MPC8548E Core is drawing more current

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

MPC8548E Core is drawing more current

1,001 Views
BASIL
Contributor II

We designed MPC8548E based board. We r using Linear Technologie's 15A power module (LTM4611) to power the core of PowerPC. We are following a powerup sequence of VDD_CORE --> I/O supply (or as specified in datasheet). Initially after board power on, we are not feeding any clock or reset to processor, We are just powering on the core voltage(Note that at this moment no IO is active), at that moment, we are noticing a strange behaviour, powerPC is getting heated up like any thing and it is drawing a current of 10A.

There is no short between VDD_CORE and GND plane (Impedance measured between these rails is 5 Ohms).

Using X-Ray we have seen the assembled PowerPC BGA, there are no traces of ball to ball short.

Then what could be the reason for this unusual behaviour?

I am attaching, the powersupply portion of the schmamtic... Please reply immediately to my query because we can not proceed futher, if this is not working

0 Kudos
4 Replies

846 Views
hwrobel
NXP Employee
NXP Employee

Assuming you followed the EC and possibly also AN3640 for any other points, you point out the problem already.

You are not providing HRESET and you are not providing a clock, i.e., you leave the whole part in a totally indeterminate state.

SYSCLK must *always* be connected, and in the EC it is stated that you must have a stable SYSCLK and HRESET for at least 100µs before letting go of HRESET.

Also, just to be sure, verify the ramp rates for your power supply per EC.


846 Views
BASIL
Contributor II

Thanks for your reply. I found it very useful. I need some more clarifications related to this matter.

We are following power on sequence of VCORE --> then VCC_2.5V (IO) --> VCC_3V3(IO) .

Immediately after turning on VCORE ie. 1.1V the current draw is becoming 6A. Also we have captured the 1.1V power cycle in an oscilloscope, that I have attached herewith.

The SYS_CLK can be applied only if I am able to power on 3.3V IO supply. But at the beginning of the power sequencing itself, 1.1V is failing so I am unable to poweron 3.3V supply in the board.

Since SYS_CLK logic in powerpc is operating at 3.3V IO, I wont be able to supply the 3.3V, If 1.1V itself is failing.

If I skip the normal power on sequence, ie If I am applying 3.3V IO supply alone and SYS_CLK , still the powerpc is getting heated up.wave.jpg

0 Kudos

846 Views
hwrobel
NXP Employee
NXP Employee

Make sure you follow all connection recommendation footnotes in the EC, check AN3640 again for incorrectly connected pins and especially ensure /HRESET activated right on power up. No /HRESET, no defined device operation.

If that won't help, you will likely have to contact your local support or issue an S/R.

0 Kudos

846 Views
BASIL
Contributor II

OK... Finally we solved the issue. It was a problem because of core power supply module. Thank you..

0 Kudos