They don't necessarily have to be calibrated to a particular salt though they often are out of the box and that's typically sodium chloride. The point is that they measure conductivity which is a function of the concentration of each ion and its mobility. If you have constant or fairly constant ratios of ion concentrations you could calibrate to the mix. So if, for example
you had a Ward Labd report like this one (the first I could find looking back)
pH 7.4
Total Dissolved Solids (TDS) Est, ppm 47
Electrical Conductivity, mmho/cm 0.08
Cations / Anions, me/L 0.7 / 0.8
ppm
Sodium, Na 2
Potassium, K 3
Calcium, Ca 11
Magnesium, Mg < 1
Total Hardness, CaCO3 28
Nitrate, NO3-N 0.3 (SAFE)
Sulfate, SO4-S 1
Chloride, Cl 2
Carbonate, CO3 < 1
Bicarbonate, HCO3 38
Total Alkalinity, CaCO3 31
we have a reported TDS estimated by conductivity of 47 whereas the sum of the reported ions give us 60. The reasonably close balance (0.1 mEq/L) says this is a pretty good report and so we can conclude that
TDS ~ 60*conductivity(mmho/cm)/0.08
If, in the spring, there is snow melt and we see conductivity of 0.04 we are probably safe in assuming that the TDS is 30, that the alkalinity is about 15 etc. If, OTOH, the conductivity suddenly goes up to 0.1 in the winter we might suspect that it is due to chloride from road treatment and this crude approximation might not be so good.
If your meter only reads TDS, i.e. you don't get an actual conductivity reading then use
TDS ~ 60*TDS_reading/47
In place of 47 use whatever your meter read on this sample. IOW, check the TDS using your meter before sending the sample to Ward Labs and compare to the calculated TDS you get by summing all the reported ion contents (don't forget to multiply the Sulfate number by 3 and the Nitrate number by 4.43).