A veritable tsunami has swept over my N scale club. Several of our newest members collect Japanese-prototype trains. It’s easy to see why: there’s a dazzling variety of sleek, colorful models available, many of them made by a manufacturer (Kato) we already know and love. The interest is now spreading to our older members, including me. In the wake of my discovery of Usui Pass, I found myself wanting models of the trains that once operated there. When Steve said, “I’m putting together a Plaza Japan order, you want anything?” I said yes.

A 489 series EMU and a pair of EF63 helper locomotives would have been ideal, but they’re unavailable at the moment, so I picked a Kato 485 as a reasonable stand-in. The 485 and the 489 both share the same quirky space-age styling, so unlike anything seen in North America. My new Kato 485 arrived last week in one of those convenient dark-green “bookcase” boxes. Steve got me decoders and light kits, too. I’ve already talked about Kato light kits, they’re the same ones I put into my 20th Century Limited. Let’s talk about the decoders.

A typical Kato EMU set requires 3 decoders–one for the motorized car in the center of the train, and one for each end car to control head and tail lights. All three are drop-ins. Just pry off a little hatch cover on the underside of the car, slip the decoder in between the contact bars, and put the cover back on. If you like the way Kato built your California Zephyr, you’re gonna love these things.

The motor car was the easiest decoder install I’ve ever done. The Kato EM13 decoder has the same ID code as a Digitrax DN163. It’s actually manufactured for Kato by Digitrax, and I’ve worked with a lot of Digitrax decoders. Pop the hatch, slot the board, put the ol’ PowerCab into programming mode, run through the addressing prompts, and it’s done. For want of an obvious road number, I set the address to 485, the same as the big numerals on the box. The motor car moved back and forth on the very first test. That’s better than most of my North American Kato locomotives did.

Installing the FL12 function decoders for the end-car lights was almost as easy. The only tricky part is orienting the two boards in opposite directions from each other, so that one displays headlights when the other lights up the taillights. No other wiring or configuration is necessary. Clever, and yet…

…also stupid, once you get to addressing the things. I tried several times, but couldn’t get a long address to take. I also tried operations (programming-on-main) mode, and that didn’t work, either. The FL12 is supposedly similar to a Digitrax TF4 function-only decoder. I’ve installed one of those, as it happens, and its instructions explicitly state that it can’t be programmed in operations mode. What to do?

The FL12 instruction sheet is entirely in Japanese. An English translation can be found deep in Kato USA’s website, but it isn’t much better. A perusal of forum posts suggested that perhaps the FL12 wasn’t even capable of long addresses, which didn’t make sense. Long addresses have been part of the DCC standard for a very long time now. Further digging revealed a possible solution.

Before we get to that, let’s briefly review how DCC addressing works. A short address is stored by a decoder in CV1. In the early days of DCC, this was the only form of addressing available. Long addresses were added to the standard later, to broaden the available address space. The 4-digit road number of any locomotive in North America can be used as a long address. It requires two bytes, so components of it are stored by a decoder in CV17 and CV18. One bit in CV29 is used to specify whether the decoder responds to the short address in CV1, or the long one in CVs 17 and 18.

Normally, your command station does the math for you when programming a decoder address. CVs 17, 18, and 29 are calculated and written in one shot. For some reason, the Kato FL12 won’t play this way. (“Hmm…ATTiny,” said Steve as he inspected the board. “First decoder I’ve seen that didn’t use a PIC or SAMD chip.” Perhaps that has something to do with it.) You’ve got to spoon-feed it the CVs one at a time.

It was worth a try. A long address of 485 breaks down into values of 193 for CV17, and 229 for CV18. (Here’s a handy online calculator to take the pain out of the math.) The value for CV29 is easy enough: for most of my roster (also configured for long addresses), it’s 38.

So, back to the PowerCab. Prog/Esc, Use Program Trk, 2=CV. Enter CV Num=17. Wait. Can Not Read CV. Enter Value=193. Enter CV Num=18. Wait. Can Not Read CV. Enter Value=229. Enter CV Num=29. Wait. Can Not Read CV. Enter Value=38. Prog/Esc. Select Loco: 485. And…

It worked. Lights responded to the direction button, and F0, just like I expect them to. Yay. The next day, I brought the 485 to our Christmas-week exhibit at the library, where I found it to be an ideal show train. It comes out of the box and onto the rails quickly, runs beautifully, is bidirectional, and attracts almost as much audience attention as Steve’s bright-pink Hello Kitty Shinkansen. Sure, it strays pretty far from my professed early ’60s Pennsycore aesthetic, but I’m enjoying it.