Absolutely amazing, thank you!. Analog Devices / Linear Tech should kidnap you and have you do all their tutorials & content creation. The channel, The Signal Path was my ultimate favorite channel, now I have two ultimate favorite channels , thanks again!!!
I've never been kidnapped, it might be an interesting experience, but a free program deserves free promotion - I'm just glad I can share what I know and bring useful input about a great program to other people.
Great video as always. For the viewers out there that are in need of a little fun I propose a drinking game. Each time Fesz is saying the word 'so' you drink. I wonder how many of you can last until the end of the video :)))
Depending on what you want to simulate, you will need a more or less complex model; If you use the complete version in a more complex circuit, your simulation might end up going quite slowly.. Anyway let me know how accurate the simulation ends being compared to the real thing!
This is brilliant! I find your knowledge of LTSpice and ability to engineer reality into it astounding. I'm not sure I would call it simple though, it looks pretty complex to me! But I suspect that you may have made an invalid assumption:- You said there is a moment when no coil is connected as the brushes cross over the small gap in the commutator. But in reality I expect that for that short moment, there are TWO coils connected to the supply, as the brush touches both contacts. Imagine a larger motor with carbon brushes, which are much wider than the gap between contacts, where this would be even more pronounced. Do you agree? Does this affect your model? Can you/have you considered folding the complexity of this model into a simply used general component that can be dropped into any LTSpice circuit? Maybe that is so simple as to not be worth mentioning, but I certainly have no clue about how to do it. I found this video because I wanted to play with simulating control of universal motors (eg: electric drill). In that case, the permanent magnets are replaced by field coils in the stator, which are in series with the rotor and brushes. I imagine that adding that aspect to the model would be easy (for you, not me!).
Hello Bertoid! I think you are right! 2 coils should be connected if the brush contact area is wider than the gap, I will try to look for the motors I used in the video just to check, and let you know. Though if that happens, its a bit more difficult to get the inductive spikes, I think I need to study the problem a bit more deep. Regarding simplifying the model, you can pack the entire thing into a single component and that makes it more "user friendly". I tried to cover that process at the end of a different video - ua-cam.com/video/0smuo8bMRn4/v-deo.html (here I make another monstrosity, and turn it into a library component) Regarding the electric drill, and modeling components in general, it highly depends on what "features" of the real thing you want to simulate. It can be as simple as a resistor to just waste some watts of power, or some monstrosity like I tried to create which simulates it down to the rotation speed and complete electrical noise. Considering that the motor is like this one with some extra fixed winding's for the stator electromagnets you can take a model for a clasical brushed motor, and just add a couple coils to simulate the electromagnets. Those coils will come with the inductance and series resistance of the real thing. The rest (rotor + brushes) is just like a simple permanent magnet motor.
Thanks for the reply. I was hoping that you might have done a video on making a library component - I'll take a look at that. I have to re-watch this video to better understand a couple of things:- 1/ How the inductive spikes were coupled back into the supply lines. As this will happen when the brush leaves the contact, there is no longer a connection into the supply? (apart from arcing, and maybe some capacitive & inductive coupling) And isn't this the same problem whether no coils or two coils were are connected - always one coil will be de-energized as the brush leaves it, causing the spike? 2/ How the permanent magnet's field was introduced into the model I can see how adding two inductors for the field coils in a universal motor is fairly straightforward, but won't the fields from them (which will vary with current) have to be coupled into the model, in place of the permanent magnet? And I imagine it all gets quite a bit more hairy when the supply is AC!
Hello Bertoid, On the first topic - even if the contact jumps from one coil to the other, or both are touched at the same time, the spike will appear when the disconnection from a charged inductor occurs. When the disconnection appears, the current previously going trough the first coil doesn't dissapear (the inductor stores current the same way the capacitor stores voltage), it turns into a very high voltage that causes the arc - this voltage is fed back into the supply line. The second coil that is already connected to the contact is also an inductor so it will not accept a fast current rise therefore it does not absorb the high voltage arc. The model I made reproduces this effect, it just doesn't have the second coil. On the second topic - the permanent magnets are not directly modeled in my example - these would be what cause the voltage source in series with the inductor and resistor to create the induced voltage (the EMF). I think that you can get the model to work with AC (turn the same way regardless of voltage polarity) by multiplying the voltage in the EMF source and the one in the mechanical circuit with the direction of the current trough the motor (*1 if its positive or *-1 if its negative). Also to remove the "generator" behavior, the voltage source that links the internal mechanics to the output needs to be multiplied with 0 if there is no current on the input. This way it shouldn't start generating current when rotated. When I get a chance I will try to edit the model to reproduce the AC motor also
Nice explanation. I was wondering about simulating actually the three coils each with it's own EMF loop. What do you think about this approach? Any advises against it (ex: simulation convergency...speed, etc...) The reason for this is that I would like to have in the simulation the amount of dV/dt generated by te coil commutation and how it affects the coils adjacent to it. I mean, I guess there is some energy passed by one coil to the other. Would like to have your comments on this. Thanks again for the nice video.
