Homemade Motor Controller

[snakeplissken]

Just as a plan b I am thinking of playing around and just experimenting with the idea again of building a motor controller this time not out of a kit . What do you guys think of this high power IGBT as a candidate toshiba MG600Q1US41? http://www.datasheetcatalog.com/datasheets_pdf/M/G/6/0/MG600Q1US41.shtml or http://www.datasheetcatalog.org/datasheets/320/425969_DS.pdfI was thinking about two of them in parallel. What kind of a driver should I use? Would some kind of an Arduino be appropriate to use as a programmer? I am not sure because the datasheet says it takes 15 volts to open the gate and if I have two of them it really might not be enough. Also I would like the micro controller to be on board the controller permanently with the programming port coming out of probably of a plastic box. Anyone have any ideas about what sensor I should use to measure current to go back to the micro to the shut off in case of high current?

[Carlton]

your link didn't load( did you use the url bbc tag )

I've skimmed through your vehicle thread but still don't understand the full project.
What motor are you using?
what size?
what voltage?
what cluch?
what transmission?

I'm a fan of solid state relays we usea nice one from Sony a lot at work I think they can be triggered on 3volts and carry 8amps each
I don't know about the responce time but probably usable with arduino

[snakeplissken]

Here the links should load now. I am using a Warp 9 series dc motor I think its 9.25 inches in diameter I was running 168 volts no clutch direct drive and the original stock Ford five speed transmission. I think most of the controllers go up in the frequency range of 15 khertz could the relays be switched that fast?

[innovate1]

That part looks like it could work. Rather than use two in parallel you should see if there is a 600V version. No need for a 1200 V switch in a 170V system. The same size part in 600 V will be able to carry a lot more current.

As far as a driver you are going to need to supply MANY AMPS to the gate to switch it fast enough. Do you see the recommended gate resistor of 2 ohms? That means peak currents close to 15/2 = 7.5 A! (Actually higher than that with the positive and negative supply.) There are a number of companies that make MOSFET/IGBT driver chips that translate logic level signals to gate drive signals. At this power level I suggest you use isolated gate drivers. It looks like this part is rated to have positive and negative gate drive. A lot of gate drivers are positive voltage only. The company that makes the power module probably also makes gate drivers or could recommend something although for something this big they may expect you to build up your own driver.

For current sensors look at LEM type. The wire passes through the sensor with the sensor detecting the magnetic field. Lots of companies make them now. Over current protection is critical.

I also think you should look at a circuit topology that allows regenerative braking. For that you normally need more than one power switch. Not completely necessary but will give you more range.

Switching frequency is normally quite a bit lower than 15 kHz for this power level - probably something in the 5 - 8 kHz range. It somewhat depends on the inductance of the motor but slower switching is generally easier.

A fuse and a contactor to shut things down when needed.

For the controller you just need something that can generate variable PWM and can look at current signals. The main control loop shouldn't have to be too fast. Do the over current protection in logic outside the processor so it always works.

And look at some high ripple current, low ESR caps for the switching converter. These and how you lay out the connections are critical. You can generate a lot of voltage across a short wire when you switch large currents quickly.

I strongly suggest you build a smaller motor controller first and implement it the same way you plan to do the large one. You will learn a lot in the process. Better to fry some smaller/lower cost hardware first - and it WILL happen. And you may demagnetize the motor at the same time. Ask me how I know...

[Carlton]

excuse me it was a Sharp Microelectronics SSR S202

I don't think the responce time in kHz will work, but it is really cool what we can use this little thing for

http://datasheetz.com/data/Relays/Solid%20State/425-2403-5-datasheetz.html

[snakeplissken]

I can't seem to find a lower voltage higher amperage igbt on ebay. I guess the only reason for wanting to buy this part is the price. That would suck if it took 7.5 amps to open the gate I could be wrong but I think I saw where the where the gate might of needed a minimum of 5 volts at 600ma to open. What does it mean to have a positive and negative supply for a gate driver? How do I have the current sensor separate from the microcontroller? How big would the caps need to be in voltage and farad? Would the caps be wired across the battery + and - of the controller? I have a series motor there are no permanent magnet in it so I don't think that regen braking would be very easy but check out this link of a guy who did it and tell me what you think http://www.cafeelectric.com/curtis/regen/index.html.

