From: Jack Correia <jcorreia@fiona.umsmed.edu>
  To  : rasmb@bbri.eri.harvard.edu
  Date: Wed, 24 May 95 17:01:49 -0500

Re: time out errors

        John Steckert just called and asked me if we have experienced time 
out problems where scans stop for some period of time, occasionally 
restarting at a later time.  He explicitly wanted to know what parameters we 
use to collect velocity data and if it affects these timeout problems.  John 
is not currently on the internet, so I 1) agreed to ask the question for him 
and 2) gave him Jeff Hansen's and Bo Demeler's phone # because I believe 
they were having the same problem with their instrument(s).  Has this been 
discussed before on the internet - I know I have discussed it with my 
service guy, Dennis Foyt.

        We collect all velocity data at 0.01 cm step size, 4 averages and 
continuous mode.  We never use the delay or the interval parameter, leaving 
them at the devault values.  We have only very seldomly seen timeout 
problems.  Once we had to restart the scan and actually lost most of the run 
because we didn't catch it fast enough.  Other times we see gaps in a series 
of scans as if a cycle of scans have been missed, but then it picked up with 
the next scan.  The question is not just how many of us see this problem, or 
how many of us are incapacitated by this recurring problem [John cannot do 
velocity runs the problem is so bad], but what is the cause and what is the 
fix for the problem.  Beckman apparently suggests setting the interval to 
exceed the single scan time, but this does not fix the problem and it limits 
the amount of data one can collect.  John is doing a survey of XLA users he 
knows to document the problem.  If you have suggestions, better yet a fix, 
or experience with this problem pass the info to me and I will summarize for 
John and the rasmb.  If you want to call John he is at (508) 623-2277 or a 
fax # (508) 623-2602 in Massachusetts, USA.

jack