How does the new HO Atlas 'Classic' RS-3 shape up as far as converting to
DCC? (I'm referring specifically to the latest bunch that are shipping now,
although I assume that the earlier 'Classic' RS-3 from 1999 and 2001 are the
same under the shell.)
Does it have a standard NMRA DCC plug?
Do the headlights need to be replaced with 12v bulbs?
Any other issues?
[The 'Road Foreman of Engines' is getting a little ticked at an intermittent
electrical short somewhere in his current power and is thinking of an
I haven't seen this particular release, but Atlas Classics always have an
8-pin NMRA socket, and they've never required you to replace the bulbs, but
that's not always a good thing, because they're also known for very limited
output. You may not even be able to replace the bulbs because resistance is
hard-wired into the circuit board.
May main complaint with some Classics is that you have to remove the entire
shell to get at the DCC socket. You have to be very careful, or you could
mangle the handrails. To me, having them fully assembled only benefits those
you are going to run them in analog mode.
What are you using to open them with, a can opener?????
I took mine apart to check what decoder I used (its been awhile since
they were installed, plus it only took a second) and Whoa, surprise!
It's an NCE DA102! No 8 pin plug, no 1.5v gow bulbs. Just the same NCE
decoder that fits Atlas, Stewart, Kato and others with 12v bulbs. And
YES, they are Atlas Classics!! I think someone must be confused with
the Kato RS-2 that uses a plug-n-play decoder with factory installed
led's. I have some of them also!!
A DCC hobbyist shouldn't have to dismantle a loco just to get at an 8-pin
socket! I don't think you speak for the masses out there, who we need to
convince to go digital if this is going to become as ubiquitous as analog is
I personally think that many people are just a little bit intimidated by
having to take apart a new loco, especially when Atlas gives so little
support in the way of instructions.
I spoke with an Atlas rep about a year ago - asked him why they hadn't
continued to pursue the removable hood access cover they had implemented in
one of their previous Classic releases, and he was unaware of such a thing!
How can they possibly know what we want when they don't even know what
And the fact that they have Lenz manufacture decoders that are supposed to
default to silent running, but actually default to precision speed - I
wonder how many buyers have never been able to see a difference in noise
levels between an analog Atlas and one of their #342-equipped silent
Stop defending Atlas, just because you have a gift - one that not all of us
What would you prefer? A pull out drawer? Maybe in the underbelly? We
could mill out some of that below the center of gravity weight in the
fuel tank, and put it there? Nevermind the loss of traction and
instability it would cause, it's eeeeaassyyy!!!!!!!
Kato, Stewart, and Atlas engines have been like this for many years.
Why is it a mystery now? Firmly grasp the front hood between your
thumb and forefinger, and pull up! Same thing with the rear.
I assume you are referring to the dynamic brake blister on the GP38'
and 40's. Probably the worst design feature on the engine. VERY damn
few of them are on straight, and can easily be put crooked by light
handling. I made the mistake of picking up one by the blister,
fortunatly it only fell about a foot back to the layout and was not
Couldnt tell ya. I swapped the plugs on the way home, never ran them
I wasn't defending Atlas, I was merely shedding some light on
inaccuracies in your prior post! Maybe more folks would switch to DCC
if they were not mislead along the way.
It was the GP40 - the last one to be released in the Classic line. I had
catalog #8910. The rep swore up and down that they NEVER had a removable
access cover in any of their Classics! When I gave him the part number of
the one I had, I never heard from him again.
I was just guessing, but a guess isn't good enough. I just reviewed all of
the information on the Atlas website for every release of the RS-3, and
there is nothing stated about an NMRA socket. That doesn't mean the latest
release doesn't have one, though. Sorry I couldn't be more helpful. I guess
you'll have to find someone who has already bought an RS-3, or try
contacting Atlas direct...
Go back and look at your 1st post. In that you clearly state
(without saying I think, or maybe, or IMHO) that the Atlas "classics"
line has always included an 8 pin plug. That got me thinking, because
I did not remember any of my RS units having 8 pin decoders other than
my Kato units. It took me all of 3 minutes to remove the couplers and
shell to look at the decoder inside.
So what do I do now. Let another modeler go buy one of these
things only to find it doesnt have a plug? Or do I respond? I saw this
one heading straight for a heartache.
Ok, I'll admit that my response was a bit sarcastic, and for that
I'm sorry. I too would like to see as many people embrace DCC as you
do, and I know the surest way to get someone to go buy a new DC power
pack is to piss them off over DCC, or to confuse them. Truth was my
No inaccuracies, no misleading, and no ego. Since you brought
this up, exactly how large does your ego have to be to have a tag line
like "Digital Railroader" after your name? Is it something you just
make up, or is it a title you earn?
Yeah, you guys have to be gentle with me -- especially on DCC matters. I'm
not sure, but I may have fried a decoder while programming a simple
plug-n-play installation tonight.
Thanks for posting the accurate information on this loco -- I always
- Mark Mathu
Whitefish Bay, Wis.
OK, I'll admit that I should have put "IMHO" in front of my original post.
My intention was not to mislead, but my lack of experience with RS3s should
have kept my mouth shut. I also apologize. BTW, I promise to stop the
name-calling if you will do the same.
