TPS22965: Power management forum: TPS22965: Non functioning of TPS22965NQWDSGRQ1 load switch

Hi Devendhiran, 

Thanks for reaching out on E2E!

Please find the answers to your questions below:

1. Is there any timing restrictions are there in the slew rate of VIN and VOUT voltages (power up and power down)?

   [AH]: The CT pin on the TPS22965/N controls the slew rate of the device. What are your operating conditions?

2. Is there any timing restrictions  between slew rate of VIN and VOUT with respect to slew rate of CT pin voltage?

   [AH]: You can refer to Table 1 for a correlation between Rise time and the CT capacitor. I'm not sure what you mean by "timing restrictions", can you elaborate? 

3. Do you have any diagnostic procedure for the verification of failure of internal MOSFET and switching circuitry?

[AH]: We can't guarantee the mode of failure for the internal MOSFET. Depedning on how the device fails, it could either fail as a short or open depending on the stress across the FET.

Could you share any schematics, oscilliscope shots, or more details regarding your operating conditions? When does the device fail? What is your output load and capacitance? 

Thanks,

Arthur Huang

To find the latest information on Power Switches, visit www.ti.com/powerswitch

Hi Arthur Huang,

Thanks for your reply.

[1] The operating conditions are listed below for your reference :

• Input Voltage = 3.3V
• VBIAS voltage = 5V
• On Control (3V3_IP_EN) = Driven as high from micro controller (LVCMOS 3.3V level) -as a default grounded with 10K resistor.
• CT pin = grounded with 1nF capacitor

Attached schematic for your easy understanding.

[2] Timing restrictions:

VIN voltage generated first, BIAS voltage will be generated second. After the proper settling of both the power rails, ON signal will be released from micro controller.

Since this switch (TPS22965NQWDSGRQ1) is not having a quick discharge feature, when the ON control negates to active low,  VIN and VOUT rail encounters different ramp down timings. So here I am referring the term timing restrictions as for instance, VOUT should be reached some threshold before VIN reaches some threshold and vice versa. In the same manner VIN, VOUT power down timings with respect to VBIAS voltage. 

[3] I agree with you. MOSFET may fail as either short or open between any two of its terminals. In such case do you have any measured values and test cases to validate its failure. Even though I am not able find the mode of failure, If I could able to isolate the failure region within the switch (either the switching MOSFET or switching circuitry) , it may give me a better insight to prevent in future.

RESULTS OF MEASUREMENT :

------------------------------------------

1. VIN ramps properly and settled at correct voltage level.
2. VBIAS also ramps properly and settled at correct voltage level.
3. ON control also driven as high from micro controller after the validation of input voltage rail.
4. But, upon the assertion of ON control signal, CT voltage is pulsating and VOUT also not observed at output.

Attached the snaps for your understanding.

Hi Devendhiran, 

Thanks for the additional clarification, the data you collected is very thorough and helpful. 

At first glance, it seems unusual that VOUT isn't rising on the output. Even with the configuration you described with EN waiting until VBIAS and 3V3 are present, the device should operate normally. 

1: What is your downstream load? Can you replicate the same test with your downstream load disconnected? 

2: It also looks like the CT pin is being pulled down by something, is there a short to GND on the CT pin or is your oscilloscope probe pulling the pin to GND? Normally, it doesn't oscillate in the behavior you've observed. 

3: What is 3V3_IP? Is this the output voltage? 

Thanks,

Arthur Huang

Hi Arthur Huang,

Thanks for your comments.

Please find my replies for your comments below:

1: What is your downstream load? Can you replicate the same test with your downstream load disconnected? 

[Reply_Devendhiran] - Our downstream load is SoC. Right now we dont have any option to isolate the load from the output of load switch.

2: It also looks like the CT pin is being pulled down by something, is there a short to GND on the CT pin or is your oscilloscope probe pulling the pin to GND? Normally, it doesn't oscillate in the behavior you've observed. 

[Reply_Devendhiran] - We have physically measured the impedance between the CT pin and GND and noted that it is in the range of >30Mega Ohms.

3: What is 3V3_IP? Is this the output voltage? 

[Reply_Devendhiran] - Yes, It is the output voltage of Load switch.