INA128: INA128UA

PCN# 20220802000.1 introduced an entirely new die. The changes are listed for revision F of the datasheet.

Which electrical characteristic is tested by your failing test?

Hello Anna,

Clemens is correct, that PCN details the addition of a new design for this device. Please provide what parameters are not meeting your specification. Do you have a schematic, or some measurements to provide us more information on the behavior?

Best,
Gerasimos

Dear Gerasimos,

I have two information:

First one, that we have checked with our multimeter in diode tester mode for +4legs and -6legs.
We have compared 2-2IC in soldered position. With the new ICs had 140mV more then the older ones.

The second one is from the end products tester's manufacturer, who gave us the following feedback: 

„In connect with the assembled board testing issue we have found that the test fails are related to the voltage drop measured on clamping diode. The ICT test we perform is a kind of low level component verification. That means, the ICT test itself doesn’t power up the circuit, there is no real voltage measuring on dedicated test points, etc. The low level ICT test is performed with goal to detect eventually missing or wrong way inserted component based on measurable internal structure of the IC. In current case the logical gate IC has clamping diodes on their inputs. Knowing the position of the clamping diodes inside of the IC by simple diode measuring we can decide whether the IC is present with correct orientation or not.

 The forward voltage drop on such diodes is around 0.6V. To be precise on evaluation for accepted level we had chosen the forward voltage drop on first tested sample (the golden sample prior the first manufacturing batch). The accepted threshold voltage has been defined to be +/-30% against the golden sample’s value.

Now , we use exactly the same specified component but surely from different manufacturing batch or even from different manufacturer. This cause slight variation on clamping diode’s forward voltage drop which result fail on test since the base parameter has not been adjusted. But as long as the right component has been built in we can be sure the whole circuit will work as expected. Based on this we can adjust the acceptance level to be narrow enough to exclude mistakes but wide enough to handle slight manufacturing variances.”

According to this information we think that the tolerance modified in the new version, similar to the measured voltage.

Considering these information, we think that something changed.
Do you have any idea why this could be happening? Did you change anything during the production process?

Thank you in advance!

Anna