Hello Jorge! Unfortunately I don't intend to spend more time modeling motors. At the moment I have quite a few videos planned ahead, but who knows maybe some time in the distant future.
Hi, I think there is a small errror in schematic: Source B11 should be I(B7) and label B10 should be changed to B11. Then B14, B15, B16 should point to label B11
Honestly I don't think this is information is present in the datasheet, since its of no use during normal design. The best way to obtain it is to try to measure it somehow...
What a great presentation. I have been looking for a good motor model recently and yours is the best I've found. Would it be possible to send me the.asc file to try? Thanks
Thank you for a very well explained simulation of DC motors. What is the motor inductance and resistance that matched your motor? I like to estimate the typical L/R time constant of DC motors. I use a DC motor with 3 rotor poles as tachogenerator for a DIY servo drive of a sewing machine. So I can confirm, this waveform. However, because of inaccuracies of small motors, the brushes or rotor system may have turned a bit, so you do not get the perfect rectification of the sine waves. It can be a bit difficult to measure shaft speed fast at low speeds with rotary encoders. Then another DC motor is a nice and cheap alternative to encoders. However I think you will chose a motor with a rotor of 5 poles - you can get such small DC-motors cheep as well. This is my setup on the sewing machine: ua-cam.com/video/075OH2rgel0/v-deo.html
I don't remember measuring the exact inductance - I think I just used a value that made the model behave similarly in the simulation as in real life; and for the resistance, I seem to have 7.1R; I did this video quite long ago, I unfortunately don't remember all the details
The best ltspice tutorial playlist as far as I know❤️
Absolutely amazing, thank you!. Analog Devices / Linear Tech should kidnap you and have you do all their tutorials & content creation. The channel, The Signal Path was my ultimate favorite channel, now I have two ultimate favorite channels , thanks again!!!
I've never been kidnapped, it might be an interesting experience, but a free program deserves free promotion - I'm just glad I can share what I know and bring useful input about a great program to other people.
The guy with an outstanding skills on LTSpice!
Please keep making these videos, these are gold.
It's great the you've shared the model. Very useful, thank you.
Thank you for sharing this work. Well done!
Great video as always.
For the viewers out there that are in need of a little fun I propose a drinking game. Each time Fesz is saying the word 'so' you drink. I wonder how many of you can last until the end of the video :)))
I wasn't really aware of how much I do this, but I will try to make it 100 at one point in a single video :D
Another great video, I really wanted to do this and I wasn't sure where to begin. Thanks a ton!
Depending on what you want to simulate, you will need a more or less complex model; If you use the complete version in a more complex circuit, your simulation might end up going quite slowly.. Anyway let me know how accurate the simulation ends being compared to the real thing!
multumim prietene foarte detaliat si precis.
Ma bucur ca ti-a placut! Mersi mult!
This is brilliant! I find your knowledge of LTSpice and ability to engineer reality into it astounding.
I'm not sure I would call it simple though, it looks pretty complex to me!
But I suspect that you may have made an invalid assumption:-
You said there is a moment when no coil is connected as the brushes cross over the small gap in the commutator. But in reality I expect that for that short moment, there are TWO coils connected to the supply, as the brush touches both contacts. Imagine a larger motor with carbon brushes, which are much wider than the gap between contacts, where this would be even more pronounced.
Do you agree? Does this affect your model?
Can you/have you considered folding the complexity of this model into a simply used general component that can be dropped into any LTSpice circuit?
Maybe that is so simple as to not be worth mentioning, but I certainly have no clue about how to do it.
I found this video because I wanted to play with simulating control of universal motors (eg: electric drill). In that case, the permanent magnets are replaced by field coils in the stator, which are in series with the rotor and brushes. I imagine that adding that aspect to the model would be easy (for you, not me!).
Hello Bertoid!
I think you are right! 2 coils should be connected if the brush contact area is wider than the gap, I will try to look for the motors I used in the video just to check, and let you know. Though if that happens, its a bit more difficult to get the inductive spikes, I think I need to study the problem a bit more deep.
Regarding simplifying the model, you can pack the entire thing into a single component and that makes it more "user friendly". I tried to cover that process at the end of a different video - ua-cam.com/video/0smuo8bMRn4/v-deo.html (here I make another monstrosity, and turn it into a library component)
Regarding the electric drill, and modeling components in general, it highly depends on what "features" of the real thing you want to simulate. It can be as simple as a resistor to just waste some watts of power, or some monstrosity like I tried to create which simulates it down to the rotation speed and complete electrical noise. Considering that the motor is like this one with some extra fixed winding's for the stator electromagnets you can take a model for a clasical brushed motor, and just add a couple coils to simulate the electromagnets. Those coils will come with the inductance and series resistance of the real thing. The rest (rotor + brushes) is just like a simple permanent magnet motor.
