Have you considered, or is it feasible to use some kind of jig to mount the probes in during calibration? Are any of these meters just count rate meters (friskers)? If so, you can use a pulser to do an electronic calibration, and then verify the detector response and/or efficiency using a small planchet source.