Direct Drive: Do I have enough torque?

[Johny]

Unfortunately the best way you can parallel AC Induction motors is to lock the shafts together then treat them as one motor.

The next best thing is to run them in V/Hz mode that can allow up to around 5% slip (speed difference) - maybe. Running in V/Hz mode limits your starting torque though. AC controllers for EVs usually run in Vector or Field Oriented Control (FOC) mode which gives better efficiency and full starting torque.

Also, alas, a typical car's 10M turning circle is about 12-15% speed difference from inner to outer wheel - so it's not really viable.

[Canberra32]

[quote="Johny"] Unfortunately the best way you can parallel AC Induction motors is to lock the shafts together then treat them as one motor.

Or here is a thought
Few years back I worked on an formular SAE car that took the diff center out of a fwd car and put it in a simple housing turned up on the lath then bolted a drive sprocket to it.

Perhaps a similar simple housing like that with a motor both sides to do the balancing?
More the light mechanical solution?
So the diff center sits between the two motors to create the affect

It's actually dirt simple to make the housing on a lathe.
Just go to a wrecker and ask for any differential center out of a front wheel drive small car cheap solution
Ill even go so far as to say you get the center send it to me and I'll send it back with a cad file of the housing

[coulomb]

Coulomb, losing one motor in this case won't cause the car to go sideways on you during normal driving. eg: plenty of four-wheel solarcars only drive one wheel.

It seems to me that if one motor or controller failed in the most common situation, so the motor had zero torque (free-wheeled), then you would be fine.

But what if one motor mechanically locked up (e.g. bearing seized), or you lost CAN communications to one controller and kept running at the last commanded torque (granted, you could probably protect against that). I could envisage the vehicle being turned into oncoming traffic faster than the driver could react, but it looks like the chances of that happening are a lot lower than I thought. I guess if the danger can be adequately managed, then like other dangerous things (e.g. high voltage), it can be used safely.

Thanks for the reminder about the solar car case.

[ Edit: "the" -> "one". ]

[unheardofinstruments]

I have been toying with building my own large diameter axial flux pancake motor so i could bolt it directly to the cv and not need a reduction drive
the magnetic field rotates 7 times faster than the rotor as there are 6 pole pairs and 7 magnet sets this gives 7 times slower speed and almost 7 times more torque but different ratios are possible

axel borgs book is great

http://www.amazingdiyprojects.com/electric_motor.html

his build;

http://www.evalbum.com/3318

overclocking calculations here

http://www.aerodesign.de/peter/2001/LRK ... o_eng.html

as well as a great motion pic
could the controller be switched to overclock at a different ratio at speed with a custom/modified controller?

one wheel only needs to have a infinite ratio torque converter if it can go backwards too for a differential effect to work

The motor could also be the gear in a differential arrangement but it would require spinning contacts

[Canberra32]

Unheardofinstrament I have a site of a good motor designer if you want
If your going large dia than also makevitvlike the Gemini electric so it uses both sides of the stator

I'm having a larger version of a synergy smart drive designed
You have the same idea as me work using mechanical advantage to make every volt count

[Canberra32]

Hey back on the topic of run two motors down the drive shafts here is an idea

http://www.alibaba.com/product-gs/45427 ... Drive.html

Two of these
Can get them as hub attachments too

[unheardofinstruments]

Cool planetary drives want the other 8? (minimum order of ten, not sure how much $ the ones we might require would be) but a great find, ones I have found have all been quite expensive by comparison so far.

They can be salvaged from some automatic gearboxes, Craig from turquoise energy did just that but is 3d printing his mechanical infinite ratio torque converter, if it works I would be keen to have him print me one.
An equivalent nu vinci drive (much more limited ratios) is $1395 just for the bicycle strength one but they do bigger torque versions (poa- read expensive)
I also just noticed the newer prius has an auto gearbox that claims to be continuously variable, are they talking the same thing? I bet the designers would kick themselves if they realized the Constantinescue patent is long lapsed, or did they use it and rename it? Anyone know how it works?

I like the gemini motor thing but I would do the same with a pancake design and instead coils both sides of the rotor magnets plus it's great to see someone else using the back emf idea, I love it when you think of something and someone has already done it.

By larger synergy smart drive do you mean a larger diameter motor linked to an I.C.E and with regen buck converter? + Tell us the site for motor design then! I wanted to use a large hallbach array, haven't found one off the shelf and getting a magnet design built as a `small' run is $10,000.
Litz wire seems well worthwhile using too.
Let me know if you have a good source of chunky neodynium magnets in a sector shape and I might give up looking for the hallbach array

I wonder if a simple electric clutch (like an aircon pump magnetic clutch activated by switch) on the primary drive shaft (before reduction) could hack the torque and release when appropriate, (say when you turn more than 5% slip) pretty simple infinite slip diff.

Sounds to me like V/Hz mode might be workable most of the time except parking/u-turns (and maybe takeoff if it can't be switched on the fly from vector/FOC mode), it isn't often you use the outer 2/3rds of the steering, acceptable scrub at low speed/tight turns? Might do strange stuff in mud.

[Canberra32]

Am talking to my machinist about a two speed box with a differential it's a simple design but will be affective

[BigMouse]

I also just noticed the newer prius has an auto gearbox that claims to be continuously variable, are they talking the same thing? I bet the designers would kick themselves if they realized the Constantinescue patent is long lapsed, or did they use it and rename it? Anyone know how it works?

\

The prius transaxle isn't a CVT in the most common configuration. It has a fixed gear ratio on the final drive and a single planetary gear set between attached to two electric motors and the ICE. The "continuously variable" action comes from varying the ratio of speeds between the two inputs to the planetary set (the second, and larger electric motor is geared directly to the differential). There are a couple really good videos that explain it on youtube.

In short, the Prius CVT works by varying shaft speeds rather than geometry.

[unheardofinstruments]

A two speed box & differential sounds like a good drop in replacement for the 924 transaxle, I would be keen to hear how it goes, sounds great. The infinite ratio torque converter would allow a much weaker/smaller motor and it isn't any more complex than a diff/box to make by the looks of it, ask your machinist if he would give the constantinescue drive a look/see, the only tricky bit is the linear converters teeth that converts the wave motion into single direction drive. Here is a great paper analysing it, the url is epic but the simplicity of the system is obvious from the first diagram which is a basic sketch of his first working prototype

http://www.google.com.au/url?sa=t&rct=j ... XTJHp-oR6g

[Canberra32]

Complex = $$$$$
Simple means more affordable.

[Johny]

In short, the Prius CVT works by varying shaft speeds rather than geometry.

Here is a good site with an interactive that helps explain it.
http://eahart.com/prius/psd/