Fried during programming? Programming on the main or on the programming
track? That would surprise me. Even if one puts the 8 pin plug in
backwards, the design is such that they will not be damaged. In fact
with a backward plug the decoder will still work, just backward. As far
as I know, the only way to fry a decoder is if you put track power into
the output wires, which should not be possible with a plug-n-play
installation. I've even put DCC equipped locomotives on the track with
other command systems (CTC type) with no damage.
By any chance were you dealing with CV29 when the problems began?
"Fried" may be an overly-harsh term; "scrambled its brains" may be more
Here's my attempt at explaining it:
Loco: HO Atlas Classic C424
DCC system: Wangrow System One
I was programming on the programming track, which is a regular part of my
layout wired with a DPDT center off toggle switch to the programming output.
My first step was to get the loco on the programming track, and try to assign
a long address. I used the option "1" of The System One programming menu,
which basically sets the standard CVs (short address, long address,
acceleration, deceleration, starting voltage) through a series of prompts.
The system seemed to be identifying things properly - it showed the
manufacturer number and decoder version at the start of programming. The
final prompt is to "SET CFG?" which I believe programs CV29. The first time
through I inadvertently selected NO, which I believe should just leave CV29
at the default (which is 2 for the D14SRP decoder).
I quickly realized my error when I switched the toggle to mainline power and
the loco wouldn't budge under command sent to the long address. So I
switched back to the programming track and went through the steps again, this
time remembering to reply YES to the "SET CFG?" prompt. I thought I was
doing everything right, but again, no response when I went back to mainline
power. No light, no movement.
Not long afterwards -- I can't recall exactly when, I spent three damn hours
trying to figure the thing out last night - the decoder would come up with
"CAN NOT READ CV" for most, usually all, of the settings, starting with the
manufacturer number and decoder version -- which are the first two things the
System One attempts to read.
So I eventually went to the System One programming option to directly set
individual CVs. I went through all sixty CVs (several times, actually)
trying to reset them to the factory defaults. It was very hit or miss.
Sometimes I would go back and try to set the same CV again immediately
afterwards; sometimes the System One would show the value that I had just
set, other times it was a completely different value, and most often it was a
"CAN NOT READ CV" reply. I tried directly setting CV17/CV18 to the long
address (#313 = 193/57, but I'd get different numbers back if I tried to
reset the values again in the programming session); I tried using the System
One to reset all the decoder defaults, I tried a Brite Boy on the programming
track, alcohol on the wheels... but I never could get a consistent repeat of
the results. Sometimes when I would start programming again the System One
would correctly identify the manufacturer number and decoder version, most
often it was the same "CAN NOT READ CV" message... and I'm pretty sure that
CV7 and CV8 aren't even read-write locations in the memory.
My wife finally prompted me today to go to the hobby shop and ask for help
(the next thing you know, she'll be telling me to ask for directions when I'm
lost while driving). I bought a new Digitrax DH123P and the DCC guru at the
shop attached and programmed it in under two minutes. And he was using an
NCE system which is basically the same as the Wangrow... so I followed his
steps and I *thought* I was doing the same steps; go figure.
By the way... Another way to fry a decoder is to install one in a Life-Like
GP30 and not include the insulation (double backed tape or similar) between
the decoder and the circuit board. Don't even ask me how I know that.
Yeah a good chance. I think that is one of the things that the System One
tries to set using System One's option "1" to set standard programming
options. I tried to set CV29 back to the factory default a few times, but
that didn't clear things up... it was usually a "CAN NOT READ CV" reply the
next time I tried to change any of the Cvs, inclduing CV29.
Although I feel that I understand the basics of DCC, I seem to have fallen a
little short in the execution.
Whitefish Bay, Wis.
It sounds like you do have a "scrambled" decoder. Life is full of
little mysteries, and some of them come on a printed circuit board.
I've has a few do really stupid stuff for no reason at all.
I would recommend packaging that little gem up, and sending it to
Jim Scorce at NCE for replacement. I would send a letter with all
that you did to try and get it working, and ask that they replace it
under warranty. You should also mention that you have already replaced
the decoder and the engine runs. You may want to not mention the "D"
word. No sense rubbing salt in a wound.
Mark Mathu wrote: (whole series greatly abbreviated)
(detailed description deleted)
It sounds like you did everything properly. It could just be a bad
decoder. Did the unit run on default channel #3 before you started
programming? I've got a LENZ 1014 that the system cannot recognize. It
gives an error that means "decoder not found" although I know it is
there and getting power. I have a friend trying to get me to "trouble
shoot it", but I'm a "back to the store" kind of person.
The only other thing I can think of would be that possibly one of the
values overflowed. That is, the valid range is 1-24 and somehow a value
got set to 27. At least on LENZ, if one does this (especially with a
throttle) the whole thing can go wacky.
Yup, been there done that (and that was back when decoders were $60
each). I had one where the power take off wire from the truck rubbed
through the insulation. Went around a tight corner and the, now
uninsulated, raw track power wire shifted up and touched the internal
part of the circuit board. Major smoke, the locomotive still smells
like fried electronics.