Good explanation sir. Complete information about SVPWM is available. Especially the teta calculation in terms of switching frequency and the output frequency is more informative and missing in almost any textbook. Thank you
Thank you for this very good presentation. I got a question about the maximum length of the vector. It seems to me, that you can get max length at the base vectors, but in the mit position of a sector, the max length of the vector is shorter like a factor 0.866. Am I right?
Yes, that is spot on. This value is the maximum non-distorted three phase output. In case you need higher voltages (vector) you can exceed the length of the vector only partially, resulting in a distorted three wave sine or as it is usually called overmodululation.
I appreciate your video but one question please, in the first place why the a, b, c vectors are placed at 0, 120, 240 degree?? I know that in a 3 phase grid power, the voltage of each phase is 120 degree apart in phase, BUT, we are not using grid power here, these a, b, c vector is phase voltage vector on each of the phase's inductor of the motor, which is generated by the switching of the inverter, not from some three phases grid that give you perfect 120 degree apart phase voltage. So how do you know that these three a,b,c vectors are always 120 degree apart as your drawing ??? Please, I have been doing a lot of reading but still no where explain this.
The three phases are created by the location of the three inductors. Even though the inductors are fed from the same voltage source they create flux in different locations of the machine. Now, if you look at the three created fluxes i.e. at their sum you will find that this sum is a rotating magnetic field at constant strength (flux vector). So, the three phases are generated by the timing of the three sine voltages applied to the inductors plus their physical location.
i dont understand when one switch is made ON the voltage appears across the phase "a" is 2/3 VDC instead of VDC, as the whole VDC will connect to the winding.
ok i love you explination , my dumb brain did not understand the last part about the integration but could i ask if you could please make a playlist for high dynamic drives. and maybe their simulink counter parts. honestly i love you. i have spend many months now tring to find the best resourc and you are 1 of the 2 best i have found. THANKS P.S please more english.
You are welcome, xganh. Regarding the integration, it is not that difficult. Think in switching periods. Here, a switching period is defined as 100e-6s (f=10kHz). During that time there is always one of the three possible vectors (100, 110, 111/000) present. Everything what happens within the period will contribute to the space vector of that particular switching period. How does every one of the three vectors contribute? It happens through integration. The physical meaning of integration here is: voltage times time. The longer a specific vector is applied to the load the "heavier" (greater) its influence on the resulting space vector gets. If you switch to the 000-vector, then there is no contribution to the resulting space vector because 0 times time is always zero. After one switching period we see what voltage vector has been integrated (created). For the next switching period we compose a new space vector (out of the three vectors). The important thing is during one switching period every tranisitor switches one time (on & off). So, if you check on your oscilloscope you will detect exactly 10 kHz switching frequency for every single IGBT. Regarding high dynamic drives, basic idea: You do not let your space vector run in "smooth" circles, changing ever so slightly its length and rotational speed. Instead, you jump around in length as well as in angle. This is best done with directly accessing the voltage space vector as opposed to the creation of the modulation by a sine-triangle-modulation. For Space Vector Modulation the sine isn't just a sine but the concept of the modulation can be equal. Since you cannot jump in time this modulation type has limits. A direct access to a space vector without having to look at its history is much more powerful.
Thanks so much for the perfect video Sir I have a question for you. How can we produce 3 phase voltage vectors using a microprocessor? Also how to control the voltage vectors produced using a microprocessor?
Thanks for your positive feedback. First of all there is only one voltage space vector. The vector results from the three voltages of the 3 motor coils. It is important to understand that the three voltages across the motor coils do not add up in space. The actually rotating voltage vector is a back-translation from the three fluxes through each of the three coils. These fluxes sum up to a rotating flux vector which in turn is back-translated to a now really rotating voltage vector = space vector, cf. ruclips.net/video/raT-hJRMtdI/видео.html Implementation will be discussed later.
It's really helpful for me, but there are still some questions that I cannot understand. The switching state goes from {0,0,0} ->{1,0,0}->{1,1,0}->{1,1,1}->{1,1,0}->{1,0,0}->{0,0,0} My question is: Why we turn off the last switch at the first stage? Why don't we turn off the second one and make it goes like {0,0,0} ->{1,1,0}->{1,1,0}->{1,1,1}->{1,0,1}->{1,0,0}->{0,0,0}?
