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

[e*clipse]

Originally Posted by MPaulHolmes

LOL. Haha.

The cheapest I have found is $145 for 1, and $100 each for 5. But I don't need 5.

Would you be willing to piece something together? Some years ago I built one by machining my own code wheel and mounting two optically interupted switches such that it worked as a quadrature encoder. It was soooper cheap - like $5 and a lot of time....

I'm not saying do that, but you can buy nice encoder wheels and seperately nice little quadrature sensors:

Qty 1 HP Q9898 Optical Encoder TTL Quadrature Output | eBay

Agilent Heds 6140 B10 Optical Encoder | eBay

Odds are it will be difficult to find the exact ID, but I'd be happy to either machine a spacer or open up the inside if it's close.

E*clipse

[e*clipse]

Originally Posted by Astro

852D+ Then you bought the same model as me. I think they updated it just after i bought it but didn't change the model number. They added a few refinements. Fan in the main unit is different to mine. Bet you even got a real power switch?

LOL! No, this one doesn't have a main power switch either. In some ways, that may be good because it has a "cool down" cycle for the hot air handle. I'll bet it doesn't have an off switch so that a person doesn't turn it off while it's still hot.

I was able to solder most of the board together yesterday, except for those resistors I managed to forget to order... It looks way better than the hand-soldered work on the opposite side. I just hope I didn't overheat the parts.

- E*clipse

[e*clipse]

The only thing I can see is that it might be a bit overkill, but that won't hurt anything.

There are also some other good possibilities, raised by this idea. For example - how about electric power steering?? I know for a lot of conversions, the whole power steering issue is a big power-hog PIA. Also, I'm looking at how to heat/cool my EV. The available options are pretty lousy, IMHO. There is no cooling option, and the heating option is expensive and totally not appropriate for a vehicle. The new technology is heat-pumps - yes, the EV1 had one, the new Leaf has one, and the Renault Zoe has one. I've been looking around, and found a pretty svelte one that uses a little BLDC motor for the compressor. - Hey! what about this controller??

Originally Posted by MPaulHolmes

Oh it's no intrusion! I think it would work fine. I'm not 100% sure I understand, but as long as you somehow tell the controller the amount of torque you want it to produce, it will do it, as long as it's hooked up to a motor.

[rapapar]

Originally Posted by MPaulHolmes

Oh it's no intrusion! I think it would work fine. I'm not 100% sure I understand, but as long as you somehow tell the controller the amount of torque you want it to produce, it will do it, as long as it's hooked up to a motor.

thanks for your answer, I'm going to learn your project in detail

[Astro]

Originally Posted by e*clipse

...Also, I'm looking at how to heat/cool my EV. The available options are pretty lousy, IMHO. There is no cooling option, and the heating option is expensive and totally not appropriate for a vehicle. The new technology is heat-pumps - yes, the EV1 had one, the new Leaf has one, and the Renault Zoe has one. I've been looking around, and found a pretty svelte one that uses a little BLDC motor for the compressor. - Hey! what about this controller??

I also am looking into eventually using a reverse cycle heat pump set up to heat and cool.
Initially i will use a ceramic heater element to replace the existing liquid heater core. That is a super cheap option. Ceramic heaters can be found for less than $20. Of course they are not as efficient as a heat pump but it will do the job initially.
And i will use the existing A/C compressor driven off of the non drive end of the traction motor. Again cheap and will get the job done.

I had thought of doing a heat pump with some sort of small electric motor to drive it.
But then i thought, why not just run the compressor off the traction motor?
As long as the traction motor has enough torque to spare then it means i save the weight of an additional motor for the compressor as well as the complexity and weight of a second controller. It is also very cheap as a standard ICE belt driven compressor can be used. Although i would like to get a nice efficent scroll compressor out of a modern split system set up.

The compressor wouldn't turn when the car was stationary but as long as it was just a brief pause like say, a set of traffic lights, the existing pressure in the system should keep it going for a little while and it will be re-pressurised as soon as the car takes off again.
Or if it was a longer stop than the car could be put in neutral and the and the motor brought up to a low RPM to operate the compressor. Much like an ICE vehicle idling.

Actually seeing as the compressor has an electrically activated clutch maybe it could be made to only activate when there was spare torque available. So not when the car was accelerating, only when it was cruising.

Actually another thought, for the guys that don't have a regen capable motor having the compressor engage on deceleration would be a way to recover some energy.

I am sure Paul's controller would be capable of running a small motor to drive the compressor but it may be a bit of overkill.

[cts_casemod]

While I doing some work this morning, It just occurred to me that one could replace the bipolar PWM switching algorithm with a Unipolar version.

Basically what this means is that instead of the battery Voltage producing Line to line voltage, one could produce twice the line voltage, using a unipolar switching PWM waveform



Notice how the PWM on the graph is referenced to GND rather than 1/2VCC

That would allow 415V motors to be operated from a 300V Battery with no conversion losses. Anyone interested in this?

[MPaulHolmes]

I would be interested!! I don't really understand though. so, you have a 300v dc input. You pwm that input to make it look like a pwm version of a sine wave. Then you pwm another one with the complement. Then you subtract the results? How do you subtract them?

[cts_casemod]

Originally Posted by MPaulHolmes

I would be interested!! I don't really understand though. so, you have a 300v dc input. You pwm that input to make it look like a pwm version of a sine wave. Then you pwm another one with the complement. Then you subtract the results? How do you subtract them?

The requirement is instead of having complementary PWM outputs for each timer, one needs to have a dedicated output, rather than a complementary mode as they now have.

By default thats not possible with the PIC, but perhaps with a bit of smart coding we could use the interrupts and a look up table to disable one of the channels, so that only the positive or the negative switch can be active at a given time.

For example, assuming an h bridge:
During the positive half cycle, PWM is sent to the high Igbt of leg 1 and the low IGBT of leg 2, so that the voltage is positive (DC-LINK)

During the negative half cycle PWM is sent to the low IGBT of leg 1 and the high IGBT of leg 2, so that the voltage is negative (-DC-LINK)

Here's an example for a single phase inverter using this topology
http://www.journal.au.edu/au_techno/...abstract05.pdf

[e*clipse]

If we could pull this off, it would be AWESOME!

Yes, I'm totally on board - not having to have 650V of batteries would really help my project.

I'll try to catch up a bit here, starting with your link, so maybe I can help a bit - but I'm definitely on board.

- E*clipse

Originally Posted by cts_casemod

The requirement is instead of having complementary PWM outputs for each timer, one needs to have a dedicated output, rather than a complementary mode as they now have.

By default thats not possible with the PIC, but perhaps with a bit of smart coding we could use the interrupts and a look up table to disable one of the channels, so that only the positive or the negative switch can be active at a given time.

For example, assuming an h bridge:
During the positive half cycle, PWM is sent to the high Igbt of leg 1 and the low IGBT of leg 2, so that the voltage is positive (DC-LINK)

During the negative half cycle PWM is sent to the low IGBT of leg 1 and the high IGBT of leg 2, so that the voltage is negative (-DC-LINK)

Here's an example for a single phase inverter using this topology
http://www.journal.au.edu/au_techno/...abstract05.pdf

[cts_casemod]

Cool.

So Far I am having a bit of trouble picturing the commutation order of the switches.

Guess its time to fire Matlab and run a simulation... Ill get back to this