We have recently begun using X5R dielectric capacitors. Although they have a better temperature coefficient over Y5V, they age much faster (5-7% logarithmically per decade). I assume the reason for using the X5R was the availability of the capacitance value in an 0603 footprint. The specific value we ar using is not available in X7R. The main problem is when testing the parts on the ICT, they will fail out of tolerance on the low end, even after being "deaged" through the convection oven. Does anyone else have any experience with X5R on in-circuit testers or in the production process?