Hi 賴冠丞, thanks for asking. The answer is simple. The vector {1,0,1} is not present in the first sector! No way you can use it during this phase. Apart from that: why do want to switch from {1,1,0}->{1,1,0}? Or was it a mistake? Maybe you'll watch the video a second time? It is not that difficult.
2-level SVM is very trivial. It becomes much more interesting and complex when multilevel inverters are involved because of the much higher number of (redundant) switching stages, as well as more sub-sectors (areas) per each 60deg sector.
Great tutorial Sir! Could you please do another video facing how to implement Space Vector Modulation which generates pulses for gate drivers on a microcontroller? And it would be great so see such a video as this about Field Orientated Control
The SVM with VSI produces a three phase voltage system at its output. Therefore it is meant to drive three phase motors. Single phase motors use other (= non-three-phase-) voltages/currents to create a rotating magnetic field. This rotating magnetic field is either - turned off after start up leaving a pulsating field fed by a single voltage phase or - remains maintained by a second auxiliary phase throughout the operation. Either way the rotating field in both cases is not symmetrical and it is not possible to feed the main winding and the startup winding (in a way that makes sense from an engineering point of view). In a nutshell: Operate single phase motors with a single phase inverter and three phase motors with the introduced three phase VSI. :)
thanks sir i want only to know something do we use reference and carrier signals in SVM as in SPWM or not and if the answer is yes so what is the difference between the two ways in operation and what is the relation between this space vector and the output voltage of SVM technique in other words how does it used to generate the output voltage sorry for taking long sir
Actually, the way I understand the problem is SVM actually focusses on the rotating space vector using the eight switching states shown in the video. In short there is no connction for the star point (not regarded and not needed). SPWM creates three sine waves from the three half brigdes against star (half DC-link voltage). So you can run three different loads against half DC-link and still have a (balanced) three phase voltage system (with regard to voltages). However, if you want to run SPWM without star connection you have to add the third harmonic for a "clean" (50 Hz rotating-space vector) ... In the end you can have equal behaviour from SVM and SPWM (after tweaking).
@@quellstrom great sir,but how the pattern of switching can be determined in SVM if it is compared with the approach used in SPWM in which we can generate pulses and control their width to know the pattern of switching do we use SPWM in SVM? in other words how we can determine the magnitude and rotating frequency of the rotating vector? tanks lot sir
The length of the pointer is proportional to the inverter ouput voltage. The difference between inverter voltage and induced motor voltage across the inductances causes the current. If you need a general idea, the times of [0,0,0] and [1,1,1] need to become longer for shorter voltage pointers.
Thanks for your comment. The capacitor is not the point. All our DC links are capacitor supported voltage sources (in case of voltage source topologies). The same goes for MMC vs. CHB. The difference is the multilevel. Multilevel means there are more switching states per leg than just two. And as always, one inverter comprises three legs (= three phases). Two level / one leg: 2 states Three level / one leg: 4 states Five level / one leg: 8 states ... Actually, I don't have the numbers at hand but there are many more switching states per voltage vector (now all three legs involved). This translates to a lower ripple voltage for the specific space voltage vector (which is still the one rotating voltage vector with constant length and constant rotating frequency). So the goal voltage vector for multivlevel inverters is the same as for two level inverters. PS. This is a very good read on the general subject. www.mdpi.com/2079-9292/8/11/1329/pdf
This is a great Video, Sir. However a question appeared as you mentioned high dynamic motor drives. What you showed in the paragraph “Working a sector” means if I’m right, that you are able to create any reference vector you want in sector 1 by varying the on and off times of the transistors in one switching period, right? Is this accurate in real time that there will be only one reference vector in sector 1 so overall 6 switching periods for one cycle, or does this depends on the frequency? So if you would have twice the frequency there will be two reference vectors each sector so two switching periods each sector and overall 12 vectors and switching periods for a hole cycle? Practically referred to your example this frequency would result with 1.8° 60°/1.8° so 33 vectors and 33 switching periods for sector 1 then following 33 switching periods passing through sector 2 and so on? And varying the t1, t2 values (which just represent the transistor on off times) let the reference vector “rotate” through the sectors? Am I right in thinking so? And specifically in motor drives you would change the frequency to adjust the motor speed by varying through the frequency the angle and the amount of vectors and switching periods in each sector? And if I’m right this information of the t1 and t2 values are used to set the compare register in the microprocessor which then outputs the signals for the gate drivers?
