MSP430F534x increased power consumption during power up.

It sounds quite long to settle, but I don't know your exact code or hardware situation.
Don't forget that at power-up, the core voltage generator has to start from scratch, drawing some current until the core voltrage capacitor is charged.

What you can do to get more info is to set a port pin which you observe with the scope along with the current. Toggle the pin at the start of main, then again when you're done with initialization. Compare it with the current consumption.

Thank you for the comments

In order to measure the MSP430's current consumption the CPU is
removed and seperated from the custom board & measured seperately.

Regarding the hardware there is nothing connected
externally to the MSP430 just the MSP and the power supply.

And regarding the code, there is just a while loop inside the pre_init
as shown below, everything is disabled and no initialization code as such.

int _system_pre_init(void)
{
....Disable WDTCTL
....Disable SVS
....Port initialization
....
While(1);
}

Most of the time it takes 500ms to settle but sometimes it takes more than 1.5 sec.
We belive even 500ms is a little bit longer.

Our customer has tested this behavior with more than 100 units of MSP430F534x...
and in most of the case it takes more than 500ms(sometimes more than 1.5 sec) to settle down.
There is no problem with the LPM modes or the Active modes once the MSP is settled down after the power up.

We wonder if this is a normal behaviour in case of MSP430F5xx series..
we believe there is possiblity of the software(as there is no software as such) or the
hardware(only minimum of external components (decoupling caps, Vcore cap, pullup on RST) are used).
note:the device revision is Rev H.

Indeed it is a bit long. It will take some time to power-up VCORE, but not that long, Did you check the capacitance on VCORE? It should be 470nF (and 4.7µF or more between Vcc and Vss). It may sound silly, but if you have 470µF on VCORE, it will take an eternity to charge and could explain the long settling time. Just an idea, actually the only one I have.

What you could do: set a port pin to signal when your code has reached the while loop. Then you can see on the scope whether the settling occurs while the MSP is already running, or before it starts.

Thank you so much for the comments.
your support in E2E is very helpful.

We have some more data about this issue.
The conclusion is MSP430F5341/2's behaviour is different from other
F5xx devices like MSP430F5310.

In order to cross-check we tried to test the same on TI's target boards with other F5xx devices
and the MSP430F5341 was abnormal when compared to other devices.

1. MSP430F5310 on the "MSP-TS430RGC64B" target board
Result: During power up, it takes few ms to settle down to a normal power consumption

2. MSP430F5341 on the "MSP-TS430RGZ48B" target board
Result: During power up, it takes more than 500ms(sometimes more than 1.5 sec) to
             settle down to normal power consumption

We wonder what is the reason for such a long startup time(settling time) in MSP430F5341?
As other F5xx devices like MSP430F5310 is very small, it is clear that there might be a problem with MSP430F5341.

Prad1 said:

Result: During power up, it takes more than 500ms(sometimes more than 1.5 sec) to
             settle down to normal power consumption

If your code is already running, i wonder why power consumption suddenly drops while the user code is already active and doesn't do anything. The only thing I could imagine is the crystal driver, which is active on startup, even though the port pins are initially set for digital operation.

It's possible that the XT1 driver tries to start the crystal and somehow goes into HF oscillation, which causes the increased power consumption. And could explain, after oscillation ceases, why power consumption drops while the user code is already executing and doing nothing to cause it drop.
You may try deselecting XT1CLK as source for the FLL and for ACLK, by setting UCSCTL3.SELREF_2 and UCSCTL4.SELA_2.

Just an idea.