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

[thingstodo]

This is sort of meant for Paul, but I thought others might chime in.

I'm looking for a project that I can do at the Lake on weekends. I have 2 of Paul's AC control boards collecting dust on my shelf (not the one that I may have killed by connecting the 48V pack backward). I'd like to do a brain transplant ...

EDIT - the idea is to load Paul's firmware for a DC to AC inverter, portable 120VAC power. I'd check with 24V or 36V to start, verify the firing is good, adjust the IGBT drivers (if required) or perhaps the capacitance? Damien McGuire has a video on adjusting adjusting 'stuff' on a DC controller to prevent ringing .. I want to see if that happens in the AC controllers as well ... then run a small 120V motor unloaded.

EDIT - so .. to make things transportable .. like in a couple of tool-box sized boxes ...

I have a ... pile .. of old industrial Variable Frequency Drives (VFDs) - sort of like AC Controllers that you feed with 3 phase power instead of a battery bank - that I got from work as salvage. Some of them had capacitor faults, some lost the Transient voltage protection, but most were replaced before the power and control had actually failed - the motors they drove were upsized, the VFDs did not communicate with the control system, the keypads stopped working, we blew an analog channel so we could not see what speed the drive was running at in the field ... and the list goes on. Lots of these have visible corrosion but operated when they were pulled. Most have been stored inside for a few years. 3 of them did not fit and were stored outside in the snow

A lot of the smaller VFDs are rated for 5 amps at 575VAC, capacitors for 1000 VDC. Those have a lot of all-in-one type of IGBT packs. I don't have a lot of info on them, but I could likely figure out how to drive them with Paul's control board. Paul's board being the new 'brain' and the existing capacitors, IGBTs, over-voltage protection, etc being the 'brawn'.

The larger VFDs, 10 amp, 20 amp. 50 amp, right up to 350 amp. The ones over 20 amps are single IGBTs per phase. One of the 350 amp VFDs uses transistor technology 2 generations older than IGBT.

I'm thinking of one of the all-in-one IGBT packs or the smallest of the separate IGBT per phase units.

Comments?

[wootwootman]

Originally Posted by MPaulHolmes

It's OK to err on the side of gentler convergence rather than most harsh convergence. Any undershoot means that it's too aggressive, and the Kp and Ki should be reduced some.

Okay, I think I'll scale it back down to kp=20,000 and ki=400. After that I'll move on to the angle-offset

Originally Posted by MPaulHolmes

I've only had to do it once too. Could there be a loose connection anywhere?

That's very possible. It's not with the control-level stuff, since I'm not getting any undervoltage warnings. I don't currently have a scope handy so I can't grab any shots of U, V, and W but it wouldn't surprise me. Attached is a pic of some plotted data I gathered (kp=20,000 ki=400), notice how the plots aren't entirely consistent from run to run. I typically discard the extreme outliers and use the rest

I'm using 2AWG cables with terminals that I crimped (not soldered) on both ends. One end goes to the IGBT through a current sensor, and the other is bolted to a short (~6") copper bar that acts as a connector terminal (prong?) to the motor, one for each phase. Maybe one of these connections is bad? I'll have to remember to upload pics later this evening.


Also, my contactor (LEV200A4NAF) takes a lot of current to hold closed, and it's eating up the Lipo I use for my 12V source. Anyone know any circuits for lowering their power consuption after they're switched on?

[MPaulHolmes]

They call them economizers I think. I've never made my own economizer before though. Make sure you reduce Kp and Ki by the same proportion that you increase the voltage. So, if your test was at 50v, and you increase the DC voltage to 150v, you would decrease Kp and Ki by a factor of 3 when running the controller at around 150v.

[wootwootman]

Originally Posted by MPaulHolmes

They call them economizers I think. I've never made my own economizer before though. Make sure you reduce Kp and Ki by the same proportion that you increase the voltage. So, if your test was at 50v, and you increase the DC voltage to 150v, you would decrease Kp and Ki by a factor of 3 when running the controller at around 150v.

Thanks, I knew they had a name! And yeah I'll have to remember to scale by 3 when changing the voltage, but for now I'm going to stick with 50V. I'm gonna double check my connections later and try to get to the point where I can so run-pi-test2 and get valid results every time.

[pablitosax]

About Nissan Leaf Motor (Unfortunately the video does not exist anymore): https://insideevs.com/nissan-leaf-100-torque-video/
But the comments about the loss of torque because a reverse EMF are interesting.
Torque falls a lot because of that reverse EMF ? Can you solve it by increasing the voltage?

[freebeard]

Back EMF is the flatlining of the torque/RPM curve. It is countered in the Toyotya HSD system with buried permanent magnets in the rotor.

http://ecomodder.com/forum/showthrea...tml#post443275

That thread split off from this one way back when.

[thingstodo]

Originally Posted by pablitosax

About Nissan Leaf Motor (Unfortunately the video does not exist anymore): https://insideevs.com/nissan-leaf-100-torque-video/
But the comments about the loss of torque because a reverse EMF are interesting.
Torque falls a lot because of that reverse EMF ? Can you solve it by increasing the voltage?

Paul would have a much more informed opinion on this one.

In my experience with induction motors, more voltage overcomes back emf, which gives you the current you are looking for, which generates your torque.

I expect that is similar for PMAC. The back emf is generated by magnets instead of induced currents in the stator.

[Piotrsko]

Yes you can improve it by more voltage, to physical limits. On my dc series brush arc increases until the arc goes all around the commutator until it hits the opposite pole then it becomes a zorch.

On an induction motor eventually you either overheat or experience dielectric breakdown.

Also: most power motors have physical rpm limits. Series dc has the lowest for size due to all that wire mass in the armature

[MPaulHolmes]

My understanding is, for motors with buried permanent magnets (high saliency they call it for some reason), you get some torque from the steel in the rotor, near the edges of the magnets, almost like an induction motor, and you get some torque from the buried magnets. To get the torque from both, you have to do a maximum torque per amp correction so that Id is negative (by some amount dictated by some hideously ugly formula based on several things including the "lambda" of the magnets, whatever that is). The correction isn't very hard to implement in the software though. Once you have the Id, then you can find how much of the rest of the current goes to Iq, based on the throttle command (you use the pythagorean theorem at that point to solve for Iq, since throttleCommand^2 = IdCommand^2 + IqCommand^2).

For a nonsalient motor (magnets on the surface) you have your throttle command, and you have to divvy it up between Id and Iq. Usually Id = 0, and Iq = throttleCommand. If you run out of voltage, and you don't like how fast the motor is spinning, you want to fight against the permanent magnets, you make Id negative, but then you waste some of the torque of the motor. But it's a tradeoff when you want the motor to spin faster (at the cost of lower torque).

[pablitosax]

This idea for a hub motor was apparently good because could be adapted to any rear drum brake (plug & play) , but , again the permanent magnets and reluctance have ruined the development:
"hub motors with permanent magnets are not the best technical solution " :

Green Energy Management (GEM) Updates | Middle Tennessee State University