

And another unidentified Philips lab part soldered on a PCB.
The lid of the package is attached with a tape.

The package contains a relatively small die.

It appears to be a test circuit. This is supported by the fact that very little of the available silicon area is used and many of the bondpads have no function. The character sequences QU2AGC1 and OQT063C1 are shown next to the Philips lettering. The first character string can also be found on the circuit board and could therefore be the name of the circuit.
The outer bondpads on the upper edge clearly represent the power supply. In addition to these two the circuit is contacted via seven further bondwires. The three bondwires that lead to the inner area of the die are particularly interesting. There are no bondpads there.
This image is also available in higher resolution:
https://www.richis-lab.de/images/tbd/04x04XL.jpg 24MB

The integrated circuit is quite clear. There appear to be two signals coming in from the left, each of which is routed via a relatively large coupling capacitor. This could be a differential input. Here you can see that unusually thin lines lead from the inner three bondwires into the circuit. It is very likely that these lines were constructed using an FIB.
An FIB (Focused Ion Beam) is a tool that generates a beam of accelerated ions. Various elements are used to generate these ions, including gallium and some noble gases. Where the ions hit, they remove material. By introducing different process gases into the reaction chamber, it is possible to ensure that material is removed selectively. In contrast, other process gases allow certain layers to grow exactly where the ion beam strikes. All in all, this provides a tool that allows integrated circuits to be extensively modified down to the smallest detail. Small testpads can also be created. These are usually testpads for test needles. In this case, the testpads were apparently large and robust enough to apply bondwires to.

The lines subsequently constructed with the FIB are approximately 1µm wide. There appear to be deposits in the immediate vicinity. The point where the retrofitted line contacts the original circuit is clearly visible, even if the details of the contact itself are not resolved.

In this line, the contact can be recognized by a taper. In this area, the metal layer has also been cut at three positions. A resistor has been configured that consists of several partially bridged elements. A remnant of the metal layer appears to have remained at the middle interruption.

A wider, diffuse stripe was already visible in the previous image. Here, this stripe can be seen once again. A second stripe crosses outside the circuit. It remains unclear how these stripes were created.
https://www.richis-lab.de/TBD04.htm 