Hello Gents and Gals

I have a processor board ...very fast very powerful

The intention is to develop FOC or DTC control.

I have cobbled a V/F control (no encoder) at the back end. The intention is FOC or DTC control.

Acquiring Stator resistance is straight forward without getting overly bogged down in some of the "thesis style methods".

The simplified approach I have taken for acquiring the stator resistance and Inductance is:

I have opted to use a sort of pulse test.

1. Pulse a delta connected motor on a single phase.
2. Measure the current.
3. Filter the measurement, (CT has some noise on it )
4. Get max DC current ...set as max current
5. Calculate Stator resistance......Rs = PulseVolatge/SteadyCurrent( often sited as a dc test )
5. Pulse again, when current is 0.63 of max current get time taken to get to that point
6. Calculate Ls = Timetaken/Resistance

It would great if somebody could see any merit, or not, to this approach or possibly site an article to this method. I would appreciate it.

Also, exactly what would I be measuring with the off pulse, "off pulse test", this would be more or less the above steps in reverse but what am I measuring then, I get a sneaky feeling that I would "see" some rotor parameters?

I have not found any information on the above approach, most approaches seem to revolve around the no-load and locked rotor test.

If I could get the motor parameters without turning the motor that would be great.

Or is the approach a bit fanciful?

Thanks
Carl

 

Hello Gents and Gals

I have a processor board ...very fast very powerful

The intention is to develop FOC or DTC control.

I have cobbled a V/F control (no encoder) at the back end. The intention is FOC or DTC control.

Acquiring Stator resistance is straight forward without getting overly bogged down in some of the "thesis style methods".

The simplified approach I have taken for acquiring the stator resistance and Inductance is:

I have opted to use a sort of pulse test.

1. Pulse a delta connected motor on a single phase.
2. Measure the current.
3. Filter the measurement, (CT has some noise on it )
4. Get max DC current ...set as max current
5. Calculate Stator resistance......Rs = PulseVolatge/SteadyCurrent( often sited as a dc test )
5. Pulse again, when current is 0.63 of max current get time taken to get to that point
6. Calculate Ls = Timetaken/Resistance

It would great if somebody could see any merit, or not, to this approach or possibly site an article to this method. I would appreciate it.

Also, exactly what would I be measuring with the off pulse, "off pulse test", this would be more or less the above steps in reverse but what am I measuring then, I get a sneaky feeling that I would "see" some rotor parameters?

I have not found any information on the above approach, most approaches seem to revolve around the no-load and locked rotor test.

If I could get the motor parameters without turning the motor that would be great.

Or is the approach a bit fanciful?

Thanks
Carl

Hello Carl,

That would be good to know about Motor type ( I mean is that single or double squirrel cage...)

As per explained on a article:
The single-cage model does not represent the squirrel-cage induction motors well, and therefore the double-cage model must be used. For the caged rotor of the induction motor, the double-cage model is better than the single-cage model. Fig. 2 shows the double-cage model used to simulate induction motors’ behaviors. The deep and narrow rotor bars have the torque-speed characteristics similar to those of a double-cage rotor.

Looks like rotor locked test is dedicated to Double squireel cage induction motor...

By the way I will have some research on that subject and come back when i will get more& precises informations...

I just have a quick look on an article , it is titled "
Squirrel-Cage Induction Motor Parameter Estimation
Using a Variable Frequency Test" and that method is used on a single squirrel cage motor..

Let see what it returns , then i will add some notes on these thread..

ControlsGuy25.

 

Hi ControlsGuy25

Hello Carl,

That would be good to know about Motor type ( I mean is that single or double squirrel cage...)

As per explained on a article:
The single-cage model does not represent the squirrel-cage induction motors well, and therefore the double-cage model must be used. For the caged rotor of the induction motor, the double-cage model is better than the single-cage model. Fig. 2 shows the double-cage model used to simulate induction motors’ behaviors. The deep and narrow rotor bars have the torque-speed characteristics similar to those of a double-cage rotor.

Looks like rotor locked test is dedicated to Double squireel cage induction motor...

By the way I will have some research on that subject and come back when i will get more& precises informations...

I just have a quick look on an article , it is titled "
Squirrel-Cage Induction Motor Parameter Estimation
Using a Variable Frequency Test" and that method is used on a single squirrel cage motor..

Let see what it returns , then i will add some notes on these thread..

ControlsGuy25.

Hello ControlsGuy25

thanks for that ,,,,I will have a look.

The test motor is a 0.12Kw 3 phase asynchronous motor. I don't believe that it will be a double cage.

It is a cheap motor( just for testing ). I am aware of how the motor characteristics are impacted by different cage construction ( NEMA )

The model I used is the typical transformer modal....so no funny business on the rotor side.

