0

I have this really simple buck converter circuit but I am getting really funny results.

Buck Converter Schematic

Simulation results

I tried different input voltages (12-650 V), no matter what I put, it doesn't affect the output. The output voltage amplitude changes with the change in PWM voltage amplitude. Changing the duty cycle doesn't change the Vout amplitude either. I tried different MOSFETS as well which also included SiCs, IGBTs, etc.

Been stuck at this for a few days now, any help would be greatly appreciated. Thank you.

Some edits after all the helpful comments. I am using the Boostrap circuit as @hacktastical suggested. This is my circuit now. enter image description here

The only problem is I am getting 0.6V ripple, are there any tips on dropping it down to 0.1V? enter image description here

Saagar
  • 25
  • 7
  • You can't use a N channel FET as a high side switch like that. You need a P channel FET or some bootstrapping system for the N channel. – Unimportant Jul 28 '21 at 01:48
  • Your question is also missing all the important details of components and the design specs. – Tony Stewart EE75 Jul 28 '21 at 03:08
  • @Unimportant, we can use both for a Buck converter. Textbooks might show P channel cuz it is simpler, manufacturers use N channel because it is cheaper. – Saagar Jul 28 '21 at 12:49
  • @TonyStewartEE75 I believe I have all the important specs on my schematic. And the design specs are well, I just wanna learn. Let me know if there is anything specific that you wanted. – Saagar Jul 28 '21 at 12:50
  • If you read and understand any Power convert simulation tool you will learn the importance of ESR for caps , DCR for coils, Ron for FETs and diodes, and ultimately Q from the series reactance ratio for damping the resonance or instability during a step load start or change. This affects the startup power to load power ratio greatly and overshoot on voltage. – Tony Stewart EE75 Jul 28 '21 at 14:41
  • @Saagar No, you have a source follower now. – Unimportant Jul 28 '21 at 14:42
  • V1 must be > Vout + 2.5 Vt but not exceed 15v + Vt in a follower. This won’t work. [Edited by a moderator] – Tony Stewart EE75 Jul 28 '21 at 14:44
  • All - Please keep things civil and remember the Code of Conduct. Polite technical disagreement is allowed. However, don't post anything that could feel personal, even when applied to content and not the person. One or more comments have been deleted. Thanks. – SamGibson Jul 29 '21 at 15:01

3 Answers3

1

To use an n-FET on the high side, you will need to bring the gate voltage above the switch node voltage to turn the FET on. Bear in mind, the switch node normally swings between GND and Vin, so you have to have at least Vin + Vgs threshold at the gate to get the job done.

The typical way to do that is to use a boostrap circuit to develop this higher-than-max voltage. Below is a DCDC that uses a bootstrap (simulate it here)

enter image description here

The scope trace on the left is the gate drive for the n-FET. You'll notice that it swings between 0 and 2x Vin (24V in case). This ensures the FET turns on fully with minimum drain-source resistance.

hacktastical
  • 53,912
  • 2
  • 49
  • 152
  • thanks a lot, @hacktastical, I used bootstrap circuit as you suggested. I will try adding the circuit in the question. I can now convert the 400 V to 8V, the only thing is my ripple is 5V, I might be using the wrong equations. Are there different equations to calculate the inductor and capacitor values for HV? – Saagar Jul 29 '21 at 14:47
  • The challenge with such a high stepping ratio is the minimum on time for the high side switch will be very short. Suggest using 2 or 3 stages to step it down. – hacktastical Jul 29 '21 at 23:01
  • That said, a step down from 600V would be better accomplished using a flyback (transformer) based topology. – hacktastical Jul 31 '21 at 15:32
  • You also need to consider limiting the max Vgs to a safe value. – hacktastical Jul 31 '21 at 16:52
0

You have M1 wired as a source follower: -

enter image description here

This means that the voltage on the source is largely unaffected by the input voltage on the drain and this is what you see. Try using a P channel MOSFET instead or use a gate-source drive voltage that can operate M1 as a proper switching transistor.

Andy aka
  • 456,226
  • 28
  • 367
  • 807
  • "Try using a P channel MOSFET instead or use a gate-source drive voltage that can operate M1 as a proper switching transistor" The P-FET would also need a "gate-source drive". – Prathik Prashanth Jul 29 '21 at 04:49
0

If you are trying to drive a MOSFET that is switching 600V, I recommend using an isolated gate drive. You could use a bootstrap driver like @hacktastical suggested, but it probably makes more sense to use an isolated gate drive instead, something like this:

enter image description here

I have simulated this circuit on everycircuit. You can take a look at it.

enter image description here

And here is an application of the isolated gate drive circuit in a half-bridge inverter:

enter image description here

Prathik Prashanth
  • 1,016
  • 4
  • 17