Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Atlas Signal System Major Problem - All Lights Red in Integrated Setup

3562 views
7 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    May 2008
  • 4,612 posts
Posted by Hamltnblue on Sunday, September 5, 2010 7:33 AM

NCSX,  can you please break your posts into paragraphs or numbered points.  I rarely comment on it but these eyes cant follow what you're trying to do and I've been in electronics a long time.  Looks like all good the info but one long paragraph just makes it hard to follow. 

I think you'll get more good responses if you go back and edit the posts real quick.

 

Springfield PA

  • Member since
    July 2006
  • From: Vail, AZ
  • 1,943 posts
Posted by Vail and Southwestern RR on Sunday, September 5, 2010 6:06 AM

The open collector outputs are working just the way they should.  If you tie the outputs of multiple devices together, when one pulls low, all of the outputs will be low, they are the some point, after all.  That's how a wired AND words.  You could easily make a circuit to combine the outputs of multiple detectors without interference.

Jeff But it's a dry heat!

  • Member since
    July 2006
  • From: Missouri
  • 132 posts
Posted by nscsxcrrailfan on Sunday, September 5, 2010 12:48 AM

Earlier today, I did some more experimenting and discovered something new.  I was told that the BD20 is an open collector design and that it would be o.k. to connect two or more detectors to one signal board.  Well, that doesn't work.  I had three detectors connected to one signal control board, and if one of the detectors detected the resistor, all the detectors would trigger.  I couldn't have three detectors connected to one control board and also have one of those detectors connected to another control board without it interfering with it.  Basically, if the resistor for the first detector was on, all three detectors would act as if they were detecting something, even though only the first should be detecting the resistor.  When I connected the third block detector to a signal control board, it caused the signal to go red, even though the third detector should not have detected anything.  I can use one detector to communicate to two or three signal boards without any problem, but not the other way around.  

  • Member since
    July 2006
  • From: Missouri
  • 132 posts
Posted by nscsxcrrailfan on Saturday, September 4, 2010 12:27 PM

O.k., first, when I tried this, it was late last night (about 11:30pm).  I did this after a long day of classes (I'm in college).  It's always a bad idea for me to try and do deep thinking that late at night after being in classes all day.  So, I should have checked the cable that I thought was a crossover cable more closely.  It turned out that it was not a crossover cable, even though it was in the pile of regular phone cables.  So, I cut a new cable and made sure it was a crossover (I followed the little indented line down the cable and made sure the plastic connectors on each end went opposite directions of each other rather than similar for a straight through cable).  I plugged that into the signal board and everything worked fine!  I made some more cables and tested it, and they worked fine too.  So, thank you for your help, but I should have checked that cable closer.  However, I have noticed something interesting.  If a train (resistor in this case) moves into a block and then reverses and moves back out of the block, the yellow 8-second timer still works.  Rather than go from red to green, it goes from red to yellow for 8 seconds before going to green.  However, in other situations, the signals work just fine.  Any time a red signal goes to yellow because the train moves on down the track, it goes to yellow and stays there.  Then, once the train moves on to the next block, the yellow immediately goes to green without the 8 second delay.  I guess the 8 second delay is retained for added safety purposes when you're reversing out of a block.  Also, here's a question that I asked on here a few months ago and no one answered.  The 9V DC transformer I'm using has a limit of 300ma (no overload protection).  Do you think I'm at any risk of overloading it using these signals, signal boards, and detectors?  What if I have 20 signals and signal boards, will it overload it?  There is one last question I have.  Do you think the BD20 could detect DC or AC current?  I'm curious because I've often thought about the possibility of switching the layout back to DC in rare cases where a friend might bring some DC locomotives over (I don't like the analog feature of DCC)?  It'd be nice to be able to use the signal system in DC, but I guess I could go without it if I had to during those rare cases.

  • Member since
    July 2006
  • From: Vail, AZ
  • 1,943 posts
Posted by Vail and Southwestern RR on Saturday, September 4, 2010 11:08 AM

It sounds like the BDs are working.  I'm doubting that the boards are bad, at least not all three of them.  Static discharge is real, but most things are not very sensitive to it.

I did some Googling, I could not find an indication of the wiring of the signal cables.  There are people on the Atlas board that have it working, with the Atlas cables, of course.  

To make the experiment easier, you could look at the output of one board with a voltmeter, with the block occupied, and not, and see where things live.  Then you could just mess with two wires to figure out which lines are the two before and one before inputs.

Jeff But it's a dry heat!

