An essential tool: ESU's Profi-Prüfstand decoder tester
If this project has one essential tool besides the oscilloscope, it is ESU’s Decoder Tester, article 53900, which ESU’s German catalogue calls the “Profi-Prüfstand für Decoder”, the professional test stand for decoders. It arrived a month before any of the RailCom work started, on the simple grounds that a project about decoders needs a way to run one without a locomotive round it. Every bench capture since has gone through it, and it deserves a post.
What it is
A small board with a socket for every decoder interface you are likely to meet, a coreless motor with a flywheel so the decoder has a real load to drive, a row of LEDs that show the headlight, rear light and function outputs, a small loudspeaker for sound decoders, and two terminals for the track feed from a command station or programmer. Plug a decoder in, connect the track feed, and it is a locomotive as far as the decoder knows, with none of the bodywork, pickups or wheels that make a real one awkward to probe.
The one on this bench is the current version, article 53900, which takes NEM 651, NEM 652, 21MTC, PluX22 and PluX16, Next18 and E24 decoders plus bare wires on screw terminals, and has an extension port for the 53901 board that adds large-scale decoders. There is an earlier version, the “Profi-Tester”, with NEM 651, NEM 652 and 21MTC only, which ESU now lists under former products without an article number; its manual on ESU’s download page is filed as article 51900, dated 2012. If you are buying, check which one is in the box.
Why it matters here
The first RailCom replies this project captured came from a ZIMO MN150 that had never been inside a locomotive. It was plugged into the tester, the tester was wired straight to the command station through a sense resistor, and the scope went across the resistor. No track, so no wheel and rail noise, no pickup dropouts, no layout wiring: just the decoder, the station and the signal between them. That is the clean baseline that everything on real track is compared with.
It also makes decoders interchangeable. A ZIMO goes in, a run is captured, it comes out; a DCCconcepts goes into the same socket and gets the same test with nothing else changed. When one of them sends nothing at all, which happened, you know it is the decoder and not the wiring.
The flywheel earns its place too. A decoder’s motor control behaves differently with a real inertial load from how it behaves with nothing on the output, and the speed a decoder reports over RailCom means nothing unless a motor is actually turning. The tester gives it one, quietly, on the desk.
The one thing it does not do
It is not track. Dirty rails, a bad joint and a locomotive losing its pickup for a tenth of a second are exactly the things a detector has to cope with, and the tester hides all of them. So the rule here is: first on the tester, to learn what the decoder sends; then on rails, to learn what the layout does to it. Posts on both to follow.
Tags: bench