The trucks used on the E8 are the same ones that are used on American Models' SD60 engines, both of which originally came out in 2001. This means that the wheelbase for the E8's trucks is a bit short, as they should be a scale 14'1". The SD60's wheelbase is 13'7", so that is a difference of 6". Jamie Bothwell measured his E8 model's wheelbase and it is indeed a scale 13'7". As a side note, Jamie also measured his River Raisin Models brass E8 and it also has a 13'7" wheelbase, so it is also 6" too short.
Mike Swederska shares the following things he noted when trying to figure out why his locomotive was stuttering.
Mike noticed that the center wheel, which does not pick up power, did not have much vertical movement due to a shallow spring pocket. The trucks were teeter-tottering and lifting very slightly to the front or back wheel, which are the power pick-ups. His solution was to remove the springs and just let those center axles float, allowing the two front and rear axles to stay in contact with any uneven rail.
The other end of the power pick-up circuit is at the location of where the stud from the gear tower protrudes through the frame tower. He discovered it was black from the factory frame painting. Even though the stud was clean and the surface where the self-locking nut was clean, it was not on the gear tower and the bottom of the frame tower. Mike used a dremel with a wire brush to clean the contact between the two, and then re-attached the two trucks to frame, tightening the two nuts down to the point that the truck swiveled without friction but only had a very slight tilt back and forth.
Chuck Kemper shares a YouTube video on how he installed a DCC decoder into an American Models E8. He takes you through the entire process.
When you order the "open coupler door" part from American Models, they actually send you one piece that is the two doors attached to each in a closed
position. To actually simulate the open coupler doors, you have to cut the two doors apart, which is not trivial. Rich Gajnak offered this advice:
It requires some careful cutting to open the coupler cover. AM didn't mold in a parting line to simulate the two door sections. Rich states that he went
around and around the cover bump with an X-acto knife to, eventually, cut it in half. It's a slow process. Rich also added a scored parting line on the
"closed" cover, to better simulate the separateness of the two doors. Rich seems to recall that he also needed to widen the resulting opening to allow for
coupler movement.
