We need to characterize a 0.5 milliohm current shunt, but our bench DMM and 4-wire setup gives noisy readings that wander with temperature. A proper micro-ohmmeter is out of budget. What techniques actually work for measuring sub-milliohm resistances repeatably — Kelvin clips, thermal EMF control, averaging tricks?