Space and field vector control

AC, DC, amps, volts and kilowatt. It's all discussed in here
Post Reply
User avatar
coulomb
Site Admin
Posts: 3885
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Space and field vector control

Post by coulomb »

Ok, time to show my ignorance Image .

Can some kind soul explain what id meant by the terms "space vector control" and "flux vector control"? Also "voltage vector control" it seems.

Edit: as seen recently in the Electric Beemer thread?

Edit: field -> flux; added voltage vector
Last edited by coulomb on Tue, 10 Mar 2009, 18:43, edited 1 time in total.

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Space and field vector control

Post by Richo »

They are they same.
They are the "buzz" word the sales people come up to try explain what the inverter does.
Rather than doing V/F they calculate/control V, F and phase.
The "space" comes from the modulations stlye.
Typical PWM has the square waves both start at the beginning and finish at different times based on duty cycle.
SVM - "Space vector modulation" have the square waves start at different times and end at different times where the middle of each square waves align in the centre.

Both cases use a theroetical mathmatical model to predict which phase should turn on and for how long to achieve the desired phase change and torque control.

So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Space and field vector control

Post by Richo »

It is more for those who what "Direct" torque control.
Even though specifically it is not direct but indirect.
The only way to get "Direct" torque control is to measure what the rotor is doing.
Which is too complicated/expensive to bother since maths can predict it.

For precise torque control an encoder should be used as this cuts down the maths to determine the speed and current position of the rotor.
This extra control allows the precise control of torque at or near 0 RPM.
Last edited by Richo on Tue, 10 Mar 2009, 19:12, edited 1 time in total.
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Space and field vector control

Post by Richo »

"Space" refers to the coordinate system used in the maths for the control loop.
"Flux" refers to the ability to directly control the level of torque in the PID in the control loop. (well the torque producing current -> Flux)
"Voltage" refers again to the coordinate system and is basically the same as the "space" term.
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Space and field vector control

Post by acmotor »

When you finish with the 3 letter acronyms, there are the 5 letter ones... SFAVM and 60°AVM

Stator flow oriented asynchronous vector modulation (space vector modulation)

60°Asynchronous vector modulation

my Danfoss does these as well and will actively switch between them for optimum performance. TJ, my unit does VVCplus SFAVM and 60°AVM and not flux control (the ultimate ?), that is the next model up.

But hey guys, I just say "that's nice".
The industrial world has been into all these goodies for decades.
Image

If you design/build a controller you will be re-inventing the wheel on one hand and EV customising a proven technology on the other.
With any luck, your EV version will also be cheaper !!
iMiEV MY12     110,230km in pure Electric and loving it !

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Space and field vector control

Post by Richo »

I was looking at some of the Industrial controllers yesterday/today.
There are some very nice 100kW units.
Voltage range is good and has torque control.
Petty they wouldn't fit in the engine bay of the BMW.
Weighed 61kg - which is more than the motor I want to put in.
And at USD$7000 they can keep it Image.

I don't praticularly like re-inventing but it's cheaper Image
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Space and field vector control

Post by acmotor »

Hey Richo, EV customised package built from scratch is what we need !
You are right, the industrial presentation is not EV friendly. I can vouch for that !
iMiEV MY12     110,230km in pure Electric and loving it !

antiscab
Senior Member
Posts: 2525
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Space and field vector control

Post by antiscab »

Richo wrote: I was looking at some of the Industrial controllers yesterday/today.
There are some very nice 100kW units.
Voltage range is good and has torque control.
Petty they wouldn't fit in the engine bay of the BMW.
Weighed 61kg - which is more than the motor I want to put in.
And at USD$7000 they can keep it Image.

I don't praticularly like re-inventing but it's cheaper Image


what controllers in particular were you looking at?

Matt
Matt
2017 Renault zoe - 25'000km
2007 vectrix - 156'000km
1998 prius - needs Batt
1999 Prius - needs batt
2000 prius - has 200 x headway 38120 cells

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Space and field vector control

Post by Tritium_James »

Sensorless flux vector control is the ultimate for induction motors, it means the motor controller is controlling the size and direction (vector) of the magnetic flux in the stator/rotor/airgap of the motor, which gives you full control over every property you could want. Torque being the usual target. With accurately measured motor parameters (inductance, resistance, etc) the mathematical model that the motor controller is using can get very close to reality, and this means there's no need for expensive and delicate encoders, tachos or torque sensors to measure the properties you're trying to control.
I was looking at some of the Industrial controllers yesterday/today.
There are some very nice 100kW units.
Voltage range is good and has torque control.
Petty they wouldn't fit in the engine bay of the BMW.
Weighed 61kg - which is more than the motor I want to put in.
And at USD$7000 they can keep it .
I don't praticularly like re-inventing but it's cheaper
And this is exactly why we're doing our new controller... However I wouldn't say the re-inventing is cheaper, even though the final product might be!

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Space and field vector control

Post by Richo »

I was at a US web site that sold nearly every controller under the sun.
Including some second hand / reburbished units.