  • Member since
    July 2006
  • From: Missouri
  • 132 posts
Posted by nscsxcrrailfan on Saturday, September 4, 2010 10:38 AM

Yeah, I tried breaking the cable already, but I don't think I've tried every configuration yet.  I'll try that.  It'd be just like Atlas to design the wire configuration of those a certain way to prevent people from making their own cables.  Also, the jumpers are set to their default positions; they're not supposed to be moved to set it up the way I'm wanting.  Now, I think it's possible the boards could be bad.  I just pick them up and move them around without taking any anti-static precautions, but I'm on concrete floor.  I've never discharged any static off onto them.  So, I really doubt that could be the issue.  However, the SCB is designed to handle 6-22V AC or DC, but the BD20 is designed to handle only DC 5-12V.  For a while, I didn't realize the BD20 required only DC and ran AC 18V to it.  They acted really weird, and that's when I realized AC wasn't supposed to be used and switched to a DC source.  I wouldn't think that would cause this problem, but it's possible I guess.  If I can't get this resolved, I'll have to contact Atlas.

  • Member since
    July 2006
  • From: Vail, AZ
  • 1,943 posts
Posted by Vail and Southwestern RR on Saturday, September 4, 2010 6:52 AM

My guess is that the issue is with your signal cables.  Unfortunately, the manual doesn't give enough clues to figure out what the cable is.  I'd be tempted to make a setup where I could change the cable configuration easily (break it in the middle, make an easy way to spice it together), and experiment with two of the board, since there are not that many combinations.  I would have though it was either straight-through or crossover, so I am a bit puzzled (it makes me wonder if maybe one of the cables you made up is faulty).  I guess the other option would be to get one Atlas cable, and see what it is.

Also, I'd make sure the jumpers are right, though I don't see a mode that sounds like what you are seeing.

Jeff But it's a dry heat!

  • Member since
    July 2006
  • From: Missouri
  • 132 posts
Atlas Signal System Major Problem - All Lights Red in Integrated Setup
Posted by nscsxcrrailfan on Saturday, September 4, 2010 1:24 AM

Hi, I'm having a major problem with some Atlas signals I purchased about a year ago.  I don't have the signals wired into my track yet, I'm just testing them out using resistors in place of track and locomotives. First, let me describe how I have this set up.  I'm using Digitrax DCC.  To power the signal system, I have a 9V DC Enermaxx transformer (When I first set this up, the system wouldn't power on at all.  I got the signals to light up by switching the common and positive wires on the transformer).  For Block detectors, I use NCE BD20 detectors instead of the Atlas ones.  The block detectors are properly wired with the signal boards (I went by the diagram that explained connecting the BD20 to the Atlas signal system).  Each block detector has a wire from the positive side of the DCC main bus looped through it 3 times that goes to a resistor.  The common wire run from the resistor back to the DCC main bus.  I have a switch that I turn on and off for each resistor (3 switches for 3 resistors).  There are three signals set up, one right after the other.  Each signal operates fine when they are not integrated with the next/previous signal.  For example, the first signal will go red if I flip the switch for the resistor on.  The block detector for the first signal will detect the resistor and drop the signal to red.  The same goes for the second signal and the third.  With this setup, the signals don't know what the next/previous signal is doing.  After turning the resistor switch off, the signal goes to yellow for 8 seconds before going back to green.  THE PROBLEM:  The problem is when I try to integrate the signals together so that if the last signal is red (resistor switch on), the signal before it will be yellow, and the signal before that will be green.  The same goes for the second (middle) signal.  If the resistor switch is on, it should go red and the first signal should go yellow.  The last signal should stay green because the resistor switch for it is off (basically, it is not occupied).  HOWEVER, this is what the system does.  If the last signal's resistor switch is on, the last signal goes red plus the other two before it.  The one before it should be yellow and the one before that should be green; instead they're all red!!  The block detectors for the two signals before the last one where the resistor is being detected don't indicate anything being detected even though the signals drop to red (the little red light on the BD20s does not light up on the first and second signals; it lights up properly on the third signal).  For the second signal, if I switch on the resistor switch, it drops red but the signal before it also goes red (it should be yellow, also the BD20 light indicator for the first signal does not light up even though this signal is dropping to red instead of yellow).  For both the last and second signal, when I flip the resistor switch off for those signals, the signal before it goes from red to green quickly, but the signal where the resistor switch was switched on and is now off goes from red to yellow and then to green; it acts as if the system is not integrated and the yellow timeout feature is still active even though it shouldn't be.  In integrated setup, the yellow timeout is not supposed to be active.  Now, at first I thought it was the telephone cable I was using to connect one signal board to another.  I don't have the Atlas signal cable so I'm using a homemade one (I have a very good crimper).  I tried two straight through cables (RJ11 4-wire) and the problem still persisted.  I tried a crossover cable (regular 4 wire telephone cable) and the problem still persisted.  I tried a straight through RJ12 cable (even though the outer two wires aren't necessary) and the problem still persisted.  Then, I took the long hard route and cut into the telephone cable and combined the black with the red and the yellow with the green; the problem still persisted.  I even tried hooking up the telephone cables exactly opposite of what Atlas said to, with the previous signal's cable going to J1 and the outgoing cable coming from J2.  That didn't help at all.  I've tried all sorts of tricks trying to get this to work right, and it will not work right in the integrated setup.  Anybody have any thoughts on what to do?

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!