output chip of my pins shows 100 ohms resistance instead of open circuit.

[anilmutha]

Hi All,

I am testing my analog chip which is flip-chip'ed on a AlN ceramic interposer and it seems there is a pin which shows 100 ohms of resistance to ground instead of open circuit. The pin is basically the bias diode of a current mirror. The pin is protected by ESD diodes on chip. I have supplied voltage to it and the current drawn by it is consistent with 100 ohms measured by the multimeter. This is recurring problem and I would like to known what the problem is.

I probably have couple of theories of whats going wrong:
1) the ESD pads have blown up due to electrostatic discharge ( would that results in 100ohm resistance?)
2) while assembly maybe the solder ball on the pin have shorted to the adjacent pin. There is a ground and supply pin next to it. The measure resistance between the pin and supply is 1.7kohms and 100 ohms between pin and ground. If there was a short I would have assumed resistance of around 1 ohms or so.

Is there somebody out there who could help me out with this I would really appreciate.

Thanks
Anil

 

[amit.kumar11]

Hi,

First could you tell me chip is BGA or any other package.I assume it is a BGA.

One thing you can check is using x ray you can see whether there is short between your pin and some other pin. Since chip is BGA there will be so many surrounding pins.If it is shorted to ground then it shouldn't show 100 ohms.It might have shorted with some other pin.

 

[betwixt]

Before looking for faults, are you certain you are not just measuring the voltage drop across the protection diodes (or other PN junctions) as your meter passes current through them? You should really be measuring with a voltage significantly lower than one PN forward voltage, maybe using 0.1V or less. Most multimeters use much higher voltages and could make the ESD diodes conduct, giving the appearance of a resistance across the pins.

Brian.

 

[anilmutha]

Hi,

First could you tell me chip is BGA or any other package.I assume it is a BGA.

One thing you can check is using x ray you can see whether there is short between your pin and some other pin. Since chip is BGA there will be so many surrounding pins.If it is shorted to ground then it shouldn't show 100 ohms.It might have shorted with some other pin.

The package is QFN. The pin two adjacent's are supply and ground. I see a resistance of 1.75kohm to supply and 100 ohms to ground.

Before looking for faults, are you certain you are not just measuring the voltage drop across the protection diodes (or other PN junctions) as your meter passes current through them? You should really be measuring with a voltage significantly lower than one PN forward voltage, maybe using 0.1V or less. Most multimeters use much higher voltages and could make the ESD diodes conduct, giving the appearance of a resistance across the pins.

Brian.

I have done a IV curve for this as well by stepping voltage from 0.1V to 0.5V and its drawing 1m to 5mA of current. So I dont think I am forward biasing the diode.