I'm in the device's oscillator circuit right now - but the whole thing is carbon comp resistors. Trying to correct some of the drift, I replaced all the associated parts with some mil-qualified metal film parts from my stocks. The device would drift 10 counts out of 85 on warmup, now it's 2 in 85. I'm not replacing the rest since those don't really matter, it's the oscillator that does the dirty work here. The rest of that 1970s shit can stay where it's at.