Pipette Calibration - Addendum

   The calibration process described in "Repair and Calibration of Pipettes"
is incomplete. The pipette should be checked at all three points specified
in the Rainin published specs. These can be found in the Pipetman instruction
manual. Below are relevant excerpts.

Model   Volume(uL)      Mean Error(uL)
--------------------------------------
P-2     0.2             0.024
        1.0             0.027
        2.0             0.030

P-10    1               0.025
        5               0.075
        10              0.1

P-20    2               0.1
        10              0.1
        20              0.2

P-100   10              0.35
        50              0.4
        100             0.8

P-200   50              0.5
        100             0.8
        200             1.6

P-1000  100             3
        500             4
        1000            8


   Often, the pipette will perform within the specified error window at one
extreme of its range but not at the other. Sometimes compromises can be made   
in calibration to satisfy both extremes.

   For example: A P-1000 has been calibrated at the maximum volume and is
measured at 999, well within the window of error. However, when checked at the 
100 microliter setting it measures 104, outside the specified window. The 
volume is reduced slightly by adjusting the calibration wheel. Now the 
measurement at the maximum setting is 993, and at the 100 microliter setting 
it is 101. Both measurements are within the specified error windows. The next 
step would be to check the 500 microliter setting. Sometimes a very tricky 
3-way compromise is needed.

   If a compromise between settings at the extremes of the pipette's range is
not obtainable, but most of the pipette's range can be calibrated within spec,
some users are satisfied to calibrate the portion of the range most used. For
example, a P-1000 that can be calibrated from 200 to maximum is often considered
usable. The user would switch to a calibrated P-200 for lower volumes.

   If the error measured at the first calibration point is very large, say 10%
or 20% of full range, the instrument may need to be rezeroed. This could be
the result of the threaded metering shaft having been completely removed for 
some reason. Bring the instrument to the repair shop for rezeroing.