Hi Stephen, The problem you describe is related to the relatively low impedance of such divider.This has always to be taken into account, like the voltage drop over connection leads. This special effect (inside the instrument) is not explicitely described in the 750A, 720A, 752A manuals, but indirectly in the Fluke book, Calibration - Philosophy in Practice, 1st edition, p. 52, description of 7105 system, page 63,where the lead wire resistance error is calculated. (This can be downloaded via google books.) In your example, you find a 20uV error. The 750A draws 1mA from each input voltage applied, so that equals an internal lead resistance of about 20 milli -Ohm, which is quite reasonable. another internal wiring would not change the game.In this sense, I do not agree, that the 750A is wired in a wrong manner, this divider techniques generall has its accuracy limitations.External lead resistances might be 10 times as high, causing 10 times higher errors. So this internal error depends on the input voltage, or equivalently on the divider resistance in use. A 20ppm error occurs only at 1.1V input, 10V create 2ppm, and so forth. This may be compensated either by using Kelvin sensing of the voltage source, or in some cases by the 721a, lead compensator. These methods are also described in detail in the Fluke Calibration handbook. Frank --- Alle Postfächer an einem Ort. Jetzt wechseln und E-Mail-Adresse mitnehmen! Rundum glücklich mit freenetMail