0
\$\begingroup\$

A typical synchronous buck:

enter image description here

The normal switch node spike that people typically see when the high side FET turns on:

enter image description here

In the past i have been fairly negligible on this part. When the switch node spike occurs, Switch node voltage - VIN is always more than a typical body diode drop, so why doesn't the body diode of the high side FET conduct?

Thanks!

\$\endgroup\$
7
  • 1
    \$\begingroup\$ Who says it doesn't? \$\endgroup\$
    – Kevin White
    Apr 2, 2021 at 23:28
  • \$\begingroup\$ I am not sure and is hoping an expert can clarify it for me. If it does conduct, then VSW_NODE - VIN would just be a typical diode drop right. In the past, I have seen designs with VSW spike way higher than VIN. So VSW - VIN is way more than a typical diode drop (> 10x). \$\endgroup\$
    – helloguys
    Apr 2, 2021 at 23:31
  • \$\begingroup\$ @helloguys Are you sure that the ringing spikes seen on screen actually happen in the circuit? What if is just a sum of actual measurement, plus ground bounce, plus magnetic field and plus electric field affecting the measurement? You need to carefully document here how exactly you performed the measurements, scope and probe bandwidths, probe settings, compensation adjustment verification etc. \$\endgroup\$
    – Justme
    Apr 2, 2021 at 23:47
  • \$\begingroup\$ @Justme Hi sir, I didn't document this one, but these spikes are very common in a buck and there are all types of info on it. The spike occurs due to the parasitic inductance (for example, the large current loop) and the parasitic capacitance (such as Coss of the low side switch). So yes, it does exist. However, I can't find any info on the high side body diode behavior when this spike occurs so I am reaching out to some experts on this forum. Here is another reference by TI. ti.com/lit/an/slyt465/… \$\endgroup\$
    – helloguys
    Apr 3, 2021 at 0:04
  • \$\begingroup\$ @KevinWhite Hi sir, would you happen to know the high-side diode behavior? Thanks. \$\endgroup\$
    – helloguys
    Apr 3, 2021 at 0:05

2 Answers 2

Reset to default
1
\$\begingroup\$

I think I know why.

  1. When the high side FET is on, the body diode is effectively shorted by the channel, taking it out of the equation.

https://e2e.ti.com/support/power-management/f/power-management-forum/842123/csd18510q5b-body-diode-clamping-behavior-for-switch-node

  1. There are parasitic inductance even from the source of the high side FET to the switch node, hence the reason why VSW ring so high but there is no catastrophic damage from the VSW to VIN path. The path from VSW to VIN is more than just the RDSon of the high side FET.

enter image description here

Thanks and have a nice day.

\$\endgroup\$
0
\$\begingroup\$

"so why doesn't the body diode of the high side FET conduct"

Adding a fast response diode in parallel to high-side N-FET (Vds) would show, if this additional diode would change measurement output, thus answering this partly question towards answering if FET body diode is conducting during switching on, what's the response time and to what amount it limits VSW spikes.
(What is mentioned in questions 6th comment from Kevin White's link also, suggesting a Schottky diode in parallel to low-side (here high-side) FET.)
To some amount Rload influences these voltage spikes also, because of magnetic field stored in the inductor coil.

\$\endgroup\$
1
  • \$\begingroup\$ I don't think that would work because it also adds extra capacitance across the source-drain. \$\endgroup\$
    – DKNguyen
    Apr 3, 2021 at 4:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.