Hot-rodding the Toyota MGR

[freebeard]

Originally Posted by Cobb

The problem was it uses 650 volts vs the 288 volts that the prius uses and in order to get that voltage you need a lot of damn batteries or some type of voltage doubler or tripler.

(permalink) #3

That's the question that drove the creation of the thread: over-volting = hot-rodding.

If anyone has one in a running vehicle, I'd like to hear about it. Mine is on a shelf, while I educate myself about junkyard OEM chargers and inverters and DC-DC inverters and heaters and air conditioning pumps. Heated and cooled battery boxes. The list goes on and on.

What's your interest? This MGR is on the small side, probably suitable for a tricycle to Beetle sized car. If you haven't taken a plunge yet, by the end of this year your best drop-in solution for muscle cars or pickups or 50s land barges or motor homes will be the Tesla drivetrain.

Reasons? It's a similar configuration with a lateral MGR ahead of the axle line feeding a single speed drivetrain to a differential. Except the Telsa's inverter is integrated opposite the motor, and it is 3 feet wide and has 460-some horsepower. It can be controlled with CANbus messages except for the hardwired throttle and brake sensor. So P-R-N-D, creep and regen can all be dialed in with a $100 Arduino board.

In a few years it will be the go-to solution, like a small-block Chevy and Ford 9-inch rear end.

[pearl62]

I don't see the 650V input to the MGR as a problem, as I am thinking of using the Toyota inverter which already has the doubler. I am planning a max pack voltage of ~315V though, and I'm not sure if that will cause a problem. I can only see specs from Toyota for nominal input voltage of 288V.

I am thinking of adding this as rear wheel drive to my already EV converted FWD car. I could add a bit more power, regen and AWD, while also taking advantage of the built in DC-DC and possibly AC inverter. The MGR is also available very cheaply... much less than I expect a Tesla motor would sell for.

[e*clipse]

freebeard's right about the voltage.

I started this thread to brainstorm ways to get a bit more power out of it.

Generally, the RPM is limited by the rotor's strength and the BEMF generated at the high speeds. ORNL has tested a similar Prius motor to about 13000 rpm. Problem is boosting the voltage over the 650V, may just allow a higher rpm or top speed, but I have no idea what the rotor can handle. However, this won't boost the output power.

I'm very curious how much a lower battery voltage will limit the motor speed. I'm going to start with my Leaf battery pack, and hopefully learn something useful from that. ORNL did some tests at 500V and 250V for the Prius, perhaps that might help.

Total output power is directly tied to the current the stator can handle. The current creates heat and how the heat is dealt with becomes extremely important. To me, one of the most limiting features of the MGR is its cooling. Basically, it relies on the air blowing over the case to cool it. There is internal oil flow designed to reduce hot spots in the stator windings, but the heated oil has nowhere else to go.

I've been working on these issues with a Siamese twin motor setup. It's getting close; the weight is near that of two OE units. The new design uses all the 'Yota gears, etc, and will be prototyped on my mill. There is an internal gerotor pump to circulate oil to an external cooler and provide lubrication. The unit is completely symmetrical, which should make mounting easier. Also, the pump can be reversed so that the motor can be mounted in a behind the axle setup.

My evil plan is to use two of these siamese units, totaling 200kw+ in my eclipse. If the cooling system works properly, I'm sure I can get another 10% or 20% out of it.

Should this work out as I hope, then I'll start work on a Prius based 120kW unit.

[freebeard]

Maybe someone else (with some better idea) will notice the activity in the thread, but my understanding is that Toyota Synergy Drive has an early iteration that uses 288v, and a later iteration including the Lexus/Highlander and later Prius that uses 650v.

What's between the rear wheels of your EV now? I got the subframe out of the Lexus (separately) but it didn't include the cross-bar to the forward mounts. It's massive—40" wide and it weigh almost as much as the MGR itself. The cross-bar is more important, be sure to get that and the cable connectors.

[pearl62]

It's a 1996 Honda Civic, and it's pretty straight-forward to swap in the rear trailing arms and crossmember (or custom) from a CRV. Honda sold a 4WD wagon in Japan, and there are a few folks who have built their own 4WD Civic from CRV pieces.

