Graeme, Yes, the transmissive calibration takes longer than 6 seconds, I would guess perhaps 10 to 15 seconds. Ok, I'll see if I can wonk myself thru a recompile, with MingW. Thanks, Ken At 09:39 PM 5/23/2011, you wrote: >Ken Conrad wrote: > >>> icoms: Write_Read called with ';D00D01^M^J' >>> icoms: About to write ';D00D01^M^J' ICOM err 0x0 >>> icoms: About to return read '' ICOM err 0x2000 >>> icoms: Write_Read Write failed - returning 0x2000 >>> Calibration failed with 'Communications failure' (Serial communications >>> failure) > >Hi, > the communications has timed out here. Unfortunately I don't actually >have a SpectroscanT so I've been unable to test this mode out. Searching >through the manual it does mention that on setting this mode "An optimum >position for the Spectrolino relative to the lamp for transmission measurements >is searched." although the entry for the command itself doesn't mention this. >So maybe it is a "Initialization" or "Move-command" which requires a longer >timeout >than the default 6 seconds, although the mode command is not listed or marked >as such. > >So you should try changing line 1492 of spectro/ss_imp.c from > > ss_command(p, DF_TMO); > >to > ss_command(p, IT_TMO); > >and recompile, and see if that solves the problem. > >Graeme Gill.