[Guest]

Howdy,

I have in my possession a partially disassembled Prius transmission. Thanks to Fred and Marion in Redmond, Wasington for donating this equipment. It is my understanding that this transmission was damaged in an accident and Fred was allowed to keep the old unit when his new transmission was installed in his Prius.

I suspect this "autopsy" may take a long time to complete as I have a couple evenings a week for this project and I wish to capture as much information as I can.

I plan to post photographs and sketches of my findings. Since this message board allows attachments I will attach documents directly to my posts. If I need to post more than one file I will zip them together and then post.

I just unloaded the transmission and it is in my back yard. My first crude measurements were of the DC resistance of MG1 and MG2 coils. I have a medium accuracy handheld Fluke digital multimeter.

MG1 measurements: I measured the resistance between U-V, U-W, and V-W terminals. In each case my meter fluctuated between 0.1 ohm and 0.2 ohm. I need a better meter. Until I get a better meter I will assume the MG1 coil resistances somewhere around 0.15 ohm.

MG2 measurements: A similar situation. In each case I measured between 0.0 ohm and 0.1 ohm. Until I get a better meter I will assume the MG2 coil resistances somewhere around 0.05 ohm.

Page 33 of the Prius New Car features book gives MG2 maximum current as 351 Amps. If we assume current flows in one wire and out a second wire (this corresponds to trapazoidal control) we can use 0.05 ohm to estimate the approximate MG2 resistor heating in this scenario. P = I^2 * R = (351)^2 * (0.05) = 6.16 kW of copper heating. Using ohms law (home slaw) we can determine the voltage drop across the 0.05 ohm resistor. V = I * R. V = 351 * 0.05 = 18 volt drop due to wire resistance when current is 351 Amps.

If we assume current flows in one wire, half out a second wire, and half out the third wire (sinusoidal control) we end up with some kind of series/parallel resistor combination that makes the total resistance look 25% or 50% smaller. We either have half the resitance in parallel with half the path or the same resistance in parallel or some combination of the two (either WYE, DELTA, or a combination of the two). This would reduce wasted power by about 25% to 50% and reduce the voltage across the resistance by about 25% to 50%. This results in copper wire heating around about 4.6 kW to 3.1 kW at 351 Amps with a voltage drop of about 13 to 9 volts due to wire resistance.

Graham, a 25% to 50% reduction in copper heating losses for the same electric current is a compelling reason to consider sinusoids over trapazoids.

Good night.

Ed

 

[Guest]

Ed,
What a great project. I'll definitely be tuned in.
If you can get hold of a high power resistor of known value, maybe you can rig up a Wheatstone bridge to measure the winding resistance with the meter you've got. A few tens of amps from a car battery should do the trick. Or maybe pass the current of a headlamp bulb through the winding and measure first the current and then the voltage.
I recall that the Wye/Delta debate settled on a Wye configuration.
When you get inside, it will be interesting to see if the windings span individual stator poles, like my animated diagram, or if they overlap several poles like the eCycle motors. Of course, I'm also waiting to see how many rotor and stator poles there actually are.

[%sig%]

 

[Guest]

> -----Original Message-----
> From: Graham Davies
>
> What a great project. I'll definitely be tuned in.
> If you can get hold of a high power resistor of known value,
> maybe you can
> rig up a Wheatstone bridge to measure the winding resistance with
> the meter
> you've got. A few tens of amps from a car battery should do the
> trick.

You have given me an idea. All I need to do is inject a known current into the terminals and then measure the voltage across the terminals. The Wheatstone bridge is not necessary. I have a volt-meter that is very accurate down to a few millivolts. I have access to an amp-meter that is fused at 10 Amps. Knowing current and voltage my old buddy Ohm can tell me DC coil resistance. I just need a way to inject around 6 amps into the windings. Seems to me a couple of 1.5 Volt alkaline D-cells in parallel and a 0.25 ohm current limit resistor should do the trick. Looks like I'll be visiting my friends at Radio Shack.

Using both amp-meter and volt-meter I should be able to determine MG coil resistance accurate to 2 (maybe 3) significant figures using 99.3% accurate meters. If I can borrow some 99.95% accurate meters from work I might get 3 (maybe 4) significant figures.

