I have a few questions. It would be great if you could help explain. The first question is if we apply a positive current in the D axis we could strengthen the rotor flux. Does that mean we can get more torque at low speeds? I have a motor and inverter in a I am working on the software. I have the constant torque working correctly. I am now trying to get the field weakening working. I have applied a negative current in the D axis but the amount of applied current to get a speed increase is very large. To get twice the speed I need to apply roughly the rated current of the motor leaving no available torque current. However the same motor used on an EV seems to be able to spin at double the rates speed and still have torque available. Is there anything I am missing?
Francisco, To your first question, in theory you could, but there are other factors that you have to consider. What is the flux density in the airgap before you strengthen the field, what are the loss components, etc... If you have a torque sensor, you can experiment for your choice of motor. One aspect to check is if your position sensor is properly aligned. If not, the negative d-axis current may not be of use. You can do a simple calculation based on motor parameters to determine how much current is needed to reach a certain speed and still have torque available. Also, make sure your parameters are correct. Can you share more information on this motor and where you see this being used? Are the two machines identical, design wise and construction wise?
@@SandunKuruppu The specs from the manufacturer are as follows: Ld is 0.19mH Lq is 0.44mH Back EMF 39.2v/1000rpm Peak torque 260Nm @380A Continuous torque 128Nm @160A The torque curve from the manufacturer shows peak torque to 3000rpm with 320Vdc. Some of the numbers do not add up. I have measured the motor back EMF to be 100v/1000RPM measuring the sine wave peak to peak. I have measured the torque (in a not very accurate way) and the torque constants are in the ballpark. The motor is from a Chinese EV.
@@haowang4306 Hao, I don't have a video on calculations as it depends on the motor design and parameters. I will try to prepare some results with calculations and experimental results.
i have one question: is lambda changable at all? as far as i recall, the d-q axis equation is derived from the three phase voltage equation with clark-park transform, and term with rotor flux even after transformation should still be only dependent on rotor magnet material, and not a "effective" "total" flux in the d-axis, right? in the previous lecture the detailed transformation was not provided, so it would be great if you could show us the abc to dq transform in detail, especially the flux linkage terms, thanks!
Hi Xiao, Lambda is called the flux linkage constant because flux has a significant contribution here. In the rotor reference frame, D-axis is where most of the flux is passing through. Therefore, introducing an opposing flux source can effectively reduce Lambda. Now, how much of a reduction in flux or would this demagnetize the magnets permanently, depend on the motor/magnets used. Does this answer your question? I will try to add another video showing the rotor reference frame model derivation soon.
@@SandunKuruppu hallo, Professor, thanks for the reply! my question was more about if we should see lambda as a constant that is only denpendent on the rotor and will not be affected by stator excitation situation? it is also linked to the torque equation, e.g. rotor flux "cross product" stator flux or Iq times rotor flux linkage. In field weakening, will the torque only be redueced/sacrificed because the rotor flux and stator flux are no longer perpendicular to each other (while the magnitude of both flux vector remain the same, since the dc-bus voltage and the permanent magner on the rotor does not change?) or the torque will reduce also with the side effect that the rotor flux magnitude will reduce due to Id in the opposite direction, in this case the torque will be even smaller?
@@xiaosun9111 Theoretically speaking, if you look at the motor model from the rotor reference frame, you lambda is a constant. Similar to Ke (back emf constant) in a DC motor. However, since this is related to magnetic saturation, practically this varies with the level of saturation. So Lambda varies a little and in a non-linear fashion as the air gap flux density changes. Since Lambda is part of the torque equation, once you weaken Lamda, your torque will also drop proportionally. The reason torque drops under proper FOC is because your effective rotor flux has been reduced, as long as you maintain orthogonality between the stator and rotor flux. However, if you do not maintain orthogonality (i.e. NOT properly field oriented), then again the torque would drop. This is not due to Lambda but due to non optimal placement of stator flux. Is this clear to you? I think you are asking about two aspects that will affect torque. Lambda is a separate issue and orthogonality of the vectors are another issue.
