Constant torque in base range

[HandleBag233]

Hi everybody.

I have a small 2kW permanently excited synchronous machine fed by an inverter. It is used for lectures at a university mostly and I am doing some semestral work on it.

Currently, I only operate the machine in base range. I use a field oriented control approach, so that I set constant values of the Direct and Quadrature currents and use the inverter to control the machine. The code works fine so far (i.e. control error converges to zero) and I get good torque responses using the torque measurement.

But it's not perfect and that's what keeps me wondering. I select constant D and Q currents and vary the speed of the machine. In base range, I would expect the torque to be on average constant with respect to speed.

But I can see a reproducable pattern, independant of variation of the machine speed.

The pattern is the following: Let's say I start at 100rpm and set the machine to some D-Q-current values. It gets to 5Nm. Now I change the machine speed to 1000rpm and I only get 4,6 Nm for the same D,Q-setpoint. Now I change the machine speed to 2000rpm and I get 5Nm again. I get this bow shape all the time.

I don't understand this. Is that a known effect? Is this an issue of the encoder angle or of the current measurements or some losses (in the machine or in the inverter)? Is there literature on this?

All the people at my department are mechanical or control engineers, so no one has much clue about the inner workings of electrical machines or inverters

How can I tacke this and how would you advise me to move forward?

Help would be very much appreciated, thank you in advance!

-FJ

 

[mikekilroy]

What exactly does "permanently excited" mean? PM magnets? Are they buried in the rotor to allow reluctance torque control?

Are there any slip rings/brushes in this motor?

How are you changing speed? changing voltage to field (is there one?)? to rotor? Both Other?

How are you measuring the 5nm torque?

How much effect do you expect torque angle advance to have? Is there are? Is it based on velocity Torque? both?

 

[mikekilroy]

what is base speed?

are you exceeding the torque rating? what is torque output rating?

 

[HandleBag233]

Hi Mike,

thanks for your answer.
I mean a machine with a permanent magnet, buried and I am crating reluctance torque. There are no slip rings or brushes on this machine.

The stator has three phases, the motor has 8 pole pairs. The stator has three phases.

I have it connected to a much better machine (asynchronous), which runs a speed controller. I measure the torque using the a shaft that is connecting both machines.

I am not sure about rated torque and speed, as it is a little university made machine without any real data sets or documentation. I have made some measurements and I found out that I need to start ramping up the field weakening d current at about 3000 rpm (depending on the torque). From the power, I guess I can do about 6.5Nm, but I was not able to get more than 5.5Nm out of the machine.

The angle question I do not quite understand. The encoder gives me the position of the rotor, which I use to create my rotation frame, in which I control the currents. Could you explain what you mean?

Thanks a lot in advance!

 

[mikekilroy]

have you tried the test with no d current? Is that possible in your setup? just use the pm field alone? does the 10% unlinearity go away?

my question on voltage supplied seems to be answered by your controller making it whatever it needs to be in order to maintain the currents that you are controlling, so you do not seem to have any info on what it is doing; can it be tracked to see if it has the 10% unlinearity perhaps due to some issue with the current controllers algorithms?

torque angle control may be similar to what you are doing with d current control? It seems you are INCREASING d current out of phase with the pm field in order to reduce it? sorry, I have no experience in this kind of field weakening. I believe, but cannot say for sure, that torque angle control does similar thing without the need of adding more current and heat to the motor? anyway google it to find lots of papers on it; I have not read these so do not know if they help understanding or not:

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&cad=rja&ved=0CFAQFjAJ&url=http://

http://www.sciencedirect.com/scienc...Yw88XjnJ4DzQPnqdQ&sig2=JKx5zBV5VkDL8Vps_SIbWA

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCwQFjAA&url=http://

http://www-3.unipv.it/electric/cad/...0ExrskE2sBgvVEYCg&sig2=zhRWFLXTCzfVaFcwPOK6Hw

 

[mikekilroy]

here is a doc that states what I was trying to allude to regarding the 2 different methods of reducing the effective motor Kt:

http://wiki-kollmorgen.eu/wiki/tiki-index.php?page=Phase+lead

 

[HandleBag233]

Thanks for all the information. That should keep me busy for some days