An Easy Explanation of Subharmonic Oscillations & Slope Compensation in Current Mode Power Supplies

[@W1RMD]

Excellent top notch quality training sir! Thank you for sharing this with us. It's like going to MIT....for FREE!

[@frankcole3196]

Excellent presentation! Excellent explanation!

[@hosseinpirhady8045]

Another great video. Thank you for all the effort and sharing 🙏🙏🙏Waiting for the next episode 👍

[@crowderglen]

Thank you for the video! Your explanation was great.

[@Graham_Wideman]

I appreciate the effort that goes into videos like this, and also the desire to tell the story clearly. I'm an EE, I know a bit about switching power supply circuits. and unfortunately I don't think this video achieves its goals. The main problem is that the explanation is unmoored from even a generic diagram of the circuit to which it pertains. The presenter identifies all the signals under discussion just verbally and rather vaguely, so it's often hard to tell what is actually being discussed. For example, around 3:30 Iref is referred to as "our demand current value". Wat does that mean? This is surely a power supply producing an output voltage, (that's the demanded value), in which there's feedback that the controller uses to set the peak current of the ramp for the next cycle. I guess Iref is that peak value? But that value will change from cycle to cycle, so why is it referred to as "Iref" as though it's a fixed reference current? Either I'm unfamiliar with the terminology or I completely don't understand the circuit under examination. Later there's a comment that the subharmonic oscillation will greatly increase ripple current. Well I suppose that's just describing the alternating short and long current ramps. But what is the significance -- is the concern about disrupting the upstream supply? Or disturbing the downstream circuity, where I would have thought that voltage ripple would be a more immediate concern? Here again, being able to point to the location in the circuit that the issue applies to would be very helpful. I was very impressed with the nice animated graphs at 10:39 ... except I didn't understand what they display. What are the axes, and what are the signals plotted in red blue and green, referred to points in a schematic that I could recognize? At the point where discussion moved on to some kind of synthetic slope I was basically lost. I thought the sloped signal was the inductor current ramp, something tangible. So I didn't see how that could somehow be made synthetic. Maybe this is trying to illustrate the signals that the controller pays attention to, and the enhancement is to revise this controller input variable by adding in an additional signal or calculation. I'm not at all sure from the verbal discussion, but would have been if the graphed variables were identified with points in a schematic. (Possibly with the schematic enhanced to show stages of the controller's internal algorithm, if that's the story.) So again, I appreciate the effort, but for me, who I think might be a representative target audience, this flew over, or at least around, my head! A schematic or two to identify signals please!

[@Graham_Wideman]

Just to add --- I did see that there's a PDF with more detailed notes. I will download and read that. So no need to answer the questions I raised per se. This is more a suggestion about ways to make future videos more followable.

[@ThePyrosirys]

In voltage mode control, it's true that the demand value is the output voltage. The control loop takes in the voltage reference as an input and directly created the duty cycle of the switch as an output. This is also the case for current mode control, because at the end of the day, you want a fixed output voltage. However, there is also a reference value of the current, but the current reference value is not fixed by the design, it's just an intermediate value which is generated by the circuit. One way to think of it is that you actually have two different control loops, an inner control loop and an outer control loop. The outer loop takes the voltage reference as an input and generates the current reference as an output. The inner loop takes the current reference as the input and generates the duty cycle as the output. When you analyze the inner loop by itself, it can be correct to think of the behavior for a single switching period. During a single period, Iref will basically be a constant, since the control loops operate at frequencies lower than the switching frequency.

[@Graham_Wideman]

@@ThePyrosirys Thanks for your comment, I think it was helpful. In your final sentence, I take you to mean that the outer (voltage) control loop operates at a frequency lower than the switching frequency. While the current control loop operates within a single switching cycle. Is that what you mean?

[@ThePyrosirys]

​@@Graham_Wideman Yes exactly. Since the inner loop controls peak current, it has to act on a cycle to cycle basis. Typically the current compensator just has a proportional gain to convert the measured current to the appropriate voltage range. There is no low pass filter, since that would only prevent the control from actually detecting the peak current.