J721EXCPXEVM: J721 EVM CP board power fault

LD7 is the LED for the main 3.3v power rail, and it should be on any time the power input is applied and the power switch is in the ON position.  LD2 LED is ON (shown in picture), may maybe verify the voltage is correct (TP20).  It should be the voltage that is input to the board.  You can measure the main 3.3V supply (TP130 or TP35).  If the input supply is available but the 3.3V regulator is not working - perhaps it was damaged.

Hello Robert,

Thanks for your reply, 

TP20 shows 12.3V & TP35 show 0V when powered, so it seems that the 3.3V regulator indeed.

So do you mean the LM5140?

What is the correct channel to proceed with the fix?

Would there be any schematics or assembly files to help with the fix?

Schematics (and other design files) are available on TI.com.

https://www.ti.com/tool/J721EXCPXEVM

Yes - the regulator for 3.3V is LM5140.  

Clear, is there a TI-provided repair channel that I can contact?

TI does not offer repair services.  I can maybe help you debug the circuit and possibly locate the problem, but repair would need to be done by someone other than TI.

Debugging the circuit would be very helpful actually, please go ahead and let me know where to look.

First step is to determine if issue is with regulator or load.  With a SOM module not installed and the board powered OFF, you can measure the resistance of the VSYS_3V3 output and GND (using a voltmeter).  If there is a failed peripheral it might be shorting the VSYS_3V3 power net to GND causing and over-current issue disabling the regulator.  You can also use a oscilloscope to monitor the VSYS_3V3 to determine if its attempting to turn on, and if so, what does the output waveform look like?

I measured VSYS_3V3 (TP130) to the GND, it showed 4.5K.

I used a scope to trace VSYS_3V3 at power up, it showed low pulse (~200mV).

Low pulse....so 0v then pulse up to 200mV, and then back to 0V?  

Can you measure voltage at C68 (should be same as input voltage)

Can you measure voltage at C64 (should be near 5V, internal regulated voltage from LM5140)

| Low pulse....so 0v then pulse up to 200mV, and then back to 0V?

yes

| Can you measure voltage at C68 (should be same as input voltage)

C68 shows 12.27V (input voltage)

|  Can you measure voltage at C64 (should be near 5V, internal regulated voltage from LM5140)

C64 shows 0.95V

The VDDA pin of the LM5140 (connected to C64) in an internally regulated voltage, so it points to an issue with the LM5140.  The only external connection is C64 - so unless its shorted, it seems the LM5140 would need to be replaced.