[Cytometry] Problem with BD CS&T laser settings
a.smith at centenary.usyd.edu.au
Tue Jul 21 10:53:53 EDT 2009
Thanks so much for writing this up in so much detail for everyone.
We had come across the same problem recently and have resorted to
pasting the correct laser delays on a card on the instrument so that
people can double-check they have the right settings before they
start. Most of them have been good at choosing to accept CS&T settings.
In fact, it seems that this may be an extension of problem we first
noticed with the introduction of the user controls in Diva 5.
Basically it was possible then to get into the situation where someone
could have the laser settings changed under them with no ability to
change them back. The situation where that could happen was much more
limited (basically you had to have a some users with full or partial
permission to change some settings, and some users with no
permissions) so maybe it wasn't common.
I did submit a detailed report to BD but don't believe I heard back -
and there are clearly some additional traps in Diva 6 that are going
to catch a lot more people.
(I can't remember if I sent a report to list - I thought I might have
but I can't find anything in the archives).
Centenary Institute, Sydney, Australia
On 21/07/2009, at 5:18 AM, Christopher Bare wrote:
> We at Rockefeller recently discovered a dramatic "feature" of the
> DiVa 6
> software utilizing CS&T to set laser delay and area scaling when
> applied in
> a multi-account (multi-user with different logins) setting that can
> drastically wrong values to be applied. That, of course, can make your
> parameters falsely negative for the entire octagon/trigon.
> The Problem:
> When a user logs in, DiVa checks the database against the Current CS&T
> values stored on disk. If there is a mismatch, the user is given a
> Box asking to "Use CS&T" or "Keep Current." If the user clicks "Keep
> Current," the instrument laser settings are changed to non-current
> (out-of-date, invalid) values EVEN IF the user does NOT have
> permissions as
> assigned in the "Administration" window.
> The laser settings that the user previously used may be very very VERY
> different from the current instrument configuration (say, if a new
> laser was
> A Solution?
> If an account does not have permission to adjust delay and scaling,
> shouldn't get the Dialogue Box asking them to keep current! BD should
> reprogram DiVa to ALWAYS apply Current CS&T settings for non-
> We have instructed our users to ALWAYS choose "Use CS&T," but not
> listens, or understands. Failure can render subsequently acquired data
> invalid and useless. And unlike compensation or airbubbles, this can
> not be
> retrospectively repaired.
> A fix is better than a workaround.
> I found this because a user complained that his CD3 APC was all a big
> negative smear. I noticed his 640nm delay was 18 instead of 32. I
> Current CS&T and his positive population magically appeared.
> We run two self-service LSRII (one 4 laser, one now 5 laser). We
> define a
> single Instrument Configuration for all (200+) users with parameters
> for laser source and PMT position (e.g. 488-B) in order to prevent any
> conflicts for mismatched configs and avoid end-user mistakes. We
> deny users
> the ability to change delay and scaling, instead administratively
> run CS&T
> weekly and post the numbers on the machine for user confirmation. We
> do not
> allow users to run CS&T themselves, or create and use alternate
> Configurations. We recently upgraded one machine from 4 to 5 lasers,
> reordered their intercepts. This changed all delays (except 488).
> The way DiVa sets a user's laser delays and scalings:
> If the user chooses "Keep Current," DiVa sets the values according
> to the
> Last Tube created by the user during the previous login. If the user
> not have any tubes in the database, DiVa uses the user values from a
> different user_account_laser_defaults table, which is set when the
> user last
> logged OUT. So if CS&T has been run since the user last logged in,
> will be a mismatch and the user gets the Dialogue Box.
> Cris Bare
> Rockefeller University
> Flow Cytometry Resource Center
> Cytometry mailing list
> Cytometry at lists.purdue.edu
More information about the Cytometry