Thank you, Timo. Frankly, I don't really understand your first paragraph with the "two reference vectors": For every switching period, 100microseconds for our example, you create a new vector. Your reasoning in the second paragraph however, is right on though. It is worth mentioning that the switching frequency is usually not synchronized with the voltage vector. That results in unequal numbers of switching periods for every sector even for constant output frequencies.
Yes, of course. Let's look at a feeding 400V-system. The maximum space vector value (constant length, i.e. undistorted, not over-modulated) is Vsp=sqrt(3)/2*VDC, with VDC being the DC link voltage (voltage source). A space vector with the length of sqrt(3)/2*VDC has an amplitude per phase of sqrt(3)/2*VDC * ( 1 - 1/3) = .577 VDC, cf. construction of Space Vector. In other words the amplitude of line-ground(star)-amplitude for maximally modulated voltage is 323 V for VDC = 550 V, which is Vlg = 323V/sqrt(2) = 228V. The is somewhat below the line-ground-amplitude (325V) that was used to obtain the 550V DC link voltage. Alternatively one can interprete the DC-link voltage VDC as peak amplitude of a line-to-line motor voltage. That makes the peak value of the line-to-ground voltage VDC/sqrt(3) = 317 V. Therefore the line-ground-rms-voltage is then Vlg = 317 V / sqrt(2) = 224 V.
@@quellstrom Hi, thank you for the reply. I am sorry I don't understand your answer. I will rephrase my question. Let's suppose that I have a 400Vrms 3-phase supply applied across a diode rectifier; the voltage that I will see across the DC link capacitor will be of about 400*sqrt(2)=565.6V(dc). This DC voltage is then fed to the inverter. If I set a reference voltage space vector with amplitude of 200V, am I asking the inverter to produce an output 3-phase voltage of 200Vrms (line to line), or am I asking to produce an output 3-phase voltage where each phase has a peak value of 200V?
There is no real quick answer to this. 200V rms is 283 V amplitude. According to the scaling shown in the vid (not transformed, VDC as the perimeter of the circle) you need a voltage vector of 283V * (1 + 1/3) = 377 V (constant length). There are formulae for calculating the vectors. That is, what duty ratios are required for bridge 1 ... bridge 3 in dependence on switching frequency, mains frequency and DC-link voltage. As already hinted at, most of the time these data are calculated in the transformated plane (Clarke and Park) and backtranslated for the actual control. There you'll have a nasty factor, which obscurs the basic physical understanding BUT helps a lot with calculating the control values ... :-)
For industrial drives there are usually IGBTs. For lower voltages you may also use MOSFETs. For very high power inverters (transmission) GTOs/IGTCs are used.
Actually, if you understood this video you are half way there. Now just simply connect an induction machine, learn about the rotational speed / torque / voltage (amplitude and frequency) diagrams and you are almost there. For recuperating and sophisticated acceleration runs you need some more time. How far do you want to go? ... :-)
This video tells you how to obtain a rotating voltage vector. According to Faraday's law this rotating voltage can be translated into a rotating magnetic field. And even though the magnetic field is lagging 90degrees -- as long as the rotating frequency is constant (nominally 50 Hz), the magnetic flux is proportional(!) to the voltage. So, from the states you do not see the magnetic field directly but learn how to create a rotating voltage vector. And what to do with it you know now. :-)
Of course. Every cycle sees 0,0,0 and 1,1,1. It is rather the other combinations that do not occur in every sector. So, apart from 0,0,0 and 1,1,1 there are always (only) two other states (depending on the sector) that will occur. In other words one cycle consists of four states.
@@quellstrombut (1,1,1) would mean active shortcut of the dc-source -> practically not done, right? and (0,0,0) could lead to massive voltage peaks -> only 6 states are left (hexagon)
@@kevinhohne7889 No, (1,1,1) and (0,0,0) shortcircuits the motor not the DC-link. Pls. watch the video again ... ;-) Every switching cycle contains a (0,0,0) and a (1,1,1) 11:01
My English isn't very good, but I can understand the main. Thanks. The best video of YT.
