This is the best video about matlab for power electronics that i have seen It is much better than the webinars from mathwor. Perfect balance to make a model to work. The webinars from matwork focuse too much in other aspecst and very little about the simscap issues For example; one webinar that uses the sepic comverter focused so much in the power electronics concepts and very little about the simscap solver
i have been using the black model for the longest time and it has some limitations and I didn't know how to operate the simscape model, and the trick was to just convert the signal !!!! thank you so much
I am developing a custom power converter. I tried using Simscape to simulate it, but the results were completely nonsensical. This video provided invaluable information about the appropriate solver type that should be used for this type of task, along with a lot of other useful information. Thank you so much!
Great video! May have just saved my semester lol. I would be curious to know how to do some of the more in depth options that you kind of skipped over for the sake of simplicity. You should definitely make some more tutorials.
Oh I see, your switching the transistors to control charge of capacitor to achieve desired voltage, I’m guessing that’s what the voltage control mechanism involves. I like this tutorial because of the emphasis on practical applications regarding the use of simscape to produce some predicted behaviors. This tutorial seems to convey the significance of simscape in that it’s an implicit modeling tool that’s more intuitive compared to simulink as an explicit modeling tool. As a result of feedback in simulink the need to avoid algebraic loop and so introduce an integrator in low pass filter. I guess you choose to impliment the control portion in simulink because there are more blocks to choose from. I say that because algebraic loops are avoided if the feedback is implemented with simscape physical signal blocks
You are correct. One rule of thumb I use to simplify the decision is this. If you are going to implement your controls in a digital/embedded system that is clocked, use Simulink, and break the loop with a unit delay representing the controller update rate. If you are implementing it analog, leave everything in simscape (granted, it's easier to tune the controller in Simulink). However, if you want to go to Simulink for the loop design, there is a trick to getting past the algebraic loop although it's a little bit of a hack. For instance, if you want to close the loop on the inductor's current, you can measure it's voltage with a sensor in Simscape and pass that signal through an integrator in Simulink, divided by the inductance. This will tell you the current very very very closely to what simscape will calculate, and having the integrator in the Simulink portion of the model breaks the loop. This would cause problems if the simulation was run for a very very long time, but the error is so small it takes a while to accumulate and is very useful for tuning gains and using the controls toolboxes for initial loop design before going back to the hardware design for implementation.
I have seen using black resistance as load in some tutorials but I couldn’t get them in my libraries and the boundary connectors are not proper there All other part of my circuit are simulink modules(black coloured) Any solution or correction do you have
the black circuit elements are part of a sub library called specialized power systems. They use a different technology implementation and are not compatible with standard simscape components. There are ways to connect them, but generally make the simulation slow or unstable. Most times, there is a direct equivalent block in the simscape library that you can use instead.
I suspect the black mosfet is just to provide more of a physical interpretation to standard simulink blocks. While simscape is true implicit modeling with predictions for both current and voltage and so power delivery occurring at the terminals So I guess the black components are masked subsystems to be more representative of actual electrical components
great video, we need more. I have a question that I don't see the answer to anywhere on the internet. Here it is: I am trying to compare a DC and AC microgrid in terms of efficiency. In Simscape there are blocks with a converter which only require that you set the efficiency based on output current.. For some reason I am unable to get this right. The way I see it. when I put a variable resistor on the output and change its resistance, I should get as a current measurement the current minus the efficiency percentage I have put. But I don't get that. I input [0,10,20,30,40,50,60,70,80,90,100] as an efficiency vector for [0,10,20,30,40,50,60,70,80,90,100] as output currents. SO I expect that for 100volts output and a 20 ohm resistance that I get 50% efficiency which should give 1/2 of 5 which is 2.5.
If you have 100 Volts out at 20 ohms, you will have 5 amps, based on your vector, 5 amps = 5% efficiency, so you should get almost nothing out of the converter. Try using a vector like this [ 50 100] for efficiency, and [0 10] for current, and you should see no losses if you draw 10 amps, and the losses should increase as a percentage of power as you increase the resistance.
