3

I would like to simulate the ZVS Synchronous rectification buck converter, I know if I operate in ZVS I need to change the switching frequency of the buck converter.

Here are my questions:

  1. How to model the variable frequency of buck converter? I think the small-signal model is not the same as the constant frequency buck converter.

  2. How to implement the variable frequency control in PSIM or SIMPLIS? I would like to know how to build the control stage, I only can design the control stage in constant frequency.

  3. Do you have some material you can suggest to me to read?

EEC
  • 403
  • 3
  • 10
Jitter456
  • 555
  • 2
  • 7

1 Answers1

6

To model the buck operated in borderline or boundary conduction mode (BCM), you need to resort to a specific PWM switch model that I derived some years in my book. From that model, you can linearize it and determine the control-to-output transfer function of the BCM buck converter. That is what I did in my newly-released book dedicated to small-signal modeling of switching converters:

enter image description here

If you derive the transfer function, you'll find a simple 1st-order response which is easy to stabilize.

I have released a set of free SIMPLIS simulations templates described here and one of them is a BCM CM buck converter:

enter image description here

It works on the free demonstration version and you have the results in a few fractions of seconds. As you can see below, the part operates in second valley and turns the switch on at almost 0 V.

enter image description here

Verbal Kint
  • 20,420
  • 1
  • 16
  • 50
  • Wow .We should chat sometime +1 – Autistic May 12 '21 at 23:19
  • Thanks for your answer, in your answer, why you said the small-signal transfer function is a control to output? if the frequency will change why we don't use frequency to output transfer function? – Jitter456 May 13 '21 at 01:37
  • 1
    This is because this circuit is a self-relaxing system: there is no clock. The control voltage sets a peak current value and defines the on-time based on operating conditions. Then, the switch turns off and the inductor demagnetizes; its current ramps down. When it hits 0 A, the switch turns back on after a certain delay to be right in the minimum voltage. In this simulation, the integration time constant on the demag. path is a bit too large and it misses the first valley. – Verbal Kint May 13 '21 at 07:55
  • My application is using synchronous rectification buck converter, so the inductor current will become negative when it operates in the light load, this we called Triangular Current Mode, and we can change the frequency to reach the ZVS. but your example is not the same as mine. Do you have some information about this? – Jitter456 May 14 '21 at 01:54
  • 1
    If you operate a buck synchronously, then it remains in CCM even in a no-load situatation. If you then want to operate the upper-side switch in ZVS, you must resort to a different control scheme with additional components. This [thesis](https://dspace.mit.edu/bitstream/handle/1721.1/113122/1016458075-MIT.pdf?sequence=1) or this [document](https://www.power-mag.com/pdf/feature_pdf/1447860008_Vicor_feature_Layout_1.pdf) from Vicor show how to achieve this but there must be a lot of IP in this field. – Verbal Kint May 14 '21 at 07:24
  • If I want to operate the upper-side switch in ZVS why I can just change the frequency. if the dead time is enough the upper-side switch will go into ZVS, if the dead time is too large, when current still go into ZVS but it will also go through the body diode, so it will generate the conduction loss, I want to know how to use variable frequency control to the syn-buck converter. – Jitter456 May 14 '21 at 15:25
  • I think I don't need to add other components in my circuit, I just need to adjust the dead time and then I can get the ZVS for all load condition, it seems like the LLC control (VCO), but I am not sure if the buck converter can do the same thing. – Jitter456 May 14 '21 at 15:26