Paul and Sabrina's Cheap 3 Phase Inverter (AC Controller) with Field Oriented Control

[JRoque]

Hi,

I looked at off the shelf AC drives too and found a couple that worked all the way down to 170V DC. Because they were only rated at 1 to 3 HP, I thought they would make a good driver for external IGBT modules (minus their own high power stage). I believe this is how Eric Tischer did his Passat conversion. BTW, off the shelf AC drives with a >50 HP rating are usually rather expensive and like to be fed 460V with under voltage lockouts at >380V

Drives can move a motor in a few ways. They can do voltage and frequency variations and hope the motor will follow the commands. This is called V/F and can have some issues in vehicle applications where lots of torque is needed at close to zero RPM. The drive might command an RPM move but the motor can't deliver it at the given current so it "slips"... but the drive has no way of knowing that. Your acceleration will be less than ideal for what your motor can really deliver.

The preferred way is vector control by which the drive measures the current/flux of each phase to determine if it needs to apply more or less voltage. There are some, like the Curtis drive, that also look at the actual motor shaft to determine how fast it's moving. This last one adds complexity but it's the most accurate method.

For some time now, I've been looking at industrial motors like you suggest. I've found very nice motors that are vector rated with excellent torque curves and relatively low voltage (220V). But what gets me every time is the motor weight. A typical 50HP motor will weigh >500 lbs. I don't want to replace my 250 lbs ICE with a 500 lbs electric motor. If anything, I was hoping to shed some poundage with the swap.

JR

[mrbigh]

Originally Posted by JRoque

Hi,
But what gets me every time is the motor weight. A typical 50HP motor will weigh >500 lbs. I don't want to replace my 250 lbs ICE with a 500 lbs electric motor. If anything, I was hoping to shed some poundage with the swap.

JR

It's not only the motor weight, have you forgot the batteries total weight in addition to?

[TheSGC]

I think something that is overlooked is that the components to create a 400+volt inverter is in the thousands of dollars compared to a "lower" voltage of up to 300 battery volts.

I did test the driver again with a few adjustments I am am pretty satisfied with the driver. I am going to protoboard a good chunk of my logic now and add the two drivers to it. Since I don't have any DC/DC converters I will just use a 9v battery for my logic and then DeWalt batteries for the drivers.

I can probably protoboard it in an hour, but the coding will take a while as I have never coded C/C++ for dsPIC.

I'll try to get the driver schematics up tonight, probably as a scanned image of my notes.

[JRoque]

It's not only the motor weight, have you forgot the batteries total weight in addition to?

Hi. I was referring to typical industrial motors weight, not the overall conversion. My point is that, while industrial motors have appealing specs otherwise, their weight is usually a prohibitive factor.

the components to create a 400+volt inverter is in the thousands of dollars

Agreed. That's a totally new class of components up there that, by not being as common, tend to be more expensive. The size of the battery pack is also going to be an issue, as mentioned above, due to weight and volume, not to mention cost. It's no secret that higher voltage yields better efficiency but overall system cost and practicality should also weigh in the equation.

JR

[ecorsa]

I would argue further the case for higher voltages by saying that more batteries of less AH capacity won't cost much different from fewer of a larger capacity and will weigh about the same for the same total energy storage (comparing like for like).

Also it's difficult to say for sure whether the component cost is higher for high voltage vs high current without looking at specific devices.

I will agree that industrial motors are far from ideal for the application, but they are certainly good value for money, and easy to get hold of.

What I have been looking for is a suitable low cost controller to allow the use of high voltages, this does not exist as yet.
EV conversion will never become mainsteam untill it can create cars with decent performance and range cheaply. I think high voltages are the answer.

[JRoque]

Good points, and I just did a cursory check on battery sizes and they seem to also get smaller at lower currents. It does add to losses, etc due to the increased number of connections but that should be offset by system efficiency gains.

JR

[TheSGC]

Originally Posted by ecorsa

I would argue further the case for higher voltages by saying that more batteries of less AH capacity won't cost much different from fewer of a larger capacity and will weigh about the same for the same total energy storage (comparing like for like).

Also it's difficult to say for sure whether the component cost is higher for high voltage vs high current without looking at specific devices.

I will agree that industrial motors are far from ideal for the application, but they are certainly good value for money, and easy to get hold of.

What I have been looking for is a suitable low cost controller to allow the use of high voltages, this does not exist as yet.
EV conversion will never become mainsteam untill it can create cars with decent performance and range cheaply. I think high voltages are the answer.

What are you considering as "High Voltage"? IGBTs and Diodes are all about the same price since they are 600-1200 volts anyways, it's the capacitors and the sensors that will drive the price up. A very "High Voltage" production EV caps off at about 330 volts, and that isn't too bad. Probably less than $800 worth of parts for the 3 phase inverter. Most of the DIY high voltage EV fall into 156-288 volts using Solectria/Azure/Baldor motors.

[ecorsa]

I think most people will class anything over 200V as high voltage, the industry will class anything below 690V as low voltage so its a bit confusing. the point is to allow the use of standard motors which are cheaper than specials. So 208/240 V should be good for a lot of applications. I would prefer to go with 480V since you can get more power throught the equipment.
I thought about a simple analog control for a cheap off the shelf v/f controller, using a circuit to combine the rotor RPM with the throttle position to move the drive's setpoint around.
for example you could use closed throttle to provide 1% lag (drive setpoint 1% lower than actual rotor RPM, one third throttle could set the setpoint to equal rotor RPM, then full throttle couid move the setpoint to 3% higher RPM than rotor RPM to give full torque. I'm not sure if i am missing something important about how the drives work, but it seems to me this would work.
You could even set the setpoint progressivley slower using the brake pedal before the actusl brakes start to work.

[JRoque]

closed throttle to provide 1% lag (drive setpoint 1% lower than actual rotor RPM, one third throttle could set the setpoint to equal rotor RPM, then full throttle couid move the setpoint to 3% higher RPM than rotor RPM to give full torque.

Sort of like an exponential acceleration curve? That's interesting and it would approximate type ICE acceleration at certain speeds. One benefit of exponential acceleration is that you would get an extra push at top speeds that could help making a lane pass for example. I like it.

BTW, I keep going back to the weight factor of standard industrial motors. There are some real gems out there in terms of specs: 50 HP, vector rated, 220/460 V, 90% efficiency, $1500 shipped... but then weight is 700 lbs.

JR

[DawidvC]

If you are really interested in using industrial motors, JRoque, please have a look here -

aeva.asn.au/forums under electric motors and controllers you will find a post - changing an induction motor voltage ( wish I can post links - this software wont let me yet )

There are some other information scattered in their forums and in DIYElectricar regarding motor ratings.
The motor you mention above, will also be rated 460V, 103Hz 86.5HP without any modifications, just by running it connected for 220V at 460V and the right V/Hz ration. However, some of the Aussies will most probably rewire that monster for 200HP nominal ( about 460HP Max ) without violating any electrical laws. They typically use 11 - 22kw motors rewired for anything from 19kw to 88kw nominal, which would give you a max of at least 42kw to 220kw. The nominal rating of an industrial S1-rated ac-motor is virtually meaningless in an EV - it only tells you at what power the drive will have to limit you if you were pushing the car very hard. The real maximum power available is hard to define, as there are lots of ways to increase the power as seen in those posts.

I ope that helps a little bit
Dawid