Lochner, K.H.K.H.Lochner2022-03-032022-03-032001https://publica.fraunhofer.de/handle/publica/20105210.1515/labm.2001.25.5-6.1772-s2.0-0034975062As a result of the AiF research project No 11503N1 "Test and Improvement of the Measuring Accuracy of Multichannel Piston Pipettes", the time for testing a multichannel pipette was reduced from 11-12 hours (using the nominal gravimetric test method) or 6 hours (using a modified gravimetric method) to 1.5 hours by developing a semiautomatic gravimetric unit with weighing cells operating with strain gauges (testing unit 1). This was achieved by simultaneously pipetting with up to 12 channels and then simultaneously weighing the dispensed fluid volume. For this purpose, a testing unit was constructed in which the distance between the weighing scales is equal to the distance between the channels of a multichannel pipette (= 9 mm). By the use of weighing cells with strain gauges, a weighing accuracy was obtained that corresponds to an analytical balance of 0.1 mg resolution. Thus, this unit can test multichannel pipettes with a nominal volume from 50 to 300 µ. It was not possible to further increase resolution and accuracy by using strain gauges. Based on previous experience, a second, significantly optimised unit (testing unit 2) was developed. The most important new aspect was the introduction of weighing cells with electromagnetic force compensation, as they are used in semi-microbalances. Using this new testing unit, pipettes with a nominal volume of up to 10 µ can be tested in significantly reduced time. For the very first time, a complete, standardised gravimetric test of a multichannel pipette was performed in less than 15 minutes. The gravimetric test of multichannel pipettes by four different producers showed that no pipettes exceeded the maximum number of errors permitted in DIN 12650-8. The dispensed volume of multichannel pipettes is adjustable between approximately 10 % of the nominal volume and the nominal volume. At the lowest adjustable volume, the uncertainty of measurement was reduced on average to approximately 50 % of that obtained at the nominal volume. Slight deviations occurred at different nominal volumes and between different producers. With few exceptions, the dispensed volumes of the individual channels were more or less equal. Compared with manual pipettes, electronic pipettes showed fewer systematic and random errors.de666616journal article