Motor Control From Scratch - Part4 | Field Oriented Control (FOC) Explained!

[@desireddysatyanarayanaredd6789]

I don't understand the factor 2/3, if i resolve the current Ib and Ic along Alpha and Beta axis, i get -Ibsin30 -Ibsin30 and add with Ia is Ia - 0.5Ib - 0.5Ic => 3Ia/2 this is what i get. Why should we multiply with 2/3? Can u advise.

[@thesubcooledmind7910]

Nice Question. 2/3 is just a scaling factor and you can use other scaling factors as well like 1 and sqrt(2/3). But why do you need to scale and why 2/3? Actually there is no such requirement of scaling but we do so for our convenience. On scaling by 2/3, we make our transform "Magnitude Invariant". What that means is sqrt(Id^2 + Iq^2) will give you the magnitude of the 3-phase currents. Basically, we have matched the magnitudes. If you did not use the scaling factor 2/3 in the Clarke Transform, you would have to account for this change in magnitude somewhere in your control loop. There is something called "Power Invariant" Transform also which uses the scaling factor of sqrt(2/3) instead of 2/3 and this tries to match the power in different reference frames. For motor control applications, the 2/3 factor is widely used. I hope this answers your question.

[@mailtoxyz123]

@@thesubcooledmind7910 Thank you for the reply, i need to analyze more to completely understand this, meanwhile can u share personal email to share my detailed understanding i cant write equations etc here.

[@mailtoxyz123]

@@thesubcooledmind7910 sorry i posted from different acnt. This acnt same as original question.

[@thesubcooledmind7910]

@mailtoxyz123 [e-mail]

[@abhijithreji7555]

@@thesubcooledmind7910 why magnitude variant is widely used in motor control applications? Any specific reason for that?

[@gustavoaguayo4007]

This is one of the best explanations that I've seen so far about this topic, congratulations!

[@chrispeng621]

amazing clear and summarized explanations which totally solved my confusion while reading text book as a non electric engineer. Appreciated for the kind knowledge sharing! Cheers

[@nitinahlawat243]

What a great video. You've done a great job explaining the topic. Also, the graphics are good. You must have put lot of time to make this beautiful. Keep it coming ✌️

[@thesubcooledmind7910]

Thanks. Your support does encourage me to keep making new videos.

[@hamdanrocks.k.k..........9218]

each time i watch this iam becoming really happy for understanding topics. thank you very much.

[@vamsikrishna5391]

Ur the person I'm looking for , superb explanation ,keep rocking with ur content

[@BreakByChance]

Your channel is soaring high with long way ahead! All the best, great content Rushi

[@CAME18471]

Really good material. Could you please enable the subtitles?

[@nikhilvaliveti5717]

Seems like you have put a Phenomenal Effort in making this excellent presentation. Got you from the Linkedin post. I still found it useful despite having experience in it. Keep Continuing this good work. Perfect explanation. Subscribed!

[@thesubcooledmind7910]

Thanks for subscribing. I am happy to know that you found it useful.

[@pujitgandhi1698]

Amazing sir......you have explained so clearly...for forward to learn more from you

[@shikharvohra1009]

Absolutely wonderful. I have been trying to understand this but have failed to it in the simple manner that you have done. Again, Thanks a ton for these videos, and the resource material that you have shared too

[@thesubcooledmind7910]

It's great to know that you have understood the concepts. Keep Learning!

[@kumarbalaji3344]

Amazing video. I'm in the field of motor control (hardware) and learning control system part by your videos. Thank you!

[@hamdanrocks.k.k..........9218]

great video. highly useful

[@thesubcooledmind7910]

Thanks for watching this video. If you have any questions, leave them in the comments and I would be happy to answer it. Cheers!

[@청동이-u6p]

thank you for great video clip. can we just put current a's sin(x) to alph since it's same quantity, I think we don't need to do alpha = a*2/3 - b/3- c/3 . please teach me if I'm wrong. and stay safe from corona virus.

[@amberkothari6594]

Great Video Rushi 👍👍

[@brushhead]

Is Id always positive?

[@srikanthsarjanaa8075]

The interrupt is to do certain tasks that have to be done at a particular point in time and the event has to drive it right. Could you tell me what are the events here that are causing interrupts in detail explanation? Thank you for your efforts are really good.

[@thesubcooledmind7910]

Yes, that's a good question. For FOC, you need to measure the currents using an ADC. And someone has to trigger this ADC. In general, we trigger this ADC when a PWM up-down counter reaches the top. This will help you to get an average current waveform. (If you have not got this part, I will be explaining it in future videos, or feel free to comment and take this discussion forward. You can also read this link: community.nxp.com/thread/469445 ). Now coming to interrupts. We want the FOC to run after the ADC has read the new values of the current and rotor position. So, this is how the flow goes. The PWM in a micro-controller will trigger the ADC to read currents and then at the End of Conversion of the ADC peripheral, interrupt will be triggered.

