Quirks


Every once in a while I get a reminder of the good old days and the lingering quirks that have been passed down to us. The other day I was attempting to revive a couple old Lenz decoders that had been asleep for about 15 years. I plugged them into the ESU decoder tester, cranked up DecoderPro, and read back the addresses—103 and 104. I plugged them into a loco and put them onto the track. When I entered the addresses into my throttle I got no response. After checking the wires and the loco I finally put them on the service mode track and checked the addresses again, still 103 and 104. As the frustration mounted I was about to toss these old decoders and then something came back to me.

Well back in the infancy of DCC we only had 2-digit addresses—after all who would ever need addresses for more than 99 locomotives? However, because computers speak hexadecimal we were able to actually uses addresses up to 127. Because the addresses I entered were in the 1-127 range they were interpreted as 2-digit and not 4-digit addresses. And because I had set CV29 to expect 4-digit addresses the system and decoder were not communicating with one another. By simply changing CV 29 to a value of 2 which tells it to expect 2-digit addresses, I was in business.

Another relict of the past are the 2-digit advanced consist addresses. I don’t know anyone who likes them and it would seem more logical to be able to use the 4-digit address on the lead locomotive for the consist address but for some reason the NMRA DCC guys apparently can’t make the necessary changes to fix this anachronism.

Yesterday at an op session, a fellow mentioned that he had entered 4-digit addresses into his decoders and the locomotives wouldn’t run. When he told me the addresses I immediately recognized his problem. He had entered addresses of 9995 and 9996 assuming no one else at the local club would also be using those addresses. So what was the problem?

Well Digitrax, which was the system he was using, only supports addresses up to 9983 not 9999. I believe those address from 9984-9999 are reserved for use by the system. File that little quirk in the back of your mind and avoid those addresses.

That’s all the little quirks I can think of right now, but I’m sure more will pop up. I get several emails every day asking about odd happenings and I’ll add more later.

9 comments

  • I automatically have always assigned 4 digit addresses, even if that means leading zeroes. For instance, Santa Fe had some diesels with two digit numbers, their Erie Builts are an example. Since I always run the ABA together they all have an engine number 0090. And to make sure I do not forget I write the engine number on the bottom with a Sharpie.

  • Agreed about the 2 digit consist NMRA specs. NCE has a work around where you build the advanced consist but you dont have to remember the 2 digit address. Their workaround in their firmware remembers the loco ids you entered and lets you use the lead loco to address the whole consist. And if you switch directions, you can use the trailing – now lead to address the consist. No fumbling with having to remember what was that consist address. The only caveat is you need to break up the consisting if you carry it to another layout.

    Tom Wilson

    • Yes, it is a nice feature the way it works transparently in the background to make advanced consists. However the fact that you don’t know the consist address is a weak point for anyone wanting to move them. But you can make up an advanced consist yourself using CV programming and keep them together that way. Also it keeps track of the lead loco and switches them automatically. I haven’t tested what it does with light, horns, and bells on locos in the middle of a consist though.

      • There is some fiddling with cvs 21 & 22 for consist functions and with cvs 23 & 24 for cosist momentum. I have not played with these enough to really know what it does. But it allows for fine tuning how your consist works as a unit.

        Tom Wilson

  • Don Arnoldus, The Netherlands

    Hi Larry,
    I always have understood, that (only) Lenz works the 2-digit addresses in the range 1 till 99 (0 is for analog locs). The address range 100 till 9999 is with Lenz decoders the 4-digit range. So, for the above mentioned addresses 103 and 104 you have with Lenz devices choose for the advanced 4-digit-range,marked in CV 17 and CV 18 (and of course bit 5 of CV29 is ON). Am I right?

    • I’ll have to blow the dust and cobwebs off my old Lenz files and see what the old info sheets say.

      OK, I found the sheet for the LE103 decoder which is what I was working with and it says the range for 2-digit addresses is 1-127.

      • Don Arnoldus, The Netherlands

        Yes, the range for CV 1 is always according to the NMRA recommandations 1-127. However, Lenz equipment uses automaticly the advanced range 100-9999 by addressing decoders above address 99. So, it’s possible to fill CV 1 with the 103 or 104 according to your example with non-Lenz equipment, but not with a Lenz station.
        For CV 1 Lenz give this information:
        Value / Bit 1-127 Meaning: Basic locomotive address. This number is used to call up locomotives in the Digital plus by Lenz ® system. The use of range 1-99 is recommended for operation with Digital plus by Lenz® devices. When writing this CV, CV19 (multiple traction address) is automatically deleted in the decoder and Bit 6 (use of extended address) is deleted in CV29.

        • Don Arnoldus, The Netherlands

          Lenz is numbering the eight bits of a byte 1-8. Therefore they call bit 6 in CV29 where others use bit 5 (using the 0-7 range). Another quirk!

    • Don Arnoldus, The Netherlands

      Today I got a reply e-mail from Thorsten Gradl of Lenz Elektronik GmbH in Germany:
      “Yes, it is right. Until of the begin of DCC, Mr. Lenz had develooped it, we use the addresses 1 to 99 for CV
      1. Of course our decoders will use 1 to 127 for other systems, but we mean, how you can
      explain a normal user, that 127 is a short address and 128 is a long one. Also we count the bit from 1 to 8.”