@@SandunKuruppu thanks! i used to think that the stator flux was generated by stator currents alone, after transformation, Id and Iq are responsible and "add up" vectorwise to the "rotating field" (of course stationary in rotor reference frame), given the voltage is limited to DC bus voltage, its magnitude (sqrt(Id^2+Iq^2)) is limited! so if a negative Id is applied, (previously the Id was 0, so all voltage can devote to Iq, and the "sum" is 90 degree to rotor) under the voltage restriction, the Iq will automatically be reduced, thus as "vector sum" of stator field will be no longer perpendicular to rotor flux😂
@@xiaosun9111 Stator flux is generated by stator currents and in rotor reference frame they are Id and Iq, right? The current limit you state is also correct and yes, your stator flux vector will not be allowing optimal flux generation. That is why the power limit curve is drawn on the torque speed curve in field weakening control region. But keep in mind, this condition only applies at rated power conditions. All other places, you can boost q-axis current to get more torque as long as you dont hit the power limit. 😂 Also, use either the orthogonal components or the resultant vector for analysis. It would be better to not to mix them up when you analyze the flux vector relations. Because in the air gap, only the effective flux makes sense.
How Can the effective flux be reduced by introducing negative Id current.? Can you elaborate on this further? does this mean that by introducing negative Id current, we can effectively reduce back emf costant? If it is so then it will also have effect on Torque equation where 'effective permanent flux' is reduced.
Yes, that is why it is called field weakening control. Weakens the field. It does have an effect on torque, but the goal of field weakening is to get to higher speeds.
@@SandunKuruppu How do we model such motor then? because 'lamda_m'(permanent flux) is not modeled as a function of Id. Could you please have any literature which states that, negative Id can reduce backemf constant?
@@swanandpisat2029 Depends on the modeling platform. In Matlab, you can reduce the flux linkage constant and torque constant based on amount of field weakening.
Sir as in stator only 3 phase current is their then how we can produce negative I'd value (i can not imagine how negative I'd we push) and as in synchronous machine if angle between two flux is greater than 90 then it moves to unstable zone and if push negative I'd mean stator flux make angle greater than 90 then only one is on q axis and. One on negative d. Sir please explain how we can imagine that i push negative I'd.
It seems you are mixing up stator reference frame and rotor reference frame. You have the three phase windings in stator reference frame. When you move to rotor reference frame, you have three RL circuits. This is a hypothetical reference frame to allow for proper control. Injecting D-axis current can be though of as shifting the three phase currents by shifting three phase voltages with respect to rotor position. It is all in the phase angles when it comes to stator reference frame. I recommend you to study the relation between rotor reference frame stator reference frame. This will answer your question.
Sir before running pmsm motor on loading condition we first do the auto tuning process. In auto tuning process motor will run in forward direction followed by reverse direction
Hello sir, i have implemented field weakening of surface PMSM in MATLAB simulation while increasing speed higher than he rated speed, torque is not decreasing automatically.. could you clarify this...
Pavan, how are you weakening the field? When you apply negative Id, you also have to update you Ke (back emf constant) along with the process. Otherwise you wont see the torque decrease.
That is correct. Your back emf limits you from going beyond your maximum speed. Through field weakening you recover voltage to reach higher speed. However, now the torque is reduced as you are reducing the flux.
Did a simulation to check my hypothesis that the D axis current does not affect the torque (and torque constant), even if you go to the point where the D axis point totally cancels all of the magnets flux. Result is the same as in the torque equation, without saliency, the id has no impact on torque at all. Of course not taking into account a non linear BH curve: just a ur of 1.05. Also did not include any steel in the simulation to keep it simple. The magnetic north pole is on top. Also didnt bother with 3 phases, just made simple 2 coil system. Just Q axis current: pasteboard.co/XQMzpgWsVJUY.png Add D axis current, such that the vertial magnetic field of the magnet itself is fully canceled: pasteboard.co/PBT0LEarDMj0.png Notice how the torque doesn't change. Hence lambda doesn't change.
Elwin, I think you maybe right! I collected some experimental results which supports that Lambda is not affected. But my machines is a low current motor, so with high current this might be possible (demagnetize the magnets). But so far your I dont see a change in Lambda. I will post the data and a new video. Also @xiao sun statement about the current vector causing the flux angle to change is correct. I will mention that too. The reason I say maybe is because, I spoke to number industry members and reviewed some papers that say Lambda changes. So, maybe with high current it is possible. I cant say until I try. Lets keep the discussion open. Thank you both for helping me learn new things too!
Good explanation. thank you.
Glad it was helpful!
Explained in layman terms..... Thank you
You are welcome
thanks!
Welcome!
I have a few questions. It would be great if you could help explain.
The first question is if we apply a positive current in the D axis we could strengthen the rotor flux. Does that mean we can get more torque at low speeds?
I have a motor and inverter in a I am working on the software. I have the constant torque working correctly. I am now trying to get the field weakening working.
I have applied a negative current in the D axis but the amount of applied current to get a speed increase is very large. To get twice the speed I need to apply roughly the rated current of the motor leaving no available torque current.
