Regarding Jim Leary's suggestions about compensation - Mark Corio's 3-color compensation electronics are indeed quite nice. However, as one adds colors, particularly when the labels used are tandems in which energy transfer is incomplete, or when there is energy transfer between dyes, as can occur with fluorescein-labeled antibodies to nuclear antigens and 7-AAD on nuclear DNA, it becomes necessary to have nonzero values for most elements of the compensation matrix, meaning more circuitry if compensation is done in hardware. This circuitry must be placed between the PMT preamp outputs and the log amp inputs, and the analog electronics needed to compensate for more than four colors, even when designed and built with a great deal of both skill and care, are almost certain to raise the noise level to the point at which dynamic range is compromised. The problem is avoided by collecting linear data with 16 or more bits' resolution, and doing both compensation and log transformation with digital computers, as is done in the Beckman Coulter Epics XL and a number of laboratory-built instruments; systems now in development incorporate digital pulse processing, which further improves performance, particularly at low signal levels, by eliminating the effects of glitches in peak detector and/or integrator circuits. -Howard
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:54:17 EST