Thanks a lot for this amazing tutorial. Are you working at Mathworks? Otherwise, do you have any blog/website where there is more materials about simscape/electronics?
The most likely culprit would be that the dc voltage source is set to 0V, or that the PWM signal isn't actually turning on the Switch (check if it is correct value, and that it is connected). Other than that, a connection error is possible. MathWorks provides working examples that you can also work off of: www.mathworks.com/help/physmod/sps/ug/buck-converter_example-ee_switching_power_supply.html
Nice work! I use Simulink heavily for EV powertrain design and modeled our HV DCDC converter exactly the way you described. But like you described there's a lot of output ripple. What other methods of control loop other than a larger capacitor will minimize the current ripple? I heard there's peak current control which I haven't looked at yet.
For a given load and a specific topology, ripple is primarily controlled by the sizing of your components and switching frequency or switching approach (not all approaches have a fixed switching frequency). You can also put converts in a parallel with phase shifted carrier signals for PWM. Different control approaches can be used to improve harmonics, but at steady state, if operating at the same switching frequency, ripple will be the same across control methods.
I'm looking into this, they released an example here: www.mathworks.com/help/physmod/sps/powersys/ref/cccvbatterycharger.html but there are no files associated with it. It's a good document to explain the nuances, but doesn't provide a good starting point. This is a file exchange entry that has decent reviews that tries to implement it: www.mathworks.com/matlabcentral/fileexchange/72570-lithium-battery-charger-block?s_tid=srchtitle. The best way to do it is to model a converter that changes control modes from voltage to current, but this requires designing two closed loop control systems and coordinating the transition between them, which isn't an easy starting point. To do something simple, you could have a controlled voltage source behind an inductance in parallel with a controlled current source. Put an ideal switch on the output of each one and switch between which you want to use (there will likely be an artifact at the cross over point unless you add some snubbers/tune circuit paramters to clean it up.)
Unfortunately that is a pretty common error. It usually means the model as set up isn't realistic, but typically requires someone to look at the model and debug it to see what is wrong.
Hello. Who know how can i model diesel-generator (synchronous generator driven by asynchronous diesel motor) or where can i find tutorial to model electrical machines.P.S. haven't time to learn it from matlab documentation because of my bad technical english.
Here is an example: www.mathworks.com/help/physmod/sps/examples/emergency-diesel-generator-and-asynchronous-motor.html Unfortunately, I can only point you to english resources online for this topic. You could try to find a local IEEE groups or other communities that can support you in your native language. I'm assuming you speak Russian, and unfortunately I haven't found any publicly available documentation from standard power system analysis tools in Russian. However, youtube has some intro videos in Russian. Here is a Russian video on electric machines. I assume it is good based on the images, but the auto translate told me he was wrapping the magnets with chickens (should have been wires?) so I can't speak to it's quality... th-cam.com/video/zNdwN3rNkX8/w-d-xo.html
This is the best video about matlab for power electronics that i have seen It is much better than the webinars from mathwor. Perfect balance to make a model to work. The webinars from matwork focuse too much in other aspecst and very little about the simscap issues For example; one webinar that uses the sepic comverter focused so much in the power electronics concepts and very little about the simscap solver
please do you have an idea about how to creat a block of loss in simcape for induction motor
i have been using the black model for the longest time and it has some limitations and I didn't know how to operate the simscape model, and the trick was to just convert the signal !!!!
thank you so much
hey bro please do you have an idea about how to creat a block of loss in simulink
I am developing a custom power converter. I tried using Simscape to simulate it, but the results were completely nonsensical. This video provided invaluable information about the appropriate solver type that should be used for this type of task, along with a lot of other useful information. Thank you so much!
This is a great tutorial. You should definitely do more.
Please upload more! Great step by step introduction combining simulink and simscape components. *Subscribing*
You are a king for making this video!
Great video! May have just saved my semester lol. I would be curious to know how to do some of the more in depth options that you kind of skipped over for the sake of simplicity. You should definitely make some more tutorials.
Lovely Tutorial. Thanks a lot. Expecting more from you.
