Posted by Asmoth: Mon Jan 29, 2007 9:07 am    Post subject:

Sebastian,

If your using the dynamic/test pattern method, try to check with your test engineer the voh/vol settings. Also check with him the IOL/IOH settings.

The dynamic method O/S test actually checks the diode drop of each digital pins. The pattern do this one at a time. While checking the diode drop of a certain pin, all the other digital pins are set to ground(0V).

The pattern(through the pin card) forces a current value in(or out) of the pin being tested. The value of the current is set on the IOL/IOH. Larger current value would result to a larger diode drop voltage. This voltage is then compared to the VOH/VOL values. If the diode drop voltage is outside the VOL/VOH range, then an O/S failure will occur.

One causes of the O/S failures is the inappropriate settings of these values. The IOL/IOH settings might be set at a value large(or small) enough such that the resulting diode drop would marginally exceed the VOH/VOL values.

Also, try running the test on "Force Through" mode, meaning it will continue testing even if it encounters failure. If all the succeeding tests pass, then this means that the pin-to-pin contact with the contactors/pads is OK. So it is likely that the O/S failure may have been cause by the inappropriate settings of IOH/IOL/VOH/VOL. If this is so, then try adjusting these values.

Also, since you just lately encountered this failure, I assume that previous lots don't have this problem, which means that the settings of IOH/IOL/VOH/VOL are working fine until this current lot. This would suggests that this O/S failures maybe related to this particular lot. If this is so, then please verify the diode drop of some pins and compare it with the diode drop of the previous lots. You can use a digital voltmeter with a "diode checker" feature on this evaluation. It uses the same principle as with the pattern method of the ATE, wherein it forces a current on the diode and measure the voltage developed accross it.

It is likely that there may be a shift on the lot in question.
_________________
Asmoth72

Posted by PW57 Guy: Thu Mar 08, 2007 5:54 am    Post subject:

Hi Sebastian,

What is the ATE measurment you're getting on the parameter you are failing? There are some factors that may lead to these failures (hopefully it helps...)

1. If you have leakage problems, you might want to check the hardware you were using. Additive resistance on the hardware (eg, DUTBOARD, LOADBOARD, PROBECARD, etc) can cause leakage failures. One thing to verify this is to tun the part open-socket. If your leakage measurments were failing the limits, definitely, your set-up is causing the problem. Process of elimination can be done after this.

2. Check with your IC guy, they might change the process on your particular lot you were testing.

3. Work with the test guy and the designer guy, thay might also change something on the part that have caused the failure.

4. Work with the test guy. Ask him what particular test that this failure has something in common on the other test parameters and your current failure. For ex., if your failing leakage on one of your logic pins, check other tests like Functional Test that uses the same pin.

5. Lastly, work with the FA guy. Ask him what analysis tools he can do on a leakage failure you were seeing. I heard from this guys that leakage failure <1uA is hard analyze compared to >1uA.

As I have said, I hope the above mentioned helps. Good Luck!
_________________
...little knowledge is very dangerous...

Posted by Paula: Mon Mar 12, 2007 5:03 pm    Post subject: Re: Leakage

Based on this, what you basically performed is decapped the units and "blindsearch" for defect that might have caused the B5 failure. Good failure analysis approach should have been to validate the failure mode and then isolate the failing pin/s before any destructive analysis.

Back to Page 1 of this archived forum thread...