I've read many articles about this topic but there was still something missing. Now I finally understand, thank you!
The best ever explanation on SVPWM!
Just the BEST explanation I have ever seen for SVPWM
Thank you. That is very kind.
This tutorials should get a million views at least....Gut gemacht und Tausend Danke
Thank you. Let's hope for 10,000 views first :)
The best explanation I have seen so far. Thank you!
It was the best svm presentation ever definately!
Thank you, Omid. Very kind.
The best explanation. Thanks teacher!
Thank you.
It is a great video for SVPWM understanding
Very nice. Thank you.
Good explanation sir. Complete information about SVPWM is available. Especially the teta calculation in terms of switching frequency and the output frequency is more informative and missing in almost any textbook. Thank you
You're welcome. Glad it was of help.
Really clear explanation. Thanks for the video!
Short and clear explanation...
Thank u😊
This is worth watching. ❤
Thank you.
The best explanation I have seen ever. Many thanks.
Thank you. That's very kind of you.
Loved the explanation. Clear precise and easy to understand
Thank you for this very good presentation. I got a question about the maximum length of the vector. It seems to me, that you can get max length at the base vectors, but in the mit position of a sector, the max length of the vector is shorter like a factor 0.866. Am I right?
Yes, that is spot on. This value is the maximum non-distorted three phase output.
In case you need higher voltages (vector) you can exceed the length of the vector only partially, resulting in a distorted three wave sine or as it is usually called overmodululation.
Sehr gut
Perfect explanation 🙌
Thank you. I am glad you liked it. :-)
Very good and simple to understand! Thanks! :)
Amazing animations!
Very nice explanation, thank you
Fantastic explanation, great job mate!
really great explanation , thanks
you r perfect man!!!!!!
Thank you for this video, it helped me a lot to understand the concept of this technique.
Thanks!!!
thank you for this explanation
Very good sir
Well Explained. Great Job
I appreciate your video but one question please, in the first place why the a, b, c vectors are placed at 0, 120, 240 degree?? I know that in a 3 phase grid power, the voltage of each phase is 120 degree apart in phase, BUT, we are not using grid power here, these a, b, c vector is phase voltage vector on each of the phase's inductor of the motor, which is generated by the switching of the inverter, not from some three phases grid that give you perfect 120 degree apart phase voltage. So how do you know that these three a,b,c vectors are always 120 degree apart as your drawing ??? Please, I have been doing a lot of reading but still no where explain this.
The three phases are created by the location of the three inductors. Even though the inductors are fed from the same voltage source they create flux in different locations of the machine.
Now, if you look at the three created fluxes i.e. at their sum you will find that this sum is a rotating magnetic field at constant strength (flux vector).
So, the three phases are generated by the timing of the three sine voltages applied to the inductors plus their physical location.
the best one,thank you
i dont understand when one switch is made ON the voltage appears across the phase "a" is 2/3 VDC instead of VDC, as the whole VDC will connect to the winding.
Have you considered the voltage divider? It is never the "full VDC" that is applied to the phases a, b, and c.
thanks for making this
Thank you, Huang. Glad you liked it.
Well explained great job,,,,,,,,,,,,,,,,,,,,,
thank you very much..you saved my life:))
Really? You are welcome. Tell your friends and have a party. Stick to the Corona rules though. ;-)
Good Explanation 🎉
You are welcome. :-)
Really good...if anyone read this comment. My suggestion is please watch this video.
Thank you, Chirag. That is very kind of you.
Good explanation
ok i love you explination , my dumb brain did not understand the last part about the integration but could i ask if you could please make a playlist for high dynamic drives. and maybe their simulink counter parts.
honestly i love you. i have spend many months now tring to find the best resourc and you are 1 of the 2 best i have found.
THANKS
P.S please more english.
You are welcome, xganh. Regarding the integration, it is not that difficult. Think in switching periods. Here, a switching period is defined as 100e-6s (f=10kHz). During that time there is always one of the three possible vectors (100, 110, 111/000) present. Everything what happens within the period will contribute to the space vector of that particular switching period.
How does every one of the three vectors contribute? It happens through integration. The physical meaning of integration here is: voltage times time. The longer a specific vector is applied to the load the "heavier" (greater) its influence on the resulting space vector gets. If you switch to the 000-vector, then there is no contribution to the resulting space vector because 0 times time is always zero. After one switching period we see what voltage vector has been integrated (created). For the next switching period we compose a new space vector (out of the three vectors).
