LTI GP20 sparking - any suggestions

have a look at the wheels to see if they touch the end of the inside of the side frame or if the flange is touching the sideframe end mounting bracket. If this happens it can lead to shorting out intermittently.

You could take one sideframe assembly off at a time to see which one is the problem, if this is the cause, and see how it runs without them

Thanks banjoflyer.  In the upper photo you can see how close the front edge of the side frames are to the steps.  They were touching in curves, especially S-Trax 29" radius curves which seem to hold the side frames in that position through the curve.  That caused sparks and the motor to stammer and hesitate, especially at slow speeds.  I filed down the bull-nose of the side frames and nicked the corner of the steps with a Dremel.  Painted the side frames gray and the steps blue.  Can't tell.  And it runs great now.  One thing I noticed about this SF engine is that it has a level stance whereas the LTI C&O GP20 I am going to work on next has that annoying high front end low rear end squat that makes it look tilted.  I think Lionel may have inverted the plastic bushing atop the front yoke in the latter case (as in an upside-down T) to lift the chassis up and above the front side frames to clear them.  However, the exploded parts diagrams for these engines shows the bushing should go in as a "T" from the top which achieves a more level stance, but less clearance then between the front steps and the side frames.  And in terms of the need for replacing the shrink tubing on the more modern Baldwin, I've run into that too on the latest NH EP5 (with black side frames and two can motors) of mine and other recent release engines from customers.  Annoying little oversight, but at least not a major repair once you know to look for it.