Stage drifting

[Chen, Jianbo (NIH/NCI) [E]]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hello everyone
We have an encoded (motorized) stage that drifts around 0.5 um (500nm) per hour to mostly but not always one direction. Putting an enclosed box to surround the microscope seems did not help. Is this something unavoidable for a motorized stage? Is using fiducial marker the only solution if experiment requires high accuracy of XY position over long period of time? I would appreciate for any thought and advice.

Jianbo Chen
HIV Drug Resistance Program
National Cancer Institute-Frederick
Frederick MD 21702

[hidden email]
301-846-1841

https://ccrod.cancer.gov/confluence/display/CCRSSSCArchive/Home

[Csúcs Gábor]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Jianbo,

Although it is difficult to remotely solve your problem, but here are
some ideas (based on our experiences with motorized stages).
The first question is whether you really see a drift? Do you stay really
at one location? Or do you acquire images at several locations and
return to them? In this case what you see is just the reproducibility
limitations of your system. Assuming that you see a "real drift": you
stay at one location and the stage continuously drifts, there are a few
things I could suggest to check.

1) Check the electrical connections of your stage and also whether all
settings in your control software are correct.
2) If you have a control joystick - check it (exchange it) - as a defect
(dying) joystick can easily cause such problems.
3) If you have linear encoders - check whether they are still clean.
Dirty encoders can also lead to strange effects
4) a drift a 500 nm/hours is certainly not negligible but there are
software tools to get rid of such movements (even without the use of any
special markers).

Cheers    Gabor

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hello everyone
> We have an encoded (motorized) stage that drifts around 0.5 um (500nm) per hour to mostly but not always one direction. Putting an enclosed box to surround the microscope seems did not help. Is this something unavoidable for a motorized stage? Is using fiducial marker the only solution if experiment requires high accuracy of XY position over long period of time? I would appreciate for any thought and advice.
>
> Jianbo Chen
> HIV Drug Resistance Program
> National Cancer Institute-Frederick
> Frederick MD 21702
>
> [hidden email]
> 301-846-1841
>
> https://ccrod.cancer.gov/confluence/display/CCRSSSCArchive/Home

--
Gabor Csucs
Light Microscopy Centre, ETH Zurich
Schafmattstrasse 18, HPM G41
CH-8093, Zurich, Switzerland

Web: www.lmc.ethz.ch
Phone: [phone]
Mobile: [phone]
e-mail: [hidden email]

[Johannes Helm]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Good evening,

I have experienced this problem earlier on an older motorized stage. The
reason at that time had been a problem with a joystick. The joystick,
which presented one out of two methods to move the table - the other being
software and RS232 connection -, upon starting the system, was
mechanically not entirely in its (0|0) position. It's position was hence
some (x0|y0) at which neither x0 nor y0 were some small values, but not
really 0. Touching the joystick after complete initialization would then
bring it to the (0|0) position, which the joystick, then, interpreted as
(-x0|-y0) and thus brought the table to a slight but steady movement.

Possibly, this is your problem.

Best wishes,

Johannes

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hello everyone
> We have an encoded (motorized) stage that drifts around 0.5 um (500nm) per
> hour to mostly but not always one direction. Putting an enclosed box to
> surround the microscope seems did not help. Is this something unavoidable
> for a motorized stage? Is using fiducial marker the only solution if
> experiment requires high accuracy of XY position over long period of time?
> I would appreciate for any thought and advice.
>
> Jianbo Chen
> HIV Drug Resistance Program
> National Cancer Institute-Frederick
> Frederick MD 21702
>
> [hidden email]
> 301-846-1841
>
> https://ccrod.cancer.gov/confluence/display/CCRSSSCArchive/Home
>

--
P. Johannes Helm

Voice: ([phone] (office)
Fax: [phone] (office)

--
P. Johannes Helm, M.Sc. PhD
Seniorengineer
CMBN
University of Oslo
Institute of Basic Medical Science
Department of Anatomy
Postboks 1105 - Blindern
NO-0317 Oslo

Voice: [phone]
Fax: [phone]

WWW: folk.uio.no/jhelm

[Daniel Murphy]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Jianbo,

You mention that you have enclosed your system with an incubator.  If the
drift you are experiencing is the sort of drift that Johannes is talking
about, enclosing your system will not help at all.  

First, you need to determine whether or not your drift is "real" (drift
caused by the automated stage moving itself because it is sticky) or not
(possibly due to thermal fluctuations causing expansion/contraction of
optical components).  If it is "real" drift, then your automated stage
should report the drift: just check the (x, y) position of your stage at t0
and at t1, before and after you notice a drift.  if there is a change in the
(x, y) position between t0 and t1, then the drift is probably due to a
sticky joystick or poorly calibrated stage.

If it is not "real", and enclosing your system with an incubator did not
work, then you may be out of luck.  You could try an objective warmer.  You
could also account for stage drift in your post-acquisition analysis or
get/write a software program to re-align your stage during acquisition each
time you go to take a new image (much like an autofocus program).

Dan