Although I do believe that the load/noload is sort of the defacto for getting these motor parameters. I would prefer not to use it as it needs power measurement and motor rotation and a locked rotor. I do know that there are some "tricks" in getting around the locked....but the accuracy of the test is ...I don't know?

Seems to be some presumptions on the test...so not sure if that's good enough for high-end motor control...

My preference is to get the parameters at standstill....that's why I opted for the above test as a starting point, it seems like a path forward, with my existing hardware....which is basically a motor control circuit, bus, CT, 30Kw stack, motor, processor...
no encoder.

Thanks for your help ...appreciate it.

 

Hi ControlsGuy25

Hello ControlsGuy25

thanks for that ,,,,I will have a look.

The test motor is a 0.12Kw 3 phase asynchronous motor. I don't believe that it will be a double cage.

It is a cheap motor( just for testing ). I am aware of how the motor characteristics are impacted by different cage construction ( NEMA )

The model I used is the typical transformer modal....so no funny business on the rotor side.

Although I do believe that the load/noload is sort of the defacto for getting these motor parameters. I would prefer not to use it as it needs power measurement and motor rotation and a locked rotor. I do know that there are some "tricks" in getting around the locked....but the accuracy of the test is ...I don't know?

Seems to be some presumptions on the test...so not sure if that's good enough for high-end motor control...

My preference is to get the parameters at standstill....that's why I opted for the above test as a starting point, it seems like a path forward, with my existing hardware....which is basically a motor control circuit, bus, CT, 30Kw stack, motor, processor...
no encoder.

Thanks for your help ...appreciate it.

Hello Carl,

Thanks for sharing that datas!

I would suggest you to have a read on the following document, it is talking /describing Unit standtill test .

By using SSFR method , here some notes from he document:

Most variable frequency tests are carried out by supplying only two phases
of the stator (namely, phase-to-phase tests) with a variable amplitude
and variable frequency sinusoidal voltage source when
the rotor speed is null, resulting in the so-called standstill frequency
response (SSFR) tests (see Fig. 1). These tests may use
a controlled voltage source with tunable frequency or the pulse
widthmodulation (PWM) voltage waveform supplied by a drive
inverter. As the torque in a phase-to-phase test is null, the SSFR
test can be conducted during maintenance periods, when impact
on plant operations is minimal. [3] proposes a phase-to-phase
test using thePWMinverter at two different frequencies to determine
the single-cage model parameters. In [4], an SSFR test for
derivation of single-cage model parameters is conducted and the
influence of skin and proximity effects on the rotor resistance is....

I Think that method can be studied and applicable /implemented for your Tests (Parameters determination/identification)

It remind me Engineering school cases studies , I liked and still like it!

I can share the designed document if needed..just send me PM as the size of the file is too big, to get attached here.

Hope this can help,
ControlsGuy25.

 

Hi ControlsGuy25
Thats sounds great,,,,here's a daft question what a PM?

 

Hi ControlsGuy25
Thats sounds great,,,,here's a daft question what a PM?

Hi Carl,
I mean Send invitation on Private Message (for PM)...you can do it by clicking on "start conversation" icone on the top of the forum page.

ControlsGuy25.

 

Hi ControlsGuy25
Thanks for that ....I can't find that

 

Hi ControlsGuy25
Thanks for that ....I can't find that
View attachment 683

Hi Carl,

You have to click on that "bubble icon " which is located top left close to the "Bell icon" , then click on " start conversation " , and search for my avatar....

ControlsGuy25.

 

Hi ControlsGuy25
Thanks for that it looks like I'm having some issues with that ...giving me some error message that I'm not allowed to do that
I have the name of the test so I can have a look around.
Did you give it any thought on my approach to the test ...I did do some experimentation and look like there might be some merit in it.
I measure 49ohms on the stator ...a single winding...this seems to be very high..I know that that stator resistance gets higher as the wattage of the motor reduces ...but 49 ohms seems to be high ...no mistake on my side with the measurement.
Thanks for the help I appreciate it.....if I can be be of assistance to you please give me a shout.

 

Hi ControlsGuy25
Thanks for that it looks like I'm having some issues with that ...giving me some error message that I'm not allowed to do that
I have the name of the test so I can have a look around.
Did you give it any thought on my approach to the test ...I did do some experimentation and look like there might be some merit in it.
I measure 49ohms on the stator ...a single winding...this seems to be very high..I know that that stator resistance gets higher as the wattage of the motor reduces ...but 49 ohms seems to be high ...no mistake on my side with the measurement.
Thanks for the help I appreciate it.....if I can be be of assistance to you please give me a shout.

Hi Carl,

Thanks for your post & comments , it is much appreciated !

If you share an email so we can exchange this interesting document....

I will have a better study on your experimental tests , then come back with some notes...

