Slip during motor starting 2

[Runsor]

Is it possible to approximate the amount of slip of an induction motor during starting without measuring the actual speed? Just the current and voltage measurements and name plate values are available.

 

[RRaghunath]

You can only distinguish whether the motor is over 90% rated speed or below.
This will be clear when we look at the motor speed vs current characteristic - the current remains fairly steady (at the rated starting current) upto almost 90% of the rated speed before it commences to fall.

 

[Runsor]

Doesn't the power factor give any more information? I thought it changes during starting: from from a low inductive value to somewhat towards one.

 

[Skogsgurra]

Yes, you are right.

Cos(phi) is the electric variable that changes the most with slip. It can be seen in the so called Ossanna diagram. It is possible to draw a curve showing cos(phi) = f(slip) from the Osanna diagram. One is shown in

http://www.ew.e-technik.tu-darmstadt.de/publics/EMD/OSSANNA-circle.pdf

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Most interesting! I do believe that is the most complex graphical solution of anything I have ever seen and I took a heavy duty Engineering Graphics course that was purely graphical solutions to complex problems.

Why is it called OSSANNA??

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

The guy was a Japanese. It also has other names, which I forget right now. Hayland is one of them, I think.

It is quite simple, actually. What you see is a closed curve, almost a circle. For qualitative purposes, a circle will do.

There are some basic rules how to find the center of the circle. Please note that the diagram has a horizontal axis for reactive currents (phi=90) and a vertical axis for active currents (phi=0).

The three red lines are Current at synch speed (the short one at 90 degrees), Current at nominal load (the middle-long one tuching the circle) and Locked rotor current (the long one). It is then relatively easy to draw a circle that goes through these three points.

This simple circle is a good representation of current and phase angle at different operating points.

The more elaborate diagram (with the cos(phi) line and points) can be used for rather accurate estimate of slip when phi is known.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[mls1]

If you can take a fourier transform of the current measurements you might be able to measure the rotor speed directly from slot harmonics. Slot harmonics are produced from the air gaps between rotor and stator iron cores. It's a common method for sensorless speed measurement.

 

[Skogsgurra]

Is it? Common? Where?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[mls1]

One need only type "slot harmonic speed sensor" into google and look at the hits. Is it common? Well that's subjective but I've used slot harmonics to measure rotor speed of a wind turbine induction generator and I was amazed at how easy and accurate it was.

 

[Skogsgurra]

Hello again,

I did a search and got 121 000 hits. But, after looking through the first 20, I hadn't found one commercial unit. All hits were either research papers, patent abstracts or discussions like the one we have here.

What make is your device? I do lots of speed measurements on motors and a working device would be very interesting.

For this purpose, measuring slip during start of a motor, would it be fast enough?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[mls1]

I used a Dewetron with National Instruments I/O cards. Standard clamp on CTs worked fine. Low speed resolution was not as good and may have been improved with tunneling CTs but I didn't try it.

Real time performance was good. I was using EasyPower measure with a nice FFT function. Off line I used Excels built in fourier transform in the Data Analysis add-in. Yes, real time I could track the slot frequency, but it was actual frequency and had to be divided by the number of slots to get rotor frequency. I didn't initially know the number of slots so I estimated it based on a separate proximaty sensor input but later confirmed it from generator data.

I stumbled on slot harmonics as I could see a high frequency component in the current waveform and wasn't sure what it was. I think you're correct that it is still researchy but the foundation is set and the analysis is simple.

 

[Skogsgurra]

That is a very good trace to follow. I do not even have to do a full FFT. I think that a simple box-car detector searcing the frequency range just below current fundamental*n, where n is number of slots, will work quite well. Finding "n" could be a challenge, though.

Why not FFT? Simply because I need to implement it in a processor with no HW multiply ability. I think that a box-car will be a lot simpler and faster than SW MAC. Also, it may be less error-prone.

Sorry, we got off-topic here. Anyhow, thanks for a good tip. Your first PLS!

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[RRaghunath]

Thanks to Skogsgurra and mls1 for the informative posts.

 

[Runsor]

Thank you SKOGSGURRA for a valuable information!
I found the Heyland-Ossanna circle diagram really informative.

ITSMOKED did ask where does the name Ossanna come from.
I googled something about Heyland and Ossanna.

Alexander Heinrich Heyland (1869-1943) was born in Iserlohn, Westfalen, Germany. While still studying at Hannover Technical University he published 1894 the ”Heyland circle diagram”.

Giuseppe Francesco Vigilio Maria Ossanna was born 1870 in South-Tyrol, Austria-Hungary, which is part of Italy today. He later used the name Johann Ossanna.
While working at Siemens & Halske in Vienna, he did publish the "Ossanna cirle diagram" 1899 ("Ossanna-Kreis" in german), evidently based on Alexander Heylands work.
A couple of years later he was selected the first professor for electrical machines at Munich Technical University. Prof. Johann Ossanna died in Munich 1952.

 

[Skogsgurra]

Thanks for that information. I always thought that Ossanna was of Japanese origin. Good to know that he actually worked for my former employer. I learned how to use that diagram at Siemens. Schools don't seem to care much about the circle diagram. Or has that changed lately?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...