Dear Teams,
My customer faced some questions about TPS62135, and here is the testing report:
1: Abnormal phenomenon: SW pin has not output and voltage drop. When PG pull down, the Enable will be pull down as well, then it will cause power down with time about 240us.
Schematic diagram showing below with power rail 12 to 5, then 5 to 3.3V:
Above picture shows: Blue line is 5V PGOOD; Yellow one is 5V; Green on is 5V SW; Pink onw is 5V Inductor current.
It can be found that SW pin has no PWM output once the 5V voltage drop. At the same time, the inductor ripple current will drop to zero. Once the voltage drop to 5V*96%=4.8V, the 5V PGOOD will pull down actiely.
After pull down about 8us, it has a spike current. The spike current with peak value 4.7A (Quite similar to the in-rush). Respectively, the PWM waveform has larger duty cycle.
However, the voltage has not been set to zero after the voltage drop to SW point. It is more like the high side MOS operated at active region. Voltage peak value is around 4.5V.
TPS52135 Temperature testing: it is based on power to calculate; JT less than 70 degree, which is much less than 160 degree:
2: Second thing: tested the 5V output current waveform and 3.3V output current waveform resepectively:
a) output current waveform for 5V (No OCP occurs, average value is 878mA)
b) output current waveform for 3.3V (763mA with no OCP occurs)
c) Fan (5V other load) output current waveform (279mA)
3: TPS62135 power performance testing:
a) 5V static state loading and dynamic loading to check whether it cause voltage drop:
i) make sure whether chip has problem:
1) No load:
2) 1A load:
3) 2A load:
4) 3A load:
,
5) 3.5A load:
b) test whether the transient response meet requirement:
testing condition 1: level A: 0.1A (rise rate 1A/us)
Level B: 3.5A (Fall rate: 1A/us)
Frequency: 1Hz; Duty cycle: 50%
Result: Output voltage drop slightly: 60mV); PGOOD not status change.
Since peripheral might affect the SYS_5V, we capture the SYS_5V reproduce the voltage drop at TX2_VDD5_CAM V/I waveform, but we did not find the voltage in-rush or current in-rush situations.
Conclusion: Binocular is not the reason to cause 5V drop.
4) There is a log occurs 4-5s. It is quite similar to the previous CRC error:
This log is for 3109's workaround: 5V voltage divider to control 3.3V EN; replace 5V PGOOD to control 3.3V ENABLE
Thanks!
Dylan
