Is it possible to have a MTBF of the complete component, which in any case
should be a overestimation of the specific failure mentioned above?
Here is the MTBF report.
Our ADIS-expert told: Our experience suggests that no matter what happens,
there is always going to be a signal, even if it is always zero. I think what
they are trying to ask is, can the device still look like it is working and
still be in error. The proper answer would be to offer that outside of some
minor parametric drift, no. Failures tend to be stuck elements, broken elements
or gross issues that are normally attributed to the assembly process.