If I can really reverse the mounting of the MGR it makes installation even easier than I thought. If I can get 50kW (67hp) then it would be worth it. However, I am confused about how much power is really possible from this motor. The table below from some service training describes 50kW output from the rear motor, yet the actual HP contribution the electric motors provide in the total for the vehicle is just 60hp, which is limited by the battery (as shown in the spec sheet I found).



If the limited power output is because of the battery, can the inverter and voltage doubler handle higher power? Maybe the confusion is from the fact that the Toyota inverter actually has 2 outputs 50kW and 20kW, presumably for front and rear.

I need to keep researching...

[e*clipse]

Boy, I understand the confusion...

Here's my best stab at it.

The highlander has both the MGR and a motor integrated into the main motor's transmission.

As you can see, the full cabability of these motors is not available at the same time.

My best guess is the pack cannot produce enough current to supply both motors . . .
and/or the controller's voltage booster is the limitation. When a system uses a booster to change the voltage from say, 288VDC to 650VDC, the full power used needs to be "processed" with the two IGBT's in the doubler circuit. Thus, they must be rated for, say 50kW at 288V. This means they must handle the full voltage - 650V AND the full current of the battery pack - 175A in this example. So, from the voltage doubler perspective, you've got the worst of both worlds - high current AND high voltage.

Now, about the MGR, that 50kW is intermittent. Better to plan on 25kW most of the time and 50kW for short bursts, like getting on the highway.

Also, in the stock mode, it is CRITICAL that it is mounted with the motor in front of the axle. Why? because oil is pumped by the ring gear into several oil gallies at the top of the MGR, then distributed to the gears and bearings and cooling for the stator. This won't work in reverse, so a backwards mounted MGR will overheat and wear out quickly.

HTH,

E*clipse

[pearl62]

Thanks again for the good info...

Didn't you mention that the oil pump can be reversed? Is there any negative impact to doing that? Is it hard to reverse the pump?

[e*clipse]

Originally Posted by pearl62

Thanks again for the good info...

Didn't you mention that the oil pump can be reversed? Is there any negative impact to doing that? Is it hard to reverse the pump?

You're welcome!

The pump can only be reversed in the new design I am currently working on. I'll post some pics soon. This new design includes a completely new pump and completely new case.

I suppose a way around the pump issue would be to purchase a reliable external pump - for example one designed for turbo oil scavenging - and carefully plumb it so the oil gallies are filled properly. BTW, Tilton engineering makes a good one.

- E*clipse

[pearl62]

You mean something like this?
http://www.summitracing.com/parts/ti...cIwhoCAaLw_wcB

It might be a good idea anyway as the MGR is in the rear and not getting much actual airflow for cooling...

Looking at the case, it seems that an external oil outlet could be placed near the bottom of the ring gear, feeding back in at the end of the motor shaft, does that seem right?

Are you really machining a custom case just for your own needs?

[e*clipse]

Originally Posted by pearl62

You mean something like this?
http://www.summitracing.com/parts/ti...cIwhoCAaLw_wcB

It might be a good idea anyway as the MGR is in the rear and not getting much actual airflow for cooling...

I'll have to look at the case.

Are you really machining a custom case just for your own needs?

Re pump - yes, that's the one. I've gotten good recommendations from many experienced folks for that one, including a good friend who does Baja 1000 support and Carroll Smith, author of several excellent books, including "Engineer to Win" and "Tune to Win." If the pump becomes a critical item - especially if it is the lube supply - it's obviously a poor place to save money...

I'll probably use that pump myself for a supplimental cooling oil pump. With the stock pump - and my new design - if the car isn't moving, no oil is being circulated. That's fine for the lubrication, but the motor might need more cooling.

Also, you're right about the cooling. Not only is the MGR getting restricted airflow, it's getting pre-heated (by the engine/radiator) airflow. d'OH!

The case I'm building is a proof of concept. If the concept works in my car - real world daily driver and track time - then it might be worth moving on with the concept.

- E*clipse