Thanks for the reply. I was hoping that you might have done a video on making a library component - I'll take a look at that.
I have to re-watch this video to better understand a couple of things:-
1/ How the inductive spikes were coupled back into the supply lines.
As this will happen when the brush leaves the contact, there is no longer a connection into the supply? (apart from arcing, and maybe some capacitive & inductive coupling)
And isn't this the same problem whether no coils or two coils were are connected - always one coil will be de-energized as the brush leaves it, causing the spike?
2/ How the permanent magnet's field was introduced into the model
I can see how adding two inductors for the field coils in a universal motor is fairly straightforward, but won't the fields from them (which will vary with current) have to be coupled into the model, in place of the permanent magnet?
And I imagine it all gets quite a bit more hairy when the supply is AC!
Hello Bertoid,
On the first topic - even if the contact jumps from one coil to the other, or both are touched at the same time, the spike will appear when the disconnection from a charged inductor occurs. When the disconnection appears, the current previously going trough the first coil doesn't dissapear (the inductor stores current the same way the capacitor stores voltage), it turns into a very high voltage that causes the arc - this voltage is fed back into the supply line. The second coil that is already connected to the contact is also an inductor so it will not accept a fast current rise therefore it does not absorb the high voltage arc. The model I made reproduces this effect, it just doesn't have the second coil.
On the second topic - the permanent magnets are not directly modeled in my example - these would be what cause the voltage source in series with the inductor and resistor to create the induced voltage (the EMF). I think that you can get the model to work with AC (turn the same way regardless of voltage polarity) by multiplying the voltage in the EMF source and the one in the mechanical circuit with the direction of the current trough the motor (*1 if its positive or *-1 if its negative). Also to remove the "generator" behavior, the voltage source that links the internal mechanics to the output needs to be multiplied with 0 if there is no current on the input. This way it shouldn't start generating current when rotated.
When I get a chance I will try to edit the model to reproduce the AC motor also
great video: simple and effective !
Nice explanation.
I was wondering about simulating actually the three coils each with it's own EMF loop. What do you think about this approach?
Any advises against it (ex: simulation convergency...speed, etc...)
The reason for this is that I would like to have in the simulation the amount of dV/dt generated by te coil commutation and how it affects the coils adjacent to it. I mean, I guess there is some energy passed by one coil to the other.
Would like to have your comments on this.
Thanks again for the nice video.
This was a great video. Well done, and thank you. Any chance you might work through a BLDC motor model?
Hello Jorge! Unfortunately I don't intend to spend more time modeling motors. At the moment I have quite a few videos planned ahead, but who knows maybe some time in the distant future.
Hello, The video is amazing, Could you please explain how to make a model for BLDC motor
Hi, I think there is a small errror in schematic: Source B11 should be I(B7) and label B10 should be changed to B11. Then B14, B15, B16 should point to label B11
can I use the timer 555 to replace your VCO circuit ? Is it possible ? Because I found it easy to implement and more understandable.
hi FesZ I could not download your model on google drive.. :(
Hi, how can one find out the motor inductance and the motor inertial from the motor data sheet? Any tips? Thanks.
Honestly I don't think this is information is present in the datasheet, since its of no use during normal design. The best way to obtain it is to try to measure it somehow...
What a great presentation. I have been looking for a good motor model recently and yours is the best I've found. Would it be possible to send me the.asc file to try? Thanks
There is a link in the description to some of the spice files used in the video.
How easily can a BLDC motor be modeled !?
Thank you for a very well explained simulation of DC motors. What is the motor inductance and resistance that matched your motor? I like to estimate the typical L/R time constant of DC motors.
I use a DC motor with 3 rotor poles as tachogenerator for a DIY servo drive of a sewing machine. So I can confirm, this waveform. However, because of inaccuracies of small motors, the brushes or rotor system may have turned a bit, so you do not get the perfect rectification of the sine waves. It can be a bit difficult to measure shaft speed fast at low speeds with rotary encoders. Then another DC motor is a nice and cheap alternative to encoders. However I think you will chose a motor with a rotor of 5 poles - you can get such small DC-motors cheep as well. This is my setup on the sewing machine: ua-cam.com/video/075OH2rgel0/v-deo.html
I don't remember measuring the exact inductance - I think I just used a value that made the model behave similarly in the simulation as in real life; and for the resistance, I seem to have 7.1R; I did this video quite long ago, I unfortunately don't remember all the details