[snakeplissken]

Check out this guys motor controller project its looks interesting he used four big Darlington transistors.
http://www.diyelectriccar.com/forums/showthread.php/blackpanther-sts-home-brew-dc-motor-32083.html On the fifth page of the thread he gives a schematic and parts list but never gets to how he current limits maybe it worked so well he just stuck with the two 555's to out put pwm.

[innovate1]

It doesn't need 7.5 A to keep it on. This is just the kind of peak current you need to make it switch quickly. Once you raise the gate voltage it needs no current to keep it raised - It looks like a capacitor. But if you have hundreds of amps and you switch it slowly there is very high power loss during the switching that can destroy the part. 5V may barely turn it on but it will not be usable at anywhere near the rated current at that gate voltage if it turns on at all.

For current sensors it isn't practical to use resistors for current sensors at this current level. For example, if you use a 1 milliohm resistor at 300A you will get 0.3V across it. .3V x 300A = 100 W. So the losses in the current sense resistor is 100 W. Better be a LARGE resistor! Resistors can be used for lower currents, under about 50A. Here is a sensor something like what you want:
http://cgi.ebay.com/LEM-LF306S-LF-306-S ... 4cee1b3e5e
There are links to the manufacturer so you can check out the details...

Sorry for missing the series motor. I usually deal with permanent magnet motors. The regen with series motor article looks reasonable. It looks like they go into some of the control issues. If you can separate the field and armature you might be able to control them separately.

[snakeplissken]

Could I use the 1000 amp 50mv shunt I have already for current protection somehow sending 50mv back to the micro? What parts would you recommend for say a 72 volt 200 amp continuous half bridge? Would I have to go lower voltage?

[Carlton]

I don't know why your in such a high voltage (you don't have to worry about line losses inside a vehicle)

what is the hp and Wattage range of the motor?

maybe we should start over in designing the electronics

[snakeplissken]

Well according to a few performance graphs I have seen the Warp 9 motor is probably about 32 horsepower at 72 volts but at a higher voltage its probably much more. About 1000 watts per hp. But I wasn't planning on using the smaller controller to run the Warp 9 I will just probably test that on a scooter motor or maybe the go kart. I picked a higher voltage so the vehicle would get more range.

[innovate1]

Yes, you could use the 50mV shunt - I hadn't read your other posts at the time and didn't know you had that. It won't give much signal to run directly into the micro so you should think about a signal amplifier. You will need a 4 wire or kelvin connection to the shunt anyway and the amplifier with differential inputs is a good way to do that.


In reference to Carlton's comment:
"I don't know why your in such a high voltage (you don't have to worry about line losses inside a vehicle) "

That's not such a crazy thing to worry about. You are dealing with lots of power. At lower voltages the currents are higher for the same power. Higher voltage lets you use smaller wire and/or have less voltage drop. Most of the hybrid vehicles have gone to even higher voltages.

Doubling the voltage will halve the current for the same power level. Given the same wire size this will reduce the voltage drop in the wires by half. And the voltage drop is now 1/4 of the percentage of the supply voltage. And you have dropped the losses in the wires to 1/4 as well.

This is one of the main reasons they are looking at going to 42V auto electric systems instead of 12V.

[snakeplissken]

How much esr is acceptable for the type of capacitor I am looking for? I was looking on digikey last night and sometimes they wouldn't have the resistance listed I wonder if that means they don't know or if it is just too insignificant to list? Do you know any websites that might have the type I am looking for?

[Carlton]

as far as I know capacitor types are only electric or ceramic

a perfect capacitor has infinite "resistance" (well in real power, for a DC car I hope your not talking imaginary power )

[snakeplissken]

I am no expert but I think esr refers to the capacitors efficiency and how much energy will be converted into heat within it. Here's is a quote from transwiki .

[edit] ESR Equivalent Series Resistance
The ESR rating of a capacitor is a rating of quality. A theoretically perfect capacitor would be lossless and have an ESR of zero. It would have no in-phase AC resistance. We live in the real world and all capacitors have some amount of ESR. To understand why let us review what a capacitor is and what they are made of and how we rate them.

I think what you mean is the integrity of the dielectric whether there is still a space between the cathode and anode and I am interested in an electrolytic capacitor.
Jim could probably provide a better explanation and more background on this.