You right to say that 49ohms is bit high for kind Motor...

Did you apply the "pulse method testing" ( I mean The method that you described previously?)

Hope this help,
James.

 

Hi ControlsGuy25
Yes I applied the method got about 49 ohms, bearing in mind that the noise from the CT is awful ..so I stuck an exponentially weighted filter in there and some other things( there is a lot going on in the measurement ) I will put a screen shot of it below.
I also just measured it with a bench ohm tester.....it is 49 ohms.
have a look at cell S1...

You can see the exponential rise of the signal which is what you would expect from an inductor. There are various plots some with very crude filters but I believe I can get a good signal if I spend more time on the algorithm....maybe a better filter ..kalman which is basically a more sophisticated version of the exponentially weighted filter( I can hear the screams from the kalman filter guys on that sentence ).

I should have, and probably will, put a conditioning circuit on the CT to filter out the common mode noise and there is also a low frequency component.....anyway I don't want to bombard you with to much info ...it's better that you draw your own conclusions

But please pass comment ....you know what is like when you look at something to long you only see what you want to see

 

Hi ControlsGuy25
Yes I applied the method got about 49 ohms, bearing in mind that the noise from the CT is awful ..so I stuck an exponentially weighted filter in there and some other things( there is a lot going on in the measurement ) I will put a screen shot of it below.
I also just measured it with a bench ohm tester.....it is 49 ohms.
have a look at cell S1...

You can see the exponential rise of the signal which is what you would expect from an inductor. There are various plots some with very crude filters but I believe I can get a good signal if I spend more time on the algorithm....maybe a better filter ..kalman which is basically a more sophisticated version of the exponentially weighted filter( I can hear the screams from the kalman filter guys on that sentence ).

I should have, and probably will, put a conditioning circuit on the CT to filter out the common mode noise and there is also a low frequency component.....anyway I don't want to bombard you with to much info ...it's better that you draw your own conclusions

But please pass comment ....you know what is like when you look at something to long you only see what you want to see

Hi Carl,

I highly appreciate your feedback !Thanks for sharing such datas..

As I am busy on a power plant project, I strongly advise you to have a read on that excellent article:
https://www.researchgate.net/public...TER_DETERMINATION_FOR_15KW_AC_INDUCTION_MOTOR
Also I suggest you to read the document, that I told you about SSFR (standstill frequency response (SSFR) tests ),
Have a search on the web with following title:
squirrel cage Induction motor parameter using SSFR

We could also exchange email , in case you did not find it...

I will study closer your last datas and come back asap...

Hope this can help!
Stay safe&healthy
James

 

Hi ControlsGuy25
Yes I applied the method got about 49 ohms, bearing in mind that the noise from the CT is awful ..so I stuck an exponentially weighted filter in there and some other things( there is a lot going on in the measurement ) I will put a screen shot of it below.
I also just measured it with a bench ohm tester.....it is 49 ohms.
have a look at cell S1...

You can see the exponential rise of the signal which is what you would expect from an inductor. There are various plots some with very crude filters but I believe I can get a good signal if I spend more time on the algorithm....maybe a better filter ..kalman which is basically a more sophisticated version of the exponentially weighted filter( I can hear the screams from the kalman filter guys on that sentence ).

I should have, and probably will, put a conditioning circuit on the CT to filter out the common mode noise and there is also a low frequency component.....anyway I don't want to bombard you with to much info ...it's better that you draw your own conclusions

But please pass comment ....you know what is like when you look at something to long you only see what you want to see

Hi Carl,

I had a look on the trends that you shared here, I have following questions:

1-What parameters have you been capable to determine hence to the test....( Rotor /Stator Resistance.. by reading you first post it looks like it is Stator resistance right ALSO CALLED "Rs" ...
2-Which leakage inductance are you measuring .(It is shown on excel file , a value for L (is that Stator leakage inductance..)

Thanks for confirming some points....

James

 

Hi James
I believe that would be stator inductance. Ls
So the graph is current vs time... the time being measured in units of cycles( haven't calculated how long one cycle takes )
So the stator is essentially made up of leakage and magnetising inductance. Send me an email and I will make it clearer. You have a PM
I believe that Ls = Lls + Lm where Lls is the leakage.
Column A is essentially some time element
Column B is the filtered values of the stator current ...on one phase of the motor
Column C is another filtered(improved) value of the stator current...on the same phase as the motor.
Thanks
Carl

 

I thought I would wrap this up. I had a conversation with a guy who is doing some research on a new type of rotor construction.
He was kind enough to send me his thesis. Coincidentally the method he used was very similar to the one I mentioned. So I can conclude that there is something in it.
I have developed some software around this and now I have a view that I might be able to get all the motor parameters without rotating the motor....