PMT Gains on Digital Cytometers

Cheryl Smith cheryla.smith at utoronto.ca
Wed Jan 25 09:55:19 EST 2006


I'm posting a message from one of our local users, Mark Luscher, who wanted
to contribute to the discussion of PMT Gains on Digital Cytometers. He's had
some problems subscribing to the group, so I thought I'd forward this along
while he works out the bugs...

-------------
 Hi All,

I've been grappling with these issues of PMT voltage settings, in something
of an improvised way since I have not seen the BD application note and don't
have anyone locally who has done the experiment.

I have looked at it twice now, once on our FACS Calibur, and once on a BD
LSR II.

Here's the experiment I've done:

I ran UltraRainbow Calibration Particles
<http://www.spherotech.com/new%20downloadable%20notes/ultrarainbow%20calibra
tion%20particles.pdf>  on each instrument and looked in each channel
simultaneously over a range of voltages. I set up 2D plots (not shown) to
uniquely identify each of the bead populations, then plotted the median of
each bead population vs. voltage. Error bars are shown as two standard
deviations each side of the mean.

I also calculated the stain index (see for example BD Newsletter
<http://www.bdbiosciences.org/pdfs/newsletters/04-790030-3A1.pdf> ), which
is an indication of resolution of adjacent fluorescent populations, for each
bead population compared to the next lower population.

Here are pointers to the data:

For the FACS Calibur
<http://mlg4-6366.med.utoronto.ca/%7Eluscher/FACScalibur%20data.zip>
For the LSR II 
<http://mlg4-6366.med.utoronto.ca/%7Eluscher/LSR%20II%20data.zip>

First, a couple of comments. In several of the plots, the CV's go a bit
crazy for the highest and lowest staining beads (depending on voltage). That
should be ignored. It's just the effect of cells bumping up on the axis.

The CV's generally get better at higher voltage. I have not yet run the
PMT's up to the top to see if I start to see additional noise from
overamplification, but it appears that up to 700 (LSR II) or 850 volts (FACS
calibur), more is better. (Obviously, this only applies if your most
positive cells remain solidly on scale).

I don't see spikes like Robin Stingley saw, but obviously I'm doing a
different experiment.

The Stain Index data are interesting, because stain index may continue to
improve with increasing voltage even when the CV's don't change too much. It
seems to me that the ability to resolve populations is the best test of
instrument resolution. Any comments on that issue?

Aside from inviting any comments on the methods and the data, and the
general issue of whether I'm on the right track here, I have two questions.

First, do you think I'm correct that we should be sticking with a voltage on
each PMT that is well into the flat portion of the CV graph, usually around
700 V in the experiments I've attached here (range 500-800) except when we
have to turn it down to keep our stained cells on scale?

Second, if you take a hard look at the Calibur and the LSR II data,
comparing similar channels, I think that the CV's on the LSR II are somewhat
higher. It turns out that the LSR II is not on service contract and does not
have a QC program at the moment. Is this an indication of a problem with the
machine, or am I somehow comparing apples and oranges when I look at CV's
across machines (same bead prep)?

Thanks for any assistance you can offer,

Mark Luscher

NOTE ADDED: I just obtained a copy of the BD PMT setting application note
from Robin Stingley (THANKS!). I see that the note shows one thing that I'm
seeing too -- namely that CV's get better as voltage gets higher, until they
reach a plateau. There's no indication of amplifier noise at high voltage.
It makes me think that we should set every PMT at 800V or so, then run a
fully stained sample (uncompensated) and back down the voltage on any
channel where there could be cells bumping up against the top axis. Once
everything is solidly on scale, we'd go back and determine target channels
for our control beads, and then we'd go ahead with the rest of the
experiment (compensation, FMO's, etc.). Is this out to lunch? It's backward
from the 'old' way (bring your negatives on scale), but it makes sense to
me...

------------------

From: Stingley, Robin L <StingleyRobinL at uams.edu
<mailto:StingleyRobinL at uams.edu?subject=RE:%20Optimizing%20PMT%20Gains%20on%
20Digital%20Cytometers> >
Date: Mon Jan 23 2006 - 09:56:37 EST
I've tried it two more times, and got spikes at different voltages each
time.  The third time, I got two spikes in the same run.  I'm trying to
use them all to get an average for the correct PMT settings.  I don't
really understand what's happening.


Robin


Robin Stingley, Ph.D.

Flow Cytometry Core Facility

Department of Microbiology and Immunology

University of Arkansas for Medical Sciences

Slot 511, Rm B504

4301 West Markham Street

Little Rock, Arkansas 72205

Phone: 501-686-6927

E-mail:  StingleyRobinL at uams.edu

http://www.uams.edu/flowcytometry/ <http://www.uams.edu/flowcytometry/>


________________________________

From: Carol Oxford [mailto:cloxford at ucdavis.edu]
Sent: Thursday, January 19, 2006 6:55 PM
To: cyto-inbox
Subject: Re: Optimizing PMT Gains on Digital Cytometers


That's really interesting- I've done it twice now, and both times my
curves look very similar to yours, although several of the parameters
that started out with lower CV's didn't show the spike.   I was about to
send the data to BD for some advice- I'd be really interested to see
what data other labs are getting.


Carol


------------------------------------------------------------------------
---

Carol Oxford	   

Manager, UC Davis Optical Biology Lab

Tupper Hall, rm 3425

University of California

Davis, California  95616

(530) 752-7205	   

(530) 752-4548 fax 

cloxford at ucdavis.edu

http://ccresources.ucdmc.ucdavis.edu

------------------------------------------------------------------------
---




 On Jan 17, 2006, at 4:09 PM, Stingley, Robin L wrote:





Dear Flow Experts,




Thanks so much for all the help on my questions about using auto comp

and the other features of FACSDiva.  Many of you mentioned that I should

establish the optimum baseline PMT gains using the BD application note

for digital cytometers to help with the rare and dim populations of

cells that we are trying to analyze.  I've just tried the protocol, and

my performance curves look weird.  I've attached a PowerPoint slide of

them.  There's a spike in the %CV of each fluorescence parameter at PMT

voltage 600.  




Should I continue on, or is there some kind of problem either with the

way I set this up or with the instrument?




Thanks,


Robin




Robin Stingley, Ph.D.


Flow Cytometry Core Facility


Department of Microbiology and Immunology


University of Arkansas for Medical Sciences


Slot 511, Rm B504


4301 West Markham Street


Little Rock, Arkansas 72205


Phone: 501-686-6927


E-mail:  StingleyRobinL at uams.edu


http://www.uams.edu/flowcytometry/ <http://www.uams.edu/flowcytometry/>







========================================================================
========================


Confidentiality Notice: This e-mail message, including any attachments,
is for the sole use of the intended recipient(s) and may contain
confidential and privileged information.  Any unauthorized review, use,
disclosure or distribution is prohibited.  If you are not the intended
recipient, please contact the sender by reply e-mail and destroy all
copies of the original message.

========================================================================
========================

This attachment - 'Baseline settings 011706.ppt' -  62.46 KBytes - can
be viewed at

http://www.cyto.purdue.edu/MD-parts/d6f3c3be0d8e0560a89c59259895ecf0070e9e40
.ppt





------------------------------------------------------------------------
---

Carol Oxford	   

Manager, UC Davis Optical Biology Lab

Tupper Hall, rm 3425

University of California

Davis, California  95616

(530) 752-7205	   

(530) 752-4548 fax 

cloxford at ucdavis.edu

http://ccresources.ucdmc.ucdavis.edu

------------------------------------------------------------------------
---










More information about the Cytometry mailing list