I am using a TPS54020 buck converter, and am seeing a glitch on SYNCOUT that corresponds to the time that the output switch turns off. Strange thing is that I see this only on one of four PCB assemblies. The glitch is about 40nS wide, and does not happen every cycle. At some load conditions (high or low) or low VIN it goes away. Scope shot is attached (1 = yellow = SYNCOUT, 2 = cyan = PH).
The glitch is causing me problems as I am using SYNCOUT to clock some other devices on the board.
My first suspicion is PCB layout or a solder issue, since only one of four boards act this way. But the layout is pretty clean, and the solder looks good.
I read that the RT/CLK pin is sensitive to noise pickup. When I look at RT/CLK with a scope (182k to ground on RT/CLK), I see basically nothing, very close to 0 volts, and what looks like noise synchronous to the switching. Is this normal? I expected a ramp, or something...
Anybody seen this phenomena?
Any ideas or suggestions for further debug are appreciated. Otherwise I may replace the IC and see if it goes away.