The important thing is during one switching period every tranisitor switches one time (on & off). So, if you check on your oscilloscope you will detect exactly 10 kHz switching frequency for every single IGBT.
Regarding high dynamic drives, basic idea: You do not let your space vector run in "smooth" circles, changing ever so slightly its length and rotational speed. Instead, you jump around in length as well as in angle. This is best done with directly accessing the voltage space vector as opposed to the creation of the modulation by a sine-triangle-modulation. For Space Vector Modulation the sine isn't just a sine but the concept of the modulation can be equal. Since you cannot jump in time this modulation type has limits.
A direct access to a space vector without having to look at its history is much more powerful.
I love the way you explained things. Excellent Video!!! 1 like and 1 subscribe. Thanks
Thanks so much for the perfect video Sir
I have a question for you. How can we produce 3 phase voltage vectors using a microprocessor?
Also how to control the voltage vectors produced using a microprocessor?
Thanks for your positive feedback. First of all there is only one voltage space vector. The vector results from the three voltages of the 3 motor coils. It is important to understand that the three voltages across the motor coils do not add up in space.
The actually rotating voltage vector is a back-translation from the three fluxes through each of the three coils. These fluxes sum up to a rotating flux vector which in turn is back-translated to a now really rotating voltage vector = space vector, cf. ruclips.net/video/raT-hJRMtdI/видео.html
Implementation will be discussed later.
The best
That is very kind. Thank you.
It's really helpful for me, but there are still some questions that I cannot understand.
The switching state goes from {0,0,0} ->{1,0,0}->{1,1,0}->{1,1,1}->{1,1,0}->{1,0,0}->{0,0,0}
My question is:
Why we turn off the last switch at the first stage?
Why don't we turn off the second one and make it goes like {0,0,0} ->{1,1,0}->{1,1,0}->{1,1,1}->{1,0,1}->{1,0,0}->{0,0,0}?
Hi 賴冠丞, thanks for asking. The answer is simple.
The vector {1,0,1} is not present in the first sector! No way you can use it during this phase. Apart from that: why do want to switch from {1,1,0}->{1,1,0}? Or was it a mistake? Maybe you'll watch the video a second time? It is not that difficult.
@@quellstrom Oh! I just type it( {1,1,0}->{1,1,0}) wrong ,but thanks!!! I got it now!!!
Great!
2-level SVM is very trivial. It becomes much more interesting and complex when multilevel inverters are involved because of the much higher number of (redundant) switching stages, as well as more sub-sectors (areas) per each 60deg sector.
Yes.
Great tutorial Sir!
Could you please do another video facing how to implement Space Vector Modulation which generates pulses for gate drivers on a microcontroller?
And it would be great so see such a video as this about Field Orientated Control
Thank you. There's more on my list ... Implementation and FOC will be addressed.
Amazing video.
What software do you use to draw your schematics?
Thank you. All drawings are made with Visio.
@@quellstrom Thanks for your reply. That is a special stencil.
@@RaedMohsen I have been using it like forever. 😀
How can I implement this SVM/VSI to Single Phase motor (splitting Run, Start and Common winding, removed Start Capacitor)?
The SVM with VSI produces a three phase voltage system at its output. Therefore it is meant to drive three phase motors.
Single phase motors use other (= non-three-phase-) voltages/currents to create a rotating magnetic field.
This rotating magnetic field is either
- turned off after start up leaving a pulsating field fed by a single voltage phase or
- remains maintained by a second auxiliary phase throughout the operation.
Either way the rotating field in both cases is not symmetrical and it is not possible to feed the main winding and the startup winding (in a way that makes sense from an engineering point of view).
In a nutshell: Operate single phase motors with a single phase inverter and three phase motors with the introduced three phase VSI. :)
thanx sir
thanks sir
i want only to know something do we use reference and carrier signals in SVM as in SPWM or not and if the answer is yes so what is the difference between the two ways in operation and what is the relation between this space vector and the output voltage of SVM technique in other words how does it used to generate the output voltage
sorry for taking long sir
Actually, the way I understand the problem is SVM actually focusses on the rotating space vector using the eight switching states shown in the video. In short there is no connction for the star point (not regarded and not needed).