[@nikhilvaliveti5717]

Actually, Field oriented control also talks about the decoupling of Field flux and Torque right, to get a good dynamic response like as of separately excited dc motor behavior i.e independently we can control torque and flux in the machine, But What I don't understand is if we change the field flux even in separately excited torque control, T= K*Phi* I. torque increases with increase in flux right? How is it independent? I only get that stator flux and field flux are made perpendicular and will get maximum torque due to this, I don't understand completely about decoupling behavior. Can you please elaborate on it, please? Please correct me If am wrong at any point Thanks in advance.

[@thesubcooledmind7910]

You are absolutely right when you say that the torque depends on flux. So, how are they independently controlled or decoupled? Let me tell you the way I think. The d-axis current controls the flux and q-axis current controls the torque. But flux indirectly controls the torque as well. However, I have never seen someone use d-axis current to "increase" flux so as to increase torque. You can always increase the q-axis current for more torque. But in field weakening, you do use d-axis current to "decrease" the net flux to achieve more speeds. (at the expense of lesser torque) In the end, do not get confused by the terminology used. The main advantage of DQ axis is that you have to deal with DC quantities which are easier to control. And if you understand all the terms related to it and the dynamics of the system, you are good to go.

[@nikhilvaliveti5717]

@@thesubcooledmind7910 thanks for the reply, so then in field weakening region, torque reduces with decrease in flux(i.e through decreasing ids) right? So they are not decoupled here right? Same is valid in separately excited dc motor field weakening as well i.e Torque reduces with decrease in flux right? If you take motor operation below rated speed in separately excited dc motor, there only voltage is controlled. Torque, current flux are kept constant right? Then how can we say that torque and flux are decoupled in separately excited dc motor? Decoupling in dc motor lead to the idea of foc in ac drives right? So if decoupling is not met, what's use of FOC then? I feel i am wrong some where, but i am not able to find it. Could you please shed some light on this?

[@thesubcooledmind7910]

Yes, they are not decoupled. I have not read extensively about DC motor. But the main use of FOC is that you do not have to deal with 3-phase AC currents. You have nice DC quantities - and the equations are similar to a DC motor. Hence, we can control a 3-phase motor almost the same way as a DC motor is controlled. If you don't use FOC - then you will find it difficult to control the currents.

[@josephma5961]

Decoupling refers to non interference between networks. We should ideally be able to change torque and flux independent of each other. In case of a separately excited DC machine, Field winding is inherently decoupled from armature - the equation describing field voltage/linkage does not involve armature currents. Hence changes in armature current doesnt reflect into field. Part 1 of decoupling done for us automatically. The other way around - the equations of armature involve field currents/linkages. which means they are not inherently "decoupled". Then we usually use feedforward or similar methods to deduce desired armature current for a certain torque requirement. Example: i_arm_ref = Te_ref/(M/Lf)*Psi_f Here by incorporating psi_f into calculation of ia, we have ensured that no matter what is the actual flux set at, the torque that you desire must be produced. Therefore, you can say armature current has finally been decoupled from field ckt and two-way decoupling is complete. In case of AC machine the structure doesnt come with inherent decoupling along any of the axis. So things have to be done on both channels. All the decoupling are subject to several constraints and are not always followed, especially during fast transients, voltage limits of inverter, etc or when we fail to estimate/use parameters closely. PS: I am also learning, thats the only reason I tried to explain.

[@sudheerkumar5966]

Very good sir 🙏🙏🙏🙏🙏🙏 Plz make more vedioes on ev

[@mark-qi6di]

10:12 Please can you explain why there is a ramp function? I am trying to simulate the Park/Clark transformations in Simulink. Do I need to put the ramp function somewhere?

[@przemysawtwardy1792]

Loves it

[@gopumohandas910]

🙌💯

[@rajatgupta619]

If you want to know more about motor control go and visit SP DAS videos on NPTEL

[@nikhilvaliveti5717]

Is it a standard that alpha leads beta, can't we consider vice versa?

[@thesubcooledmind7910]

You can definitely use your own convention. But you need to keep in mind that all other calculations need to be based on that. For eg, if you swap beta and alpha axis, the Clarke Transform will change. Similarly, if you consider anti-clockwise rotation as leading, then you need to take care of the your rotor position convention. To understand this better, I would suggest you to simulate once with A leads B leads C, and then next with A lags B lags C. In second case, dq quantities will not be DC quantities unless you make changes to your ramp (theta). At the end, the only thing you need to make sure is that the back-emf should be in-phase with the phase-currents. Irrespective of the convention, this is the only requirement. So, whichever paper or doc you read, make sure that the convention used there matches with what you have chosen, otherwise you will be in trouble.

[@prayanshuthakur1985]

Best

[@z.g6785]

insan şuna bi ingilizce alt yazı ekler . You should Add english substide