> I recall that the Wye/Delta debate settled on a Wye configuration.
> When you get inside, it will be interesting to see if the windings span
> individual stator poles, like my animated diagram, or if they overlap
> several poles like the eCycle motors. Of course, I'm also waiting to see
> how many rotor and stator poles there actually are.

Looks like I'll be inside MG2 sooner than I had anticipated. I understand why Freds transmission was replaced rather than repaired. MG2 appears to be seized up ... I'm guessing a failed MG2 rotor bearing. It makes sense that it would be cheaper to buy a new transmission than to replace a MG2 rotor bearing. The labor required to access and replace a MG2 rotor bearing would be many hours and would surely cost a bunch of money at the typical dealer labor rate of $70 per hour.

Hmmm ... if I replace the MG2 rotor bearings and races while I'm tinkering I may have a good as new transmission when I'm finished. Unless of course I break something else.

Ed

PS. I have attached a photograph of the transmission (third line of this post). Note that the transmission is in two pieces. MG1 is in the smaller piece on the left side of the picture. MG2 is in the big piece under all the gears on the right side of the picture. Note that the planetary gears, silent chain, counter gears, final gears, and differential are all located inside this center part of the transmission sandwiched between MG1 and MG2. The differential pokes out just below MG1 and MG2 where the drivelines from each wheel connect. The whole unit is very compact. The transmission currently lives in a rubbermaid plastic container. My neighbor and I can both lift it so I would be really surprised if the whole unit weighs more than 200 pounds. I'll have to weigh it sometime.

 

[Guest]

Ahhhh, the wonders of technology!

I cant wait for more posts/info on our cars!

Does this mean Im a Techno-Geek?

Thanks for taking time to do this stuff, and sharing your project with us ..

[%sig%]

 

[Guest]

>Using both amp-meter and volt-meter I should be able to determine MG >coil resistance accurate to 2 (maybe 3) significant figures using 99.3% >accurate meters. If I can borrow some 99.95% accurate meters from work >I might get 3 (maybe 4) significant figures.

If you go to that much trouble, be sure to measure the temperature
as well. If it's copper it has 0.393 %/K temperature coefficient.

[%sig%]

 

[Guest]

Ed,

I have no doubt that you are going to be inundated with questions about your investigations, so I'll get mine in close to the top of the list

Can you see where the cooling water passages are? Is there any suggestion that the windings are immersed in water (something to look out for when you open MG2 up)?

Regards,
Julian.

 

[Guest]

> Can you see where the cooling water passages are?

When I get some time I'll be sure to identify the cooling water passages.

> Is there any suggestion that the windings are immersed in water
> (something to look out for when you open MG2 up)?

I can already answer this. MG2 windings are not immersed in water. There are two holes in the housing wall that separates MG2 from the gear box. If MG2 windings were immersed in water the water would leak through the holes and spill onto the gears.

I can look into those two small holes and see MG2 windings. I'm guessing maybe #10 or #12 wire. Since the motor is rated at 351 Amps there must be a bunch of coils connected in some parallel configuration. It will be interesting when I finally open it up. MG1 has no holes in its housing so all bets are off.

By the way, I have still not made my resistance measurments. Would you believe I'm having trouble finding the right size screw to fasten wires to the U, V, and W terminals.

I've got some other activities that will keep me busy for the next week. Hopefully I'll post something interesting next week.

Ed

 

[Guest]

> If you go to that much trouble, be sure to
> measure the temperature
> as well. If it's copper it has 0.393 %/K
> temperature coefficient.
>

I went to that much trouble. I borrowed two calibrated 6 1/2 digit multimeters from my employer. Thanks for telling me about the 0.393 %/K copper temperature coefficient. I performed the measurements during the evening as the ambient temperature dropped from about 85 F to 80 F and witnessed a 1% variation in my resistance measurements. This is in agreement with the copper resistance temperature coefficient.

MG2 measurements were performed with average ambient temperature of 83 F (28.5 C)

I made 6 measurements. The resistance between terminals U and V with W open. Between V and W with U open. Between W and U with V open. Between U and V with V connected to W. Between V and W with W connected to U. Between W and U with U connected to V.