SPWM creates three sine waves from the three half brigdes against star (half DC-link voltage). So you can run three different loads against half DC-link and still have a (balanced) three phase voltage system (with regard to voltages).
However, if you want to run SPWM without star connection you have to add the third harmonic for a "clean" (50 Hz rotating-space vector) ...
In the end you can have equal behaviour from SVM and SPWM (after tweaking).
@@quellstrom great sir,but how the pattern of switching can be determined in SVM if it is compared with the approach used in SPWM in which we can generate pulses and control their width to know the pattern of switching
do we use SPWM in SVM?
in other words how we can determine the magnitude and rotating frequency of the rotating vector?
tanks lot sir
اخي لم افهم حساب التردد. اريد منك مثالاً لوقت كل مرحله للتبديل بين المراحل على فرض تردد ٥٠ هرتز. ممكن تخبرني وقت كل مرحله كم وشكراً
How to control the output power in SVPWM scheme? How the gate driving traces would look like at reduced power?
The length of the pointer is proportional to the inverter ouput voltage. The difference between inverter voltage and induced motor voltage across the inductances causes the current.
If you need a general idea, the times of [0,0,0] and [1,1,1] need to become longer for shorter voltage pointers.
What if ur dc link is capactor, i mean in MMC's. Any reference to read?
Thanks for your comment. The capacitor is not the point. All our DC links are capacitor supported voltage sources (in case of voltage source topologies). The same goes for MMC vs. CHB.
The difference is the multilevel. Multilevel means there are more switching states per leg than just two. And as always, one inverter comprises three legs (= three phases).
Two level / one leg: 2 states
Three level / one leg: 4 states
Five level / one leg: 8 states
...
Actually, I don't have the numbers at hand but there are many more switching states per voltage vector (now all three legs involved). This translates to a lower ripple voltage for the specific space voltage vector (which is still the one rotating voltage vector with constant length and constant rotating frequency).
So the goal voltage vector for multivlevel inverters is the same as for two level inverters.
PS. This is a very good read on the general subject. www.mdpi.com/2079-9292/8/11/1329/pdf
Superb...sir can we have simulation in MATLAB for this
Thanks for the reply. No, I don't have matlab simulations at this time.
this shit is fxcking goooood!
How shoud I change it for current source converters?
This is a great Video, Sir. However a question appeared as you mentioned high dynamic motor drives. What you showed in the paragraph “Working a sector” means if I’m right, that you are able to create any reference vector you want in sector 1 by varying the on and off times of the transistors in one switching period, right? Is this accurate in real time that there will be only one reference vector in sector 1 so overall 6 switching periods for one cycle, or does this depends on the frequency? So if you would have twice the frequency there will be two reference vectors each sector so two switching periods each sector and overall 12 vectors and switching periods for a hole cycle?
Practically referred to your example this frequency would result with 1.8° 60°/1.8° so 33 vectors and 33 switching periods for sector 1 then following 33 switching periods passing through sector 2 and so on? And varying the t1, t2 values (which just represent the transistor on off times) let the reference vector “rotate” through the sectors? Am I right in thinking so?
And specifically in motor drives you would change the frequency to adjust the motor speed by varying through the frequency the angle and the amount of vectors and switching periods in each sector? And if I’m right this information of the t1 and t2 values are used to set the compare register in the microprocessor which then outputs the signals for the gate drivers?
Thank you, Timo. Frankly, I don't really understand your first paragraph with the "two reference vectors": For every switching period, 100microseconds for our example, you create a new vector.
Your reasoning in the second paragraph however, is right on though. It is worth mentioning that the switching frequency is usually not synchronized with the voltage vector. That results in unequal numbers of switching periods for every sector even for constant output frequencies.
can anyone suggest me any video or documented material on how to implement this concept in MATLAB!!
احتاج كود اردوينو لتشغيل متور دسي ثلاجه وبمقابل اجر مادي. ارجو الرد
Hi there, thanks for the video. Does the amplitude of a reference space vector corresponds to the rms or peak value of the output voltage?
Yes, of course. Let's look at a feeding 400V-system.