I did have a good look at the cmg units as there are a lot of cmg resellers around.
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

fuzzy-hair-man
Groupie
Posts: 118
Joined: Wed, 12 Nov 2008, 16:40
Location: Canberra

Space and field vector control

Post by fuzzy-hair-man »

Tritium_James wrote: Sensorless flux vector control is the ultimate for induction motors, it means the motor controller is controlling the size and direction (vector) of the magnetic flux in the stator/rotor/airgap of the motor, which gives you full control over every property you could want. Torque being the usual target. With accurately measured motor parameters (inductance, resistance, etc) the mathematical model that the motor controller is using can get very close to reality, and this means there's no need for expensive and delicate encoders, tachos or torque sensors to measure the properties you're trying to control.
So it is mostly concerned with deriving a mathematical model of the motor it's connected to? once it has the appropriate model it doesn't need the motor parameters accurately measured anymore? I'm just remembering about Siemens motors and controllers and remember they were sold as a pair of motor and matched controller, it occurred to me they might be running the motor on a controller that could build the accurate mathematic model then loading the controller they wish to sell with the appropriate mathematical model for the motor and selling them as a pair, it would seem to give the benefits with a simpler controller, of course the downside would be when you want to change motors. Image

This is a stab in the dark though but I was a bit intrigued why they would match controller and motor when everyone else's controllers were happy with whatever was plugged in, Image Siemens seemed one of the leaders in the area so I didn't imagine it was for nothing.

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Space and field vector control

Post by Johny »

There do appear to be compromises in setting up to run ANY controller with ANY motor. For instance both Danfoss and Lenze can not auto-tune to a motor that is too small in power for the given controller rating.

You can still enter manual data to work with smaller motors and I don't know whether it's because the manufacturer thinks that the smaller currents are outside the sensing systems preferred range or it's just market(person) driven.

I'm running a 2.2kw on the Lenze 30kW controller without any trouble and on the low-speed loads I can give it on the bench it appears to work nicely.

I think that auto-tune functionality may have superseded Siemens "matched pairs". (IMO)

User avatar
Richo
Senior Member
Posts: 3737
Joined: Mon, 16 Jun 2008, 00:19
Real Name: Richard
Location: Perth, WA

Space and field vector control

Post by Richo »

The issue Jony is the resolution on measuring the voltage and current.
If the motor is too small it won't accurately pick up the values too geat for a good approximation of the model to run well.

In most cases this is not a problem beacuse people are unlikely to buy a really big controller to run a really small motor.
It's only a problem if you want to do some form of bench testing.

The vaules change over temp so matching controller with motor over temp may improve the response.
So the short answer is NO but the long answer is YES.
Help prevent road rage - get outta my way!

User avatar
Johny
Senior Member
Posts: 3729
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Space and field vector control

Post by Johny »

Hmmm. Not enough bits - that plagues my everyday world (DSP embedded).
I wondered if that was the issue. Probably mainly current - scaling the current sensor to the motor size gives the best resolution.
It still leaves the "matched pairs" as a question. Hype maybe?

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Space and field vector control

Post by Tritium_James »

The matched pairs is so they don't have all the support nightmares with customers not managing to get the auto-tuning thing working properly - they can just ship it out the door knowing the whole things works, and it's not going to cost their support guys a day on the phone later on!

Yes, the issue with small motors on the big controller is not enough bits, usually in the current sensing. You need to be taking fairly fast measurements, and 10 to 12 bits is usually about all that you get in most motor control type DSPs.

Basically, the more accurate your measurement results, the more accurate the model, and the less chance of the controller algorithms getting out of shape and regulating currents in weird ways.

As more processing grunt becomes available, I'm sure we'll see real-time tracking of motor parameters, so the control algorithms can dynamically update the measured parameters on the fly. The next step is actively checking for failures, doing things like harmonic analysis, looking for early indications of bearing failures reflected back via the motor current, that sort of thing. It's all pointing towards even less maintenance for EVs!

User avatar
acmotor
Senior Member
Posts: 3603
Joined: Thu, 26 Apr 2007, 03:30
Real Name: Tuarn
Location: Perth,Australia

Space and field vector control

Post by acmotor »

I noted the new FC series Dans (and probably other brands) include such features as pump run dry, drive belt breakage and bearing friction analysis.... and the one I really like ... the fire mode ! Basically the controller will run at full (nominated) speed and ignore all errors including phase losses, mains droop and motor faults and continue whatever even if the controller ends up cooked. This is for the new building fire standards that require air to be pumped to save lives (or fuel the fire).
edit: I should note that this is aimed at drives used in airconditioning systems.

There is something that appeals to my sense of humour in this fire mode.
A piece of electronics that will drive itself into the ground despite all its warnings, just because you told it to ! How brave ! Image
Last edited by acmotor on Wed, 11 Mar 2009, 19:06, edited 1 time in total.
iMiEV MY12     110,230km in pure Electric and loving it !

Tritium_James
Senior Member
Posts: 683
Joined: Wed, 04 Mar 2009, 17:15
Real Name: James Kennedy
Contact:

Space and field vector control

Post by Tritium_James »

That is quite cool. Probably not a good mode to accidentally trigger in your EV, though!

Post Reply