This is by far, the best presentation of the analysis and design of a CCM Flyback SMPS. I am in the middle of a design cycle and this video is indispensable. Thanks for the wonderful contribution, Eleazar and WE.
Does anyone know the URL to the presentation PDF or PPTX? I would like to be able to read some of the symbols and text, which is un-readable at 480kpbs video. Thanks.
Thank you for the question. The right-half-plane zero (RHPZ) is not created by any component as such, but by the way the energy is transferred from the input to the output stage of the converter during each switching cycle, so it depends on the converter topology. Especially, topologies in which energy transfer to the load only occurs during the off-time of the control transistor, like the boost, buck-boost and flyback converters, have a RHPZ on its control-to-output transfer function, whereas topologies in which there is energy transfer during the entire swiching period (Tsw), like the buck and forward converters, do not have a RHPZ. What the RHPZ ‘physically’ means can be intuitively understood as follows. Consider a sudden increase step in the load current of a flyback converter. The output voltage will start to decrease and the loop will react to increase the duty-cycle of the control transistor, so that to increase the primary peak current and with it the amount of energy transferred to the output in order to bring the output voltage back up to its target level. However, as the duty-cycle has increased, the off-time of the control transistor (energy transfer window) is now shorter (toff=(1-D)*Tsw). So if the peak current has not increased high enough (e.g. because too high magnetizing inductance (Lm) slowing down dI/dt), then the average current on the secondary winding of the transformer, which equals the converter output current, would actually be lower than before, and the output voltage will drop even more. So the control loop has reacted trying to increase the output current, but the result is the opposite, due to either too high Lm and/or too fast control-loop reaction speed. If the magnetizing inductance is lower (i.e. higher dI/dt and peak current increase for a set delta of the duty-cycle), and/or the control loop is slowed down (i.e. lower crossover frequency resulting in a lower rate of change of the duty-cycle) then the issue can be solved. Note how, for a set increase of the duty-cycle, a lower Lm will allow the primary peak current amplitude (and thus secondary peak current) to increase high-enough each switching cycle so that to compensate for the shorter energy-transfer window, and thus make the average secondary current to effectively increase, which is what the control-loop is trying to achieve. This also intuitively explains why the RHPZ frequency depends on the inductance (see formula on the presentation): lower Lm (higher dI/dt), then higher RHPZ frequency, and higher the crossover frequency can be set without any issues. Hope this answers your question.
This is by far, the best presentation of the analysis and design of a CCM Flyback SMPS. I am in the middle of a design cycle and this video is indispensable. Thanks for the wonderful contribution, Eleazar and WE.
Glad it was helpful!
Does anyone know the URL to the presentation PDF or PPTX? I would like to be able to read some of the symbols and text, which is un-readable at 480kpbs video. Thanks.
Thank you for asking! You can download the PDF of the slides here- bit.ly/FeedbackLoopCompenstation_Aug2022
Can i use the concept that you mention like the equations if i want to control the output current in boundary conduction mode ?
Where does come from this right half plane zero on the plant transfert fonction ? What are the components creating this zero ?
Thank you for the question. The right-half-plane zero (RHPZ) is not created by any component as such, but by the way the energy is transferred from the input to the output stage of the converter during each switching cycle, so it depends on the converter topology. Especially, topologies in which energy transfer to the load only occurs during the off-time of the control transistor, like the boost, buck-boost and flyback converters, have a RHPZ on its control-to-output transfer function, whereas topologies in which there is energy transfer during the entire swiching period (Tsw), like the buck and forward converters, do not have a RHPZ.
What the RHPZ ‘physically’ means can be intuitively understood as follows. Consider a sudden increase step in the load current of a flyback converter. The output voltage will start to decrease and the loop will react to increase the duty-cycle of the control transistor, so that to increase the primary peak current and with it the amount of energy transferred to the output in order to bring the output voltage back up to its target level. However, as the duty-cycle has increased, the off-time of the control transistor (energy transfer window) is now shorter (toff=(1-D)*Tsw). So if the peak current has not increased high enough (e.g. because too high magnetizing inductance (Lm) slowing down dI/dt), then the average current on the secondary winding of the transformer, which equals the converter output current, would actually be lower than before, and the output voltage will drop even more. So the control loop has reacted trying to increase the output current, but the result is the opposite, due to either too high Lm and/or too fast control-loop reaction speed. If the magnetizing inductance is lower (i.e. higher dI/dt and peak current increase for a set delta of the duty-cycle), and/or the control loop is slowed down (i.e. lower crossover frequency resulting in a lower rate of change of the duty-cycle) then the issue can be solved. Note how, for a set increase of the duty-cycle, a lower Lm will allow the primary peak current amplitude (and thus secondary peak current) to increase high-enough each switching cycle so that to compensate for the shorter energy-transfer window, and thus make the average secondary current to effectively increase, which is what the control-loop is trying to achieve. This also intuitively explains why the RHPZ frequency depends on the inductance (see formula on the presentation): lower Lm (higher dI/dt), then higher RHPZ frequency, and higher the crossover frequency can be set without any issues. Hope this answers your question.
"Watts" the point of having a extremely loud und lang intro in front of low volume that sounds like its recorded with a can