MG2 winding resistance at 83 F

U-V: 0.0347 ohm
V-W: 0.0345 ohm
W-U: 0.0346 ohm
U-VW: 0.0260 ohm
V-WU: 0.0260 ohm
W-UV: 0.0260 ohm

MG1 measurements were made a little later and the average temperature during the measurements was about 81 F (27 C)

U-V: 0.0474 ohm
V-W: 0.0471 ohm
W-U: 0.0473 ohm
U-VW: 0.0355 ohm
V-WU: 0.0355 ohm
W-UV: 0.0353 ohm

I was surprised by just how small the resistance was. Even at 100 Amps the heating of the copper wires is not much more than a few 100 Watt light bulbs.

By the way. In order to access MG2 terminals I had to open up the MG2 housing. The attached photo (third line of this post) shows one side of MG2. From this I was able to see that MG2 wires are overlapped in a complex sequence that I have not yet unravelled (but I will sometime - even if I have to unwind the rascals). This will require removing the MG2 machine from the transmission housing. I'm not sure when I'll get time to do this but I look forward to it. I'm sure I'll learn about the MG2 cooling water path at that time.

Another intersting note was that the MG2 windings are covered with oily pink automatic transmission fluid. If you look at the photograph you will notice this pink fluid on both MG2 rotor and stator.

The good news is that I was able to tell that the MG2 stator (the outside part of the machine that remains stationary) is made of 48 electromagnets. Here is a drawing I have made.

Each one of the 48 points on the inside of the MG2 stator is essentially a piece of iron with wire wrapped around it forming an electromagnet. When I get some time I will enhance this gif and add detail showing the wires on the stator and the magnets and iron laminations in the rotor.

Later,

Ed

 

[Guest]

Ed,

This is phenomenol! Is the rotor for MG2 8 pole?

You had asked for the Max out of the NiMH. Last Saturday I recorded a 100ms draw of 98.4 Amps outgoing. The lowest voltage recorded while at max was 272V for 100ms. I was going about 5 mph entering onto a major 70 mph highway and needed to punch it to fall into oncoming traffic. It was late afternoon and 88 F. The max happened somewhere between 5-10 mph @ full throttle.

Regen on that same run was rather high also with 62.8 Amps for a minimum of 100ms with a voltage read of 335V. The highest regen amperage for 100ms thus far, is 68.3 Amps and highest voltage during regen for 100ms is 346 VDC. This 346 VDC regen max happened on a 61 F night.

The highest voltages don't seem to coincide with the highest Amperages on regen. It seems it has more to do with SOC.

So far I have made 18 runs and have an overall average as follows:

Avg Temp for all runs 80.1 F

Avg Standing NiMH voltage after runs 304.8 VDC

Avg NiMH Amperage under full load 83.2 Amps
Avg NiMH Voltage under full load 276.7 VDC

Avg NiMH Amperage under full regen 59 Amps
Avg NiMH Voltage under full Regen 335.7 VDC

Air conditioning was running on 50% of the runs and Lights on 21%

Thanks again for sharing the disassembly, photos, drawings and data!

Wayne
Pittsburg, TX

 

[Guest]

also if you want to include images that you uploaded to the files section in your posts the url for all the files are <http://www.priusonline.com/files/data/filename> that way you can upload a picture and use it inline in your post with the img tag without putting the image on other servers.

[%sig%]

 

[Guest]

Thanks Jeff,

I'm going to try again. Thanks for adding the capability to upload files to this web site.

Each one of the 48 points on the inside of the MG2 stator is essentially a piece of iron with wire wrapped around it forming an electromagnet. When I get some time I will enhance this gif and add detail showing the wires on the stator and the magnets and iron laminations in the rotor.

Wayne, thanks for posting your HV battery current and voltage measurements. This data has surprised me. You have reported both current and voltage measurements higher than I had expected.

To answer your question about the MG2 rotor having 8 poles. I don't know yet. Although from the photograph in my previous post I see slots in the rotor that may indicate the presence of 8 permanent magnets. So right now I'm thinking 16 poles. Eight magnetic (permanent magnet) poles and eight salient (steel lamination) poles. I'll know more when I pull MG2 out of its housing. Following this thread will take patience as I'm not sure when I'll get time to pull apart MG2.

Regards,

Ed