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

[cts_casemod]

Great news, Paul

When you say hard acceleration/deceleration, what exactly do you mean?

On mine, I limited the maximum allowable rate of change. The motor should take no less than 3 seconds from zero to 3500RPM (120Hz). This is actually much faster than what I have seen on operation. The reason being, if there is a fault, the actual encoder value will drift faster than the speed ramp allows and therefore I can measure this parameter as a fault output to switch the inverter off.

This is generally a better indicator for an encoder fault than waiting for the pulse rate to go to zero, specially if the fault is intermittent.

When you say field weakening, do you mean after it runs out of voltage or during low torque demand?

[cts_casemod]

Originally Posted by thingstodo

Between single-pedal driving, the width of the 'coast' throttle range, the max regen on the throttle pedal versus pushing on the brake ... there are a lot of options that people can experiment with.

How one would configure these options, keeping the various variables or setpoints straight in your head ... remains a challenge.

The GEVCU uses a webserver built into the ethernet interface to deal with the settings, showing them on a web page and allowing for graphical representation of the options they support. This is not as complex as you have already programmed, but it's somewhere that is available to review code ... and borrow code that is already running.

The GEVCU is Arduino based, so I don't think that the code will not be verbatim, but the web interface should be close.

Generally the issue is how to best ramp the motor to avoid commanding huge amounts of torque. During acceleration this is not generally an issue, but during regen you have all the friction losses aiding and therefore any subtle movement can command a LOT of torque. In addition, as the voltage increases the field also gets stronger and therefore, even more torque, however at a certain point the battery will not be as receptive, so either rheostatic braking is used in conjunction or the torque will go down to keep the DC-LINK within limits.

What gear one is using also affects the performance.

On a PM motor is very easy to relate current to torque and therefore only a few equations are needed, but on an IM this is not as linear.

[e*clipse]

Originally Posted by MPaulHolmes

I was clamping Vd and Vq using a scale factor. So, it was like this:

VdClamped = scale*Vd
VqClamped = scale*Vq

So that dist(VdClamped,VqClamped) <= MAX_RADIUS.

Id is proportional to Vd, and Iq is proportional to Vq, so why not clamp IdRef and IqRef with the same scale factor?! And then let them ramp up quickly to where they wanted to be. Each time Vd and Vq bump into the ceiling, Id and Iq now can still track IdRef and IqRef.

It works REALLY REALLY well. Hard accelerations, hard decelerations, oh ya, no problems whatsoever. Field weakening happens automatically. No messing around with the actual RPM. I think I have a winner winner chicken dinner.

Sounds like the "motor men" are getting good info and squabbling less.

BTW, since I will be attempting to make a BPMBLDC motor work with this program, I'll need to "comment out" the code about slip, but still adjust the Id Iq angle. Is this part relatively easy to find in the code?

Thanks a bunch!
- E*clipse

[MPaulHolmes]

By hard acceleration I just mean a large step in command of torque doesn't cause any problem anymore. The motor is only rated for 10 amp per phase while Wye connected, and Id and Iq hit their commanded current in around a millisec, and hold it until the RPM is so big that Vd and Vq are no longer possible given the bus voltage, so then IdRef and IqRef are reduced and updated at 10khz so that Vd and Vq continue to bump into their max. I'll definitely build a ramp between the torque command and throttle position. I'm just glad it isn't needed any longer to allow the motor to run properly.

That change will be good for the permanent magnet controller too. The main difference is the "ComputeRotorFluxAngle" function. In the PMSM case, you just read the angle! haha. in the acim case you do an unholy pile of crap. But both motors will run out of voltage, and have Id chase Idref (but IdRef will be zero in PMSM), and IqRef will sometimes have to reduce to allow Iq to converge to IqRef. hmm. and to increase speed, they say do negative Id to fight against the magnets, but that will only be an option when dist(Vd,Vq) < MAX_RADIUS.

I'm going to clean up the code now, and change some things to functions. It will slow it down a bit, but make it more readable. I also changed it to 15MHz, since there is still plenty of time for everything.

Here is my son twisting throttle both directions. this is with 10amp max for Iq and Id, at 72v dc on 480vac motor:son tests out ac controller: http://youtu.be/1Yr-dkuhzxg

[Cyruscosmo]

Hello Paul

Ok I will send you a pulley that I think is about the diameter of that motor. If it is larger then you will have to mount the motor with the output shaft hung over the edge of the desk or up on a wood block that is secured to the desk.

Keep in mind that with the extra inertia the motor will want to flip over when you hit the regen hard so secure it to the desk.

When you talk about the motor shafts you say back side and front side. Are you talking about the cooling fan side being 1.125 and the output side having been turned down to 1 inch so that the encoder will fit?

If so then I will send you an H bushing and key for the turned down shaft at 1 inch.

Shoot me a PM with address and I will get this on a UPS headed your way.

Cyruscosmo

[Cyruscosmo]

Pardon my silly question but if the throttle is set up for regen when below a certain point is there not the possibility that the car could go backwards?

I would hate to let off the throttle at a stop light and find that I have suddenly met the person behind me by accident...

Cyruscosmo

[MPaulHolmes]

exactly. the fan side is 1.125" diameter. I wad just going to cut the fan off if necessary. its plastic and could easily be grinded off. the front side sounds good to me. I have it bolted to the desk. thst sucker would have been launched off the desk during the locked rotor test. this is hugely helpful to me. I really appreciate it.
oh heck my address
paul holmes
54825 w hart place
Maricopa, AZ 85139

[MPaulHolmes]

yes! the car will go backwards if you dont make sure that as soon as rpm is below some value, that you set regen to ZERO!!!!

But, you would also hit someone in front of you if theottle didnt respond, so its all good.

[Cyruscosmo]

Hey Paul

Sorry for the delay, we had a wind storm here last weekend and I had a bit of roof damage that I have finally finished fixing. I got the pulley in to UPS last night.

The tracking number is 1ZA1129E0329359226

In the box you will find a large pulley, split taper bushing, a key and some bolts. Two of the bolts and the lock washers are used to tighten the bushing, the other two are extra just in case.

The key is a standard size key for that application. I know you had a machine shop turn down that shaft so if they did it correctly they also re-cut the key slot. If the key is too thick just use s bench grinder to remove a bit from one side. Since you are not using this motor for continuous service the key may not be needed but ya have it just in case.

Here is a link with installation instructions if ya need it. http://file.seekpart.com/keywordpdf/...2210131315.pdf

Cyruscosmo

[MPaulHolmes]

Thank you Cyruscosmo! I can't tell you how much I appreciate that. I would have hated to have to mail out the controller to the beta testers, who are going to test things at hundreds of amps, having not really tested it on a loaded motor.