I had not seen this issue before with trackballs and decided to
visit Logitech's web site, to see if they have any ideas about it.
Here is what I found on their site concerning cleaning their Trackman
devices:
===================================================
 Logitech TrackMan Cleaning Instructions

The following instructions apply to pointing devices with the following
Model Numbers (Model Numbers are located on the bottom of the
unit, following M/N:)
T-CC2
T-CD2
T-CC3
T-CE3
T-SC3
T-AE2 (for Macintosh).

In order to achieve the best possible performance from your
TrackMan pointing device, the following steps should be taken
approximately once every two months. These instructions should
also be followed if your TrackMan is moving poorly in one direction.
For example, it moves well vertically, but not horizontally.

Materials Required:

a. Isopropyl alcohol
b. Lint free swabs, such as "Radio Shack 44-1001 or 44-1094
    anti-static, lintless swabs" or equivalent.
c. Lint free cleaning cloth (such as McMaster Carr number 7340T21
    or equivalent. McMaster Carr phone number is 310-692-5911.)
d. Small rigid or semi-rigid smooth plastic sheet for testing the
    correct amount of alcohol, such as a transparency.
e. Paper clip.

    Precautions:

IF TOO MUCH ALCOHOL IS APPLIED DURING THE CLEANING
PROCESS, IT MAY EITHER REMOVE THE GREASE INSIDE THE
TRACKBALL BASE PLATE OR CAUSE THE GREASE TO
SPREAD AND THEREBY DEFEAT THE PURPOSE OF THE
CLEANING IN THE FIRST PLACE.

See below for the method for ensuring the proper amount of alcohol.

    Procedure:

1. Using the end of the paper clip, turn the retaining ring on top of the
    ball counter-clockwise and then remove the ring and ball. Place them
    safely aside.

2. Clean the ball with the lint free cloth and alcohol. Set aside until
    reassembly. Do not immerse the ball in alcohol.

3. Dip the lint free swab in alcohol. Reduce the amount of alcohol on
    swab by applying it to the lint free cloth or by squeezing the swab tip
    with tweezers. You can use any method for reducing the amount of
    alcohol on the swab as long as the swab and alcohol are not
   contaminated.

Note: To test for the proper amount of alcohol, apply the swab to the
piece of plastic held vertically and draw a horizontal line with the
swab. Apply enough pressure to simulate cleaning the plastic sheet
surface. IF ALCOHOL RUNS DOWN THE PLASTIC SHEET,
THERE IS TOO MUCH ALCOHOL ON THE SWAB. If it does not run
but paints a line the width of the swab, it has the proper amount of
alcohol. The objective is to have sufficient alcohol on the swab to
clean oil, grease, and remove other contaminants but not flood the
trackball base plate with alcohol.

4. Once the proper amount of alcohol is on the swab, lightly rub the
    rollers back and forth to remove any contaminants. Then gently turn
    each roller and repeat. Repeat this procedure to make certain that
    the entire roller surface has been cleaned. Inspect the rollers for
    cleanliness as you go.

5. Clean each of the 3 red ruby bearings inside the ball cavity as
    well, making certain there are no loose particles in the area of the
    bearings.

6. Replace the previously cleaned ball and retaining ring.

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

At 11:03 PM 4/10/99 , Joan Rapier wrote:
>I use a Logitech Trackman Marble pointing device.  I've also used and
>experienced the same problems w/ a Logitech Trackman Vista pointing
>device.
>
>The problem is the trackballs get sticky - so I take them apart and
>clean them using electronic-grade alcohol.  This seems to leave the ball
>and the rubber disks on the x and y axis too dry causing the trackball
>to still be sticky - this time not because there's gunk in there but
>because there's too much traction or grip.  Another problem is there's
>slop in the ball for one of the pointing devices.  So moving it around
>w/ this play coupled w/ the sticking problems is really making for a
>pretty miserable experience when trying to make any precise movements.
>
>I do make sure to align the rubber disks on the x and y axis at the apex
>of the ball curve.  This helps some as it seems they have a tendency to
>shift w/ use - but the problem w/ sticking persists.
>
>I really prefer a trackball but it's looking like I'm gonna have to
>revert back to using a mouse if I can't get this resolved.  Should I be
>using a different type of solvent to clean the trackballs with?  Do you
>have any other suggestions?

               PCBUILD mailing list is brought to you by:
                            The NOSPIN Group
                            http://nospin.com