TPS543820: IC Vin input impedance after power-off

I am not able to replicate this issue. 

I measured the resistance of the TPS543820EVM (no board modifications) on the bench with a DMM (34401A agilent) on the VIN to PGND test points. 

When powered the DMM reads 'OVLD' and after disabling the HP6652 bench supply ( method 1 via disabled input supply or method 2 enable pin/then input supply)

, the DMM display flashes and very quickly changes numbers and within (1 to 2 seconds) reads ~100ohm and settles to 93ohm in a few seconds. 

Has the TPS543820 issue been verified on an evm board or non ICT test setup?     I will ask some teammates for suggestions. 

The manufacturing test team is the one that brought up this issue to my attention. 
I have been able to replicate this issue on one of our circuit packs, but then, I have multiple POLs on the card, so I cannot for sure say it is the TPS543820. I do not have an EVM. 

Today, I have just tried isolating Vin from the rest of the POLs, and measuring high impedance once the power is removed, and the external power supply is removed.  So, it might be a non-issue. I need to get more information from the manufacturing team. Thanks for your help. 

Hello Luc, 

Have you found a resolution to the impedance issue,  if so I will close the thread.   If you need assistance post and I will help 

Hi David,

No, I have not yet a resolution for it. 
More information: The low impedance at Vin only occurs only when they test the BP5 and Boot capacitors. This is done by injecting a low voltage (2.8V @ ~2mA) on the capacitor pads and verifying that the component is there. They state that it passes the first time, then they program the card and power it up, then power off the card and run it though ICT again, and this when they see the short at the TPS543820 input.  I have two of these TPS devices on this card and they are both performing the same. 
They have decided to not test these capacitors for now until the issue is resolved. 

Here is the procedure they sent me:
-Provider inserts an unpowered fresh card at ICT

-ICT execute and test the “SHORT” stage passing all the net impedance measurements including TPS’s reported

-Once completed and pass  the short test , ICT programs the XML UCD load at U30 and U49 with the TI POD 2006

-After program completion ICT provides the source voltages in order to power up the card now that Loads are programmed

-The card powers up and runs the rest of the powered ICT test, test in now completed and card is removed from the fixture.

-When you place again the card at the ICT fixture and retest since the beginning ; now when ICT performs the “SHORT” stage the ICT reports the unexpected shorts and the ICT test fails

ICT uses the same procedure as the xxxx Functional PMA’s, unexpected shorts has been confirmed after program and test a card on the functional test as same as the ICT

Hello Luc,

I am trying to the reproduce the behavior using the TPS543820 EVM and having no luck.  

I added a diode between the lab supply and evm so that I do not measure the kelvin sense resistors (100ohms) of the bench supply. 

I apply 12V to the input, the device regulates and then power down and measure the input impedance, it is always greater than 100kohm.   

I take an additional 3V/10mA supply and probe BP5 and BOOT to SW and reverse the order Boot to SW, BP5 to GND

The input impedance changes as the meter updates as I apply the supply to the different pins but is never low. 

Can I have an evm sent to you so that one of the ict tested devices can be installed and experimented on the evm.   

Hi David,

The manufacturing team will send me the circuit card with the ICT tested devices that they are seeing the issue with. I will hopefully be able to replicate on this card what they are seeing.

Will update you when I receive the card. 

thanks for your update.