However the same motor used on an EV seems to be able to spin at double the rates speed and still have torque available. Is there anything I am missing?
Francisco,
To your first question, in theory you could, but there are other factors that you have to consider. What is the flux density in the airgap before you strengthen the field, what are the loss components, etc... If you have a torque sensor, you can experiment for your choice of motor.
One aspect to check is if your position sensor is properly aligned. If not, the negative d-axis current may not be of use. You can do a simple calculation based on motor parameters to determine how much current is needed to reach a certain speed and still have torque available. Also, make sure your parameters are correct. Can you share more information on this motor and where you see this being used?
Are the two machines identical, design wise and construction wise?
@@SandunKuruppu The specs from the manufacturer are as follows:
Ld is 0.19mH
Lq is 0.44mH
Back EMF 39.2v/1000rpm
Peak torque 260Nm @380A
Continuous torque 128Nm @160A
The torque curve from the manufacturer shows peak torque to 3000rpm with 320Vdc.
Some of the numbers do not add up.
I have measured the motor back EMF to be 100v/1000RPM measuring the sine wave peak to peak.
I have measured the torque (in a not very accurate way) and the torque constants are in the ballpark.
The motor is from a Chinese EV.
@@SandunKuruppu Thank you for your video, do you have tutorial to do these calculation?
@@franciscoshi1968 Were you able to check if your position sensor alignment is accurate? Otherwise field weakening will not work.
@@haowang4306 Hao, I don't have a video on calculations as it depends on the motor design and parameters. I will try to prepare some results with calculations and experimental results.
i have one question: is lambda changable at all? as far as i recall, the d-q axis equation is derived from the three phase voltage equation with clark-park transform, and term with rotor flux even after transformation should still be only dependent on rotor magnet material, and not a "effective" "total" flux in the d-axis, right? in the previous lecture the detailed transformation was not provided, so it would be great if you could show us the abc to dq transform in detail, especially the flux linkage terms, thanks!
Hi Xiao, Lambda is called the flux linkage constant because flux has a significant contribution here. In the rotor reference frame, D-axis is where most of the flux is passing through. Therefore, introducing an opposing flux source can effectively reduce Lambda. Now, how much of a reduction in flux or would this demagnetize the magnets permanently, depend on the motor/magnets used. Does this answer your question? I will try to add another video showing the rotor reference frame model derivation soon.
@@SandunKuruppu hallo, Professor, thanks for the reply! my question was more about if we should see lambda as a constant that is only denpendent on the rotor and will not be affected by stator excitation situation? it is also linked to the torque equation, e.g. rotor flux "cross product" stator flux or Iq times rotor flux linkage. In field weakening, will the torque only be redueced/sacrificed because the rotor flux and stator flux are no longer perpendicular to each other (while the magnitude of both flux vector remain the same, since the dc-bus voltage and the permanent magner on the rotor does not change?) or the torque will reduce also with the side effect that the rotor flux magnitude will reduce due to Id in the opposite direction, in this case the torque will be even smaller?
@@xiaosun9111 Theoretically speaking, if you look at the motor model from the rotor reference frame, you lambda is a constant. Similar to Ke (back emf constant) in a DC motor. However, since this is related to magnetic saturation, practically this varies with the level of saturation. So Lambda varies a little and in a non-linear fashion as the air gap flux density changes. Since Lambda is part of the torque equation, once you weaken Lamda, your torque will also drop proportionally. The reason torque drops under proper FOC is because your effective rotor flux has been reduced, as long as you maintain orthogonality between the stator and rotor flux. However, if you do not maintain orthogonality (i.e. NOT properly field oriented), then again the torque would drop. This is not due to Lambda but due to non optimal placement of stator flux. Is this clear to you? I think you are asking about two aspects that will affect torque. Lambda is a separate issue and orthogonality of the vectors are another issue.
@@SandunKuruppu thanks! i used to think that the stator flux was generated by stator currents alone, after transformation, Id and Iq are responsible and "add up" vectorwise to the "rotating field" (of course stationary in rotor reference frame), given the voltage is limited to DC bus voltage, its magnitude (sqrt(Id^2+Iq^2)) is limited! so if a negative Id is applied, (previously the Id was 0, so all voltage can devote to Iq, and the "sum" is 90 degree to rotor) under the voltage restriction, the Iq will automatically be reduced, thus as "vector sum" of stator field will be no longer perpendicular to rotor flux😂
@@xiaosun9111 Stator flux is generated by stator currents and in rotor reference frame they are Id and Iq, right? The current limit you state is also correct and yes, your stator flux vector will not be allowing optimal flux generation. That is why the power limit curve is drawn on the torque speed curve in field weakening control region. But keep in mind, this condition only applies at rated power conditions. All other places, you can boost q-axis current to get more torque as long as you dont hit the power limit. 😂
Also, use either the orthogonal components or the resultant vector for analysis. It would be better to not to mix them up when you analyze the flux vector relations. Because in the air gap, only the effective flux makes sense.
hello sir,
will field weakening effects the magnetic fields of permanent-magnet of PMSM ??
