If you're keen....
Dear Ernst,
I'm not going to weigh into the "should you accept such performance out-of-box" or not argument,
that's a personal decision, which has many factors to consider.
I'd only make the comment that I have bought 7 x "current spec" Athearn SW1500s over the last 18 months,
all have been bulletproof on analog DC control,
smooth, slow, and sure-footed.
(I also have 2 original versions, one which is a good runner except for it's inefficient motor,
the other has always had a strange rear truck axle-alignment issue which I am yet to 110% nail...)
I would also make the point that in any debug-situation, "simpler is better".
it only takes a length of flextrack, and a regular 9VDC alkaline battery to test/diagnose an "analog DC" locomotive.
However, for the purpose of todays exercise, I'll assume that you've got your reasons for insisting/sticking with diagnosing the issue with the decoder/DCC-system "in circuit".
For myself, I had a close look/listen to your video, and if you are willing to _have_a_look_Yourself_,
(with the full knowledge that it _may_ void warranty), I'd be interested to know the following:
Q1 - Do you have the decoders in _both_ locos set to TCS factory defaults?
Q2 - I ask Q1, because the RED SW1500's headlight appears to be flashing in a regular manner, which is unlikely to be a "decoder-driven-led VS camera scan-rate" issue. This _may_ indicate
- a wiring issue (did you remove the JST jumper PCB as per the Athearn instructions?)
- deliberately wiring the headlight to a "flasher" function wire
(which would also presumably mean you've tweaked some CVs in the decoder, see Q1 above)
- or possibly mean you have a speed-step mismatch between the decoder and the DCC command system
(seem to recall some decoders will "flash" their headlights if there is a Speed-Step mismatch, because they are "hunting" between valid values???)
Q3 - I hear a low "whirr" noise, which may be the flywheel whipping/glancing past the interior of the shell,
or part of the mech. Does it make the same noise/have the same performance when the shell is OFF?
Q3a - If NO to Q3, check for the motor being mounted slightly 'tilted' off-centre,
any scrape marks on the inside of the shell,
or any wiring which may be getting push into interferrence with the rotating mech parts when the shell is in place...
Q4 - Based on the apparent "regular bind/stalling" evident in the slow L> R pass in the video starting at 1:25,
without disassembling the shell, you can
- tip the loco upside down in a service craddle,
- unclip the truck cover plates (do one truck at a time),
- remove the wheels/axlesets from the truck,
- and apply power.
With ONE of the truck's axles removed, does the sound/performance improve?
If YES, you may have found a binding component. Check for flash om the gears.
If NO, replace those axles, and do the same test with the _other_ truck's axles removed.
If Still no joy after testing _both_ trucks seperately, its unlikely to be specifically axle-related.
Q5 - with all 4 axles removed, apply _LOW_ power/low speed setting,
and note if either of the truck assemblies "kick" as the internal truck gearing moves.
A "kick" in either truck may indicate a geartrain position which is creating a temporary bind.
If you _really_ want to dig into it, and only ONE of the trucks is "kicking",
carefully disassemble that truck and inspect all gears and the top worm for flash or other casting issues...
From here, you could give up, or delve deeper,
depending on whether you are "mechanically inclined" or not.
Basic loco mech debug procedure is not particularly hard, if you apply a methodical approach.
Not being afraid to revert to regular analog DC control for diagnostic purposes at the least,
is also a key "ace-in-the-hole" ...
Hope this Helps,
Happy Modelling,
Aim to Improve,
Prof Klyzlr