I was having headache with the simulink to PS signal connection..
This saves my life..TT
Great video waiting for more
This is a reallyy good video! Thank you so much
I wish you have more videos like this
Oh I see, your switching the transistors to control charge of capacitor to achieve desired voltage, I’m guessing that’s what the voltage control mechanism involves.
I like this tutorial because of the emphasis on practical applications regarding the use of simscape to produce some predicted behaviors. This tutorial seems to convey the significance of simscape in that it’s an implicit modeling tool that’s more intuitive compared to simulink as an explicit modeling tool. As a result of feedback in simulink the need to avoid algebraic loop and so introduce an integrator in low pass filter.
I guess you choose to impliment the control portion in simulink because there are more blocks to choose from.
I say that because algebraic loops are avoided if the feedback is implemented with simscape physical signal blocks
You are correct. One rule of thumb I use to simplify the decision is this. If you are going to implement your controls in a digital/embedded system that is clocked, use Simulink, and break the loop with a unit delay representing the controller update rate. If you are implementing it analog, leave everything in simscape (granted, it's easier to tune the controller in Simulink). However, if you want to go to Simulink for the loop design, there is a trick to getting past the algebraic loop although it's a little bit of a hack. For instance, if you want to close the loop on the inductor's current, you can measure it's voltage with a sensor in Simscape and pass that signal through an integrator in Simulink, divided by the inductance. This will tell you the current very very very closely to what simscape will calculate, and having the integrator in the Simulink portion of the model breaks the loop. This would cause problems if the simulation was run for a very very long time, but the error is so small it takes a while to accumulate and is very useful for tuning gains and using the controls toolboxes for initial loop design before going back to the hardware design for implementation.
I have seen using black resistance as load in some tutorials but I couldn’t get them in my libraries and the boundary connectors are not proper there
All other part of my circuit are simulink modules(black coloured)
Any solution or correction do you have
the black circuit elements are part of a sub library called specialized power systems. They use a different technology implementation and are not compatible with standard simscape components. There are ways to connect them, but generally make the simulation slow or unstable. Most times, there is a direct equivalent block in the simscape library that you can use instead.
I suspect the black mosfet is just to provide more of a physical interpretation to standard simulink blocks.
While simscape is true implicit modeling with predictions for both current and voltage and so power delivery occurring at the terminals
So I guess the black components are masked subsystems to be more representative of actual electrical components
great video, we need more.
I have a question that I don't see the answer to anywhere on the internet. Here it is:
I am trying to compare a DC and AC microgrid in terms of efficiency. In Simscape there are blocks with a converter which only require that you set the efficiency based on output current.. For some reason I am unable to get this right. The way I see it. when I put a variable resistor on the output and change its resistance, I should get as a current measurement the current minus the efficiency percentage I have put. But I don't get that. I input [0,10,20,30,40,50,60,70,80,90,100] as an efficiency vector for [0,10,20,30,40,50,60,70,80,90,100] as output currents. SO I expect that for 100volts output and a 20 ohm resistance that I get 50% efficiency which should give 1/2 of 5 which is 2.5.
If you have 100 Volts out at 20 ohms, you will have 5 amps, based on your vector, 5 amps = 5% efficiency, so you should get almost nothing out of the converter. Try using a vector like this [ 50 100] for efficiency, and [0 10] for current, and you should see no losses if you draw 10 amps, and the losses should increase as a percentage of power as you increase the resistance.
An extremely useful video, much appreciated! Waiting for more :)
(P.S. liked and subscribed)
Brilliant tutorial. Fair play to you.
It would be nice if you implement the buck converter in current mode control . Thank you
Please make more videos on electromechanical system like Horn of bike, magneto assembly etc.. We will be very grateful to you please🙏🙏🙏
Thanks a lot for this amazing tutorial.
Are you working at Mathworks?
Otherwise, do you have any blog/website where there is more materials about simscape/electronics?
