[RASMB] radial shift

[Beld, Joris]

Dear all,
 
Lately we have encountered a strange problem with our XL-A. For a SE run we normally scan three times (with ten averages) per cell per wavelength. In principle these three scans are not further than 10-20min apart. Still, we can see a relatively big radial shift (0.2-0.4 mm). The scans are superimposable if shifted over the x-axis. I was doing two consecutive runs (with different proteins at different speeds, from 4k to 30k) with scans both at 235 and 280nm and this strange behavior is more pronounced at 235nm than at 280nm. Of course, one could think of samples that have not reached equilibrium yet but we estimated the time to equilibrium with ultrascan and let the system equilibrate varying from three days to 20h. So, I think the samples are at equilibrium, and moreover, within 10-20min one should not see such big radial shifts.

We have actually two questions:
- do you think we can still use this data? (in principle fitting should not be influenced by radial shifts, or am I making a mistake in that assumption?!)
- does anyone has some experience with this problem and/or knows a solution (before I call Beckman)?

Thanks a lot in advance.

To comment quickly on lamp cleaning. We clean the lamp monthly (with toothpaste) and so far that seems to be sufficient.

Kind regards,

Joris Beld

Swiss Federal Institute of Technology
Zürich, Switzerland