Hello,
What you're seeing is probably a combination of three things:
1)While the 4071 is a high accuracy voltage measurement device, it is not accurate enough to measure the 1.018638V output voltage of a Weston cell. Because our 4071 has a 105% overrange, we can use the 1V range of the 4071, which has an offset error of 800nV. In addition, the gain error at 1.018638V is ~15uV, which is higher than the several uV error you're reading.
2)Fundamentally, any current flowing into or out of a Weston cell is detrimental to the accuracy of the cell voltage. The 4071 has a typical steady state DC input bias current around 30pA, which equates to around 186 million electrons flowing through the Weston cell per second, which will rapidly change the cell voltage. Anytime current flows through a Weston cell, it can take considerable time for the cell to settle back to its nominal value (up to hours, depending on how much charge is transferred).
3)The 4071 has a 90pF effective input capacitance, which means that when we first connect our DMM - and each time we autozero (once per 7.5 digit measurement) or perform offset nulling (once per 7.5 digit measurement) - a large current flows to charge the 90pF capacitor until we reach ~1.01830. The current is only limited by the ESR and the reactive load elements of the signal path, which means we could momentarily exceed the 1mA maximum momentary discharge current of a typical Weston cell. Exceeding 1mA causes an immediate chemical reaction that permanently alters the voltage output of the cell.
Overall:
The 4071 is not an ideal instrument to connect directly to a Weston cell. If you want to continue using the 4071, we'll need to disable autozero and offset compensation, which will lower our accuracy considerably.
Instead of the 4071, I recommend using an electrometer, which has input impedance in excess of 10E10 Ohms, thus reducing charge flow while we measure the voltage of the Weston cell.
If an electrometer is out of your price range, I recommend connecting the Weston cell to our PXI-4022 guard card and then connect the output of the guard card to our 4071, but note that the 4022's 2uV/degC offset voltage drift will further decrease the accuracy of the measured voltage.
Regardless of where you go from here, I recommend precharging the test equipment with a series resistance - and then shorting it out once charged - to prevent a current surge in excess of 1mA.
Also, I recommend you get your Weston cell recalibrated; your testing up to now has probably damaged it.
-John Sullivan
Problem Solver
