[RASMB] Photomultiplier tube disturbances?

[Borries Demeler]

> 
> Greetings All:
> 
> As a hardware jockey, I'd like to put in my $0.02.
> 
> The beauty of a dual-beam spectrometer (which the XL-A is in true absorbance mode) is that it is self correcting for moderate to low frequency error (noise and drift).  When the reference and sample signals follow the same path then irregularities like lamp, PMT or electronic drift are removed when the I/Io ratio is taken.  Something like drift due to temperature changes during warm up (or caused by the other AUC in the room stopping and dumping a lot of heat into the room) will be difficult to correct for using pseudo-absorbance.  The demands of the instrument's stability are greater using pseudo-absorbance.
> 
> I don't mean to say that pseudo-absorbance is wrong, but that moderate to low frequency error that might be corrected out using absorbance can show up on the same time scale as the sample signal using pseudo-absorbance.  I would suggest devoting one channel to water or buffer to serve as a validity check, and possibly serve as the Io signal for the other channels.
> 
> Glen

Glen, that's exactly what UltraScan is doing: You always need to collect
one channel of water, which does provide the time-dependent intensity
variation. The signal is averaged over a small radial range and the
average is used as the Io for the pseudo-absorbance correction.

If the fluctuation in intensity is more rapid than it takes to scan the
previous cell then even that method will not help much, and you may indeed
be better off using a separate correction from each reference channel,
which should then contain water. But even then, you would be better off
in most cases with intensity than using absorbance optics. The variations
would be small enough to be effectively fitted with ri noise correction.
YMMV depending on the health of your instrument.

-b.