To remove the effects of influence with quarter bridge measurements, what temperature stability of the quarter bridge completion resistor makes sense?
5 ppm sounds like it would work well. Upon closer examination, we can see it is not sufficient:
Strain gauge 350 Ohm, 5 ppm/K = acc. 1.750 mOhm/K, a 1000 µm/m strain has an effect of 2 mOhm/Ohm at k=2, total 700 mOhm, a 5 ppm/K completion resistor provides an error of 2.5 µm/m each Kelvin or better 25 µm/m per 10 K = 2.5 %, which is poor.
Gantner’s strain gauge module, the A116, uses each strain gauge channel, a 120 Ohm and a 350 Ohm completion resistor, with a stability of 0.05 ppm/K. You can remove temperature influence with this easy solution. See the A116 Data Sheet and Application Report for more information!