Hi,
I often use VISHAY - Z201 Series bulk foil resistors which are specified to have ± 0,05ppm/°C and are really nice (and expensive).
Unfortunately they have one feature that no one wants to share: after soldering (even by hand) they tend to drift about 30...50ppm too high (in my case) and the resistance slowly decreases in 1 or 2 months (!) to a stable level (1ppm). l performed some measurements using ASL AC thermometry bridge to compare DUT resistance against stable Leeds & Northrup resistance standard. My colleague from calibration lab used better equipment and also confirmed that. The results can be worse when using machine soldering when the component is exposed to high temperature for longer time.
The first chart shows resistance measurements before soldering and after soldering and cooling down to ambient temperature (first 100seconds).
The second one shows measurement over first 24h form the moment of soldering.
Bartlomiej Radzik, www.rfscientific.eu
This hysteretic effect is well known for this kind of technology, as it is similar to strain gauge resistors.
Avoid heating more than 30°C over room temperature by using Cu plated clamp.
I have described that in the LTZ blog on VHP201Z type resistors.
If this shift arises, do a symmetric thermal cycling, like degaussing.
The metal foil resistors are NOT specified 0.05ppm/K, that's an idealistic value only, Vishay call it the typical value, but to the experience of several customers, that is also not true.
Most types are specified 2ppm/K max.
Typical values are mor between 0.2..1ppm/K
Frank