.

There is a deadtime of zero witch will be fine for the Gen2 but not in the Gen3. For testing with load you should add some deadtime. Look at the discussion: viewtopic.php?f=14&t=817&start=80#p12653

Martin

[bexander]

Thats even better as that is the setup I plan to use, Yaris inverter with Damien's board. Right now just doing som testing.
Martin: What battery voltage are you using?

[Martin1775]

I'm currently still experimenting, and the MGR doesn't have enough torque for my possible project (VW T2 pickup `74).

I tested the MGR with 36V and 60V small lead acid battery on the Yaris inverter without load.

Plan is, I'll try with 120V to 150V to see if it is enough to use most of the Yaris Inverter as it is.

I still have an Auris hybrid gearbox with which I want to drive the original transmission so I don't need max rev. of the motor and it possibly works with the low voltage.

[tom3141]

Does anyone know the distance between the outside faces of the two drive flanges for the MGR?

Wondering about using one and would like to see how well it would fit and whether my driveshafts would need lengthening or shortening.

[ZooKeeper]

Does anyone know the distance between the outside faces of the two drive flanges for the MGR?

Wondering about using one and would like to see how well it would fit and whether my driveshafts would need lengthening or shortening.

I knew I had that written down somewhere.... Apx. 280mm (11" if you measure Yankee style).

[tom3141]

Perfect. Thanks!

[Issac]

Did anyone success with controlling that motor ?

Just to summarize after reading whole thread - it's DC-DC permament magnet brushles motor that RPM coresponds with voltage ? (why tho it has 3pins?)

[arber333]

Did anyone success with controlling that motor ?

Just to summarize after reading whole thread - it's DC-DC permament magnet brushles motor that RPM coresponds with voltage ? (why tho it has 3pins?)

It is 3phase motor still yes? It is not DC motor like brushed motors, we call it PMSM motor because it uses magnets and synchronus motor because slip is 0. It is still 3phase and takes similar voltage and control like your ACIM, it just does not induce current in its core and therefore has no slip. It has magnets to take care of the pull force.

I am sure it could be run directly with Lebowski drive. It could sport FW up to 45%. I had this control working up to 130km/h in my Leaf motor now. You can hear when motor changes pitch when it switches from full voltage mode to FW mode. (95% of DC voltage is reached).
What i still dont have is full resolver converter for it. Here Openinverter has clear advantage and simplicity of using original sensors.

[Martin1775]

I have the MGR running connected to an Yaris inverter with 60V, but so far only tested in idle, as I will probably switch to an Auris gearbox as drive. The max RPM is different for forward and reverse, but it stops with no trottle and can be controlled with the trottle signal. With the MG2 of the Auris it is the same, the MG1 runs in both directions at nearly the same speed.

There are a few points to note when connecting.

The connections on the Toyota motors are not always wired in the same way.

For example Toyota RX 450: MG1: 1=U, 2=V, 3=W; MG2 and MGR: 1=W, 2= V, 3=U.

It is best to compare the connection designations U, V, W from the documents, so that phase 1 of the inverter correspond to U, phase 2 = V, and phase 3 = W, then the connections to the resolver (for acquisition the rotor position) must also be correct(sin, cos, polarity).

If the connections in FOC mode are mixed up, it does not work or only works incorrectly.

[ZooKeeper]

Yup, I have a working MGR with a 2005 Prius (Gen2) inverter Granted, this was after much wire swapping of the inverter, resolver, phases etc

I would say this, if you are doing something that is uncommon, I HIGHLY recommend using an ammeter on the DC supply and keeping current <10A until you absolutely 100% confirm proper functionality. Unless you have Bussman/Cooper stock or just like plasma :O

A picture is worth 1000 words.

[Issac]

Guys, with what VDC in ?

Martin1775 - 60v is 1/5 of From factory was used with 288V

[konstantin8818]

Guys, with what VDC in ?

Martin1775 - 60v is 1/5 of From factory was used with 288V

I've tested it with 100V with gas pedal both directions - it runs smooth and fast in forward, spun it up to 8000 rpm without load no problem. Still need to tweak of syncofs for smooth bakward spin though.

[bexander]

I belive these connectors are used to mate to the resolver and temp sensor on the MGR.
Resolver: Toyota 90980-11034, Yazaki 7283-7062-40
Temp sensor: Toyota 90980-11143, Sumitomo TS090 6248-5317

I have a wiring loom from a Lexus that fits the MGR and the connectors are marked 11143 and 11034.

[ZooKeeper]

In my application, I will need 5454RPM @ 62mph/100kph, so nowhere near 400v, or even 300v

For those who are more concerned with application than theory...

160v will get you 4200 rpm all day long @ 8.5A no-load