It could if it is prolonged and severe.
How Can the effective flux be reduced by introducing negative Id current.? Can you elaborate on this further? does this mean that by introducing negative Id current, we can effectively reduce back emf costant? If it is so then it will also have effect on Torque equation where 'effective permanent flux' is reduced.
Yes, that is why it is called field weakening control. Weakens the field. It does have an effect on torque, but the goal of field weakening is to get to higher speeds.
@@SandunKuruppu How do we model such motor then? because 'lamda_m'(permanent flux) is not modeled as a function of Id.
Could you please have any literature which states that, negative Id can reduce backemf constant?
@@swanandpisat2029 Depends on the modeling platform. In Matlab, you can reduce the flux linkage constant and torque constant based on amount of field weakening.
@@SandunKuruppu Can we get in touch with some other way? can you share your email Id?
@@swanandpisat2029 Its listed under channel 'About'
Sir as in stator only 3 phase current is their then how we can produce negative I'd value (i can not imagine how negative I'd we push) and as in synchronous machine if angle between two flux is greater than 90 then it moves to unstable zone and if push negative I'd mean stator flux make angle greater than 90 then only one is on q axis and. One on negative d. Sir please explain how we can imagine that i push negative I'd.
It seems you are mixing up stator reference frame and rotor reference frame. You have the three phase windings in stator reference frame. When you move to rotor reference frame, you have three RL circuits. This is a hypothetical reference frame to allow for proper control. Injecting D-axis current can be though of as shifting the three phase currents by shifting three phase voltages with respect to rotor position. It is all in the phase angles when it comes to stator reference frame. I recommend you to study the relation between rotor reference frame stator reference frame. This will answer your question.
Hello Sir, I have one small question regarding IPMSM drive why we have to perform autotuning process for IPMSM
Are you referring to auto tuning in C2000?
Sir before running pmsm motor on loading condition we first do the auto tuning process.
In auto tuning process motor will run in forward direction followed by reverse direction
I don't know about C200.can you explain bit more about that
Hello sir,
i have implemented field weakening of surface PMSM in MATLAB simulation
while increasing speed higher than he rated speed, torque is not decreasing automatically..
could you clarify this...
Pavan, how are you weakening the field? When you apply negative Id, you also have to update you Ke (back emf constant) along with the process. Otherwise you wont see the torque decrease.
@@SandunKuruppu okay sir
how and where could i change the ke...is it in flux weakening algorithm ??
field weakening won't demagnetize flux of permanent-magnet of PMSM ??
It could if it is prolonged and severe.
What's will be effect of positive d-current to a permanent magnets?
So basically, flux, weakening, enables higher rpm with lower torque
That is correct. Your back emf limits you from going beyond your maximum speed. Through field weakening you recover voltage to reach higher speed. However, now the torque is reduced as you are reducing the flux.
Did a simulation to check my hypothesis that the D axis current does not affect the torque (and torque constant), even if you go to the point where the D axis point totally cancels all of the magnets flux. Result is the same as in the torque equation, without saliency, the id has no impact on torque at all. Of course not taking into account a non linear BH curve: just a ur of 1.05. Also did not include any steel in the simulation to keep it simple. The magnetic north pole is on top. Also didnt bother with 3 phases, just made simple 2 coil system.
Just Q axis current:
pasteboard.co/XQMzpgWsVJUY.png
Add D axis current, such that the vertial magnetic field of the magnet itself is fully canceled:
pasteboard.co/PBT0LEarDMj0.png
Notice how the torque doesn't change. Hence lambda doesn't change.
Elwin, I think you maybe right! I collected some experimental results which supports that Lambda is not affected. But my machines is a low current motor, so with high current this might be possible (demagnetize the magnets). But so far your I dont see a change in Lambda. I will post the data and a new video.
Also @xiao sun statement about the current vector causing the flux angle to change is correct. I will mention that too.
The reason I say maybe is because, I spoke to number industry members and reviewed some papers that say Lambda changes. So, maybe with high current it is possible. I cant say until I try. Lets keep the discussion open.
Thank you both for helping me learn new things too!