Hi, I follow your work but I cannot get the same result as yours. it just almost a straight line in the scope. Please advise
The most likely culprit would be that the dc voltage source is set to 0V, or that the PWM signal isn't actually turning on the Switch (check if it is correct value, and that it is connected). Other than that, a connection error is possible. MathWorks provides working examples that you can also work off of: www.mathworks.com/help/physmod/sps/ug/buck-converter_example-ee_switching_power_supply.html
really cool tutorial, great work
what about code generation for this dc-dc converter? can we generate code?
Great tutorial! Thank you!
Nice work! I use Simulink heavily for EV powertrain design and modeled our HV DCDC converter exactly the way you described. But like you described there's a lot of output ripple. What other methods of control loop other than a larger capacitor will minimize the current ripple? I heard there's peak current control which I haven't looked at yet.
For a given load and a specific topology, ripple is primarily controlled by the sizing of your components and switching frequency or switching approach (not all approaches have a fixed switching frequency). You can also put converts in a parallel with phase shifted carrier signals for PWM. Different control approaches can be used to improve harmonics, but at steady state, if operating at the same switching frequency, ripple will be the same across control methods.
Is it possible for you to create such a video for a 3-phase power system project?
Thanks for the great explanation, I wish to know how to simulate constant voktage constant current charging of a battery using simscape . Thanks
I'm looking into this, they released an example here: www.mathworks.com/help/physmod/sps/powersys/ref/cccvbatterycharger.html but there are no files associated with it. It's a good document to explain the nuances, but doesn't provide a good starting point. This is a file exchange entry that has decent reviews that tries to implement it: www.mathworks.com/matlabcentral/fileexchange/72570-lithium-battery-charger-block?s_tid=srchtitle.
The best way to do it is to model a converter that changes control modes from voltage to current, but this requires designing two closed loop control systems and coordinating the transition between them, which isn't an easy starting point. To do something simple, you could have a controlled voltage source behind an inductance in parallel with a controlled current source. Put an ideal switch on the output of each one and switch between which you want to use (there will likely be an artifact at the cross over point unless you add some snubbers/tune circuit paramters to clean it up.)
@@electrificationenthusiasts8611 please i wanna know how to creat a block of loss in simcape for an induction motor
Very good tutorial, thank you!
Well done. Comparing to LTspice. Can you do C code blocks in Simscape Electrical?
You can bring c-code blocks into Simulink, you would need to interface them to the simscape network through controlled current and voltage sources.
excellent explanation, thank you
You are welcome!
very good and pedagocial video....
Excellent.
Brother how to take dc motor and centrifugal pump output in simscape
To get simscape outputs, you use sensor blocks or the probe under simscape utilities.
Please make more tutorial wrt Power systems and power electronics
Keep up the good content 👍
Sir how to eliminate the time 0.00not finite in matlab simulink please help us sir
Unfortunately that is a pretty common error. It usually means the model as set up isn't realistic, but typically requires someone to look at the model and debug it to see what is wrong.
Hello. Who know how can i model diesel-generator (synchronous generator driven by asynchronous diesel motor) or where can i find tutorial to model electrical machines.P.S. haven't time to learn it from matlab documentation because of my bad technical english.
Here is an example: www.mathworks.com/help/physmod/sps/examples/emergency-diesel-generator-and-asynchronous-motor.html
Unfortunately, I can only point you to english resources online for this topic. You could try to find a local IEEE groups or other communities that can support you in your native language. I'm assuming you speak Russian, and unfortunately I haven't found any publicly available documentation from standard power system analysis tools in Russian.
However, youtube has some intro videos in Russian. Here is a Russian video on electric machines. I assume it is good based on the images, but the auto translate told me he was wrapping the magnets with chickens (should have been wires?) so I can't speak to it's quality... th-cam.com/video/zNdwN3rNkX8/w-d-xo.html
Great video.
Keep it up man !
man thank you for the right click tipp. I was constantly deleting routes to get a new wire to click -.-
I switched to ode23t and jeez it's so much faster. Thank you really I appreciate it
Great Video
thank you
You're welcome
Thanks a lot, nice
Thanks a lot
Dude literally save my ass orz