>Date: Fri, 6 Aug 1999 11:58:10 +0000 >To: Pizzo@nso1.uchc.edu (Pizzo,Eugene) >From: Andy Riddell <ar3@alf1.mrc-lmb.cam.ac.uk> >Subject: Re: Calibur data acquisition >Cc: >Bcc: >X-Attachments: > >>Colleagues, >> >>Recently we purchased a new FACSCalibur which runs Cellquest on a 400MHz G3. >>I noticed that during initial acquisition of a sample there is a 'spurt' of >>activity that >>produces an image of low fsc vs. ssc on such a plot which within several >>seconds >>equilibrates to a normal appearance of midscale when acquiring normal mouse >>PBL's. >>BD tells us and I agree that this is simply a result of the machine >>equilibrating pressure >>and that most wait a few seconds before acquiring after placing the sample >>on the >>intake port. In fact this also relates to the question that has appeared >>elsewhere on this list concerning >>carryover from sample to sample. It is clear that this fact was not >>apparent on older machines >>because the equilibration had resolved itself before the slower processors >>could pick it up. >>Thus my question is, was it necessary to wait a few seconds on the older >>machines anyway >>because the finalization of the 'data-values' was commited to by the >>cytometer or wasn't it >>necessary because only now with the more rapid CPU does it become truly >>necessary to >>wait a few seconds for equilibration prior to acquisition? >> >>Gene/UCONN Health > >Hi, > >I've only seen this when their was a pressure problem on my FACSCalibur. >When I changed the hydrophobic filters in the side panel and cleaned out >the cuvette the effect went away. > >I used to have the old HP computers running on my FACScan. these computers >had Direct Memory Access (DMA) that allowed the data generated on the >FACScan to be dumped directly into the HP's memory. So what you saw on >screen was real time acquisition of your data. Processor speed is not >necessarily the bottleneck when it comes to data acquisition. I had a >Quadra attached to my FACSCalibur before I upgraded to a G3, and saw a >delay in acquisition. There must have been some buffering going on. The >NuBUS interface in the Quadra operated at 20MHz but only transferred data >to memory at 10MHz. Even at these speeds the memory should be adequately >updated to allow real time acquisition to be displayed. There must have >been some other bottleneck that meant buffering was needed, does anyone >know? > >In answer to your question, I do not believe that the effect that you are >seeing is normal and I would ask more questions to BD about it. On the >HP/FACScan system you did not have to wait and the data was always >available immediately. As for the G3/FACSCalibur, again I see the data >available immediately. However I do agree that the pressure system need >time to equilibrate, but I think a few seconds is too long for your >instrument, I am guessing but I would say a few milli seconds would be >more appropriate. I would recheck the tubing, filters and the cuvette for >blockages. > > >Hope this helps. > > >Andy. > > Andy Riddell PNAC Division MRC LMB Hills Road Cambridge UK tel: (0)1223 402218 fax: (0)1223 412178 email ar3@mrc-lmb.cam.ac.uk
This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:54:20 EST