PFET Breakdown- Buck Converter (Solved)

Thread Starter

stasbala

Joined May 24, 2021
11
PFET Datasheet: ZXMP7A17G (diodes.com)
LM3489 Datasheet:
https://www.ti.com/general/docs/suppproductinfo.tsp?distId=10&gotoUrl=https://www.ti.com/lit/gpn/lm3489

Hello,
I'm pretty new to power supply design; this is more or less my first solo foray into an SMPS. Schematic attached as LM3489_SMPS.pdf
PCB layout images in SMPS_Layout_3.docx

A couple notes:
- Depopulated Ceramic capacitors due to them causing an ESR problem on output of converter: this converter requires a minimum of ~ 50 milli-ohms ESR on output for the switching frequency to be high enough for good operation. 50 milli-ohm resistor added in series with bulk output capacitor to ensure ESR.
- Output loaded with test-load of 68 ohms resulting in expected current draw of 73.5 mA baseline.
- Inductor saturation current is at 1.5A, 20% tolerance on inductor value. currently have a 400 mA current limit on input to supply.

I have a buck converter design, using the LM3489 QMM/NOPB from texas instruments, in which the supply will switch during startup, but then latch the PFET gate high, although the output voltage will continue to track the input voltage. This design was specified to function up to 33 VDC at input, but I am testing at 8.50 VDC from a bench supply at the input to avoid destroying my parts en masse.

The PFET behaves as if it is being ruined by transient voltages, but I am unable to catch transient spikes on my scope capture. I have verified this with several fresh PFETs, and each one is consistent in this behavior. Scope images are the startup waveforms of the supply for switching node, gate, and output. The FET failure behavior is as follows:
Power is off for measurements:
Rdg = 9.5 ohm - typically displays high-impedance
Vds = 0.54V - Body diode is broken down?
Vsd = 0.54V

I'm fairly certain, that at 8.50 VDC input, this PFET should be only strained at a bit under half of its rating in any category, and yet it is apparently unable to handle the strain.

Would anyone be able to shed some light for me on how to properly go about fixing this issue? Much thanks in advance!
 

Attachments

DickCappels

Joined Aug 21, 2008
9,317
Yeah, they are very hard to turn off when installed backwards. I'm glad you found the problem.

It happens to all of use sooner or later.
 
Top