The maximum space vector value (constant length, i.e. undistorted, not over-modulated) is
Vsp=sqrt(3)/2*VDC,
with VDC being the DC link voltage (voltage source). A space vector with the length of sqrt(3)/2*VDC has an amplitude per phase of sqrt(3)/2*VDC * ( 1 - 1/3) = .577 VDC, cf. construction of Space Vector. In other words the amplitude of line-ground(star)-amplitude for maximally modulated voltage is 323 V for VDC = 550 V, which is Vlg = 323V/sqrt(2) = 228V. The is somewhat below the line-ground-amplitude (325V) that was used to obtain the 550V DC link voltage.
Alternatively one can interprete the DC-link voltage VDC as peak amplitude of a line-to-line motor voltage. That makes the peak value of the line-to-ground voltage VDC/sqrt(3) = 317 V. Therefore the line-ground-rms-voltage is then
Vlg = 317 V / sqrt(2) = 224 V.
@@quellstrom Hi, thank you for the reply. I am sorry I don't understand your answer. I will rephrase my question. Let's suppose that I have a 400Vrms 3-phase supply applied across a diode rectifier; the voltage that I will see across the DC link capacitor will be of about 400*sqrt(2)=565.6V(dc). This DC voltage is then fed to the inverter. If I set a reference voltage space vector with amplitude of 200V, am I asking the inverter to produce an output 3-phase voltage of 200Vrms (line to line), or am I asking to produce an output 3-phase voltage where each phase has a peak value of 200V?
There is no real quick answer to this. 200V rms is 283 V amplitude. According to the scaling shown in the vid (not transformed, VDC as the perimeter of the circle) you need a voltage vector of 283V * (1 + 1/3) = 377 V (constant length).
There are formulae for calculating the vectors. That is, what duty ratios are required for bridge 1 ... bridge 3 in dependence on switching frequency, mains frequency and DC-link voltage.
As already hinted at, most of the time these data are calculated in the transformated plane (Clarke and Park) and backtranslated for the actual control. There you'll have a nasty factor, which obscurs the basic physical understanding BUT helps a lot with calculating the control values ... :-)
Best evr
Where does the 50hz in Tmains come from plz ?
It's the mains frequency unless you live on the American continent (middle and north) or parts of Japan. Then it would be 60Hz.
@@quellstrom ohhhh thnx
What kind transistor of that in your circuit?
For industrial drives there are usually IGBTs. For lower voltages you may also use MOSFETs. For very high power inverters (transmission) GTOs/IGTCs are used.
Can you help me? I need the Arduino code with this explanation. For a fee
i want to know how tramcars work
Actually, if you understood this video you are half way there. Now just simply connect an induction machine, learn about the rotational speed / torque / voltage (amplitude and frequency) diagrams and you are almost there. For recuperating and sophisticated acceleration runs you need some more time. How far do you want to go? ... :-)
@@quellstrom thank you
this is the magnitic field in all states ?
This video tells you how to obtain a rotating voltage vector. According to Faraday's law this rotating voltage can be translated into a rotating magnetic field. And even though the magnetic field is lagging 90degrees -- as long as the rotating frequency is constant (nominally 50 Hz), the magnetic flux is proportional(!) to the voltage.
So, from the states you do not see the magnetic field directly but learn how to create a rotating voltage vector. And what to do with it you know now. :-)
gibt es auch Deutsche Variente? auf Englisch fühle ich mich nicht zu Hause:)
Leider nicht. (Habe momentan auch nichts in Planung.)
@@quellstrom Danke, selbstverständlich, da ist die Zielgruppe deutlich höher.
In reality (0,0,0) and (1,1,1) is never active, no?
Of course. Every cycle sees 0,0,0 and 1,1,1. It is rather the other combinations that do not occur in every sector.
So, apart from 0,0,0 and 1,1,1 there are always (only) two other states (depending on the sector) that will occur. In other words one cycle consists of four states.
@@quellstrombut (1,1,1) would mean active shortcut of the dc-source -> practically not done, right? and (0,0,0) could lead to massive voltage peaks -> only 6 states are left (hexagon)
@@kevinhohne7889 No, (1,1,1) and (0,0,0) shortcircuits the motor not the DC-link. Pls. watch the video again ... ;-) Every switching cycle contains a (0,0,0) and a (1,1,1) 11:01
Inverter