N Scale Signal System Questions

[C855B]

A bit off topic for this thread but just for the record: after polishing the FVM treads, I am getting highly reliable detection with them.  Now I'm buying up FVM wheels right and left. 

Good thing, then. If you were buying them up right and right or left and left, the flange would be on the wrong side of one wheel and it would derail.

 

[peteski]

A bit off topic for this thread but just for the record: after polishing the FVM treads, I am getting highly reliable detection with them.  Now I'm buying up FVM wheels right and left. 

While the suggestion of polishing the treads worked, I would still be cautious about long term results.  And to me it would also make sense to have more than one resistive wheelset in the train.

[GaryHinshaw]

Definitely agree.  That's why I'm buying them up.

[nkalanaga]

Mastertech:  Basically you have CTC without block signals.  No reason the railroads couldn't do that if they had enough passing tracks.  On the prototype, CTC usually controls the entrance/exit signals for the passing tracks, and the signals along the single track in between are ABS, controlled by the track detection circuits.  If the single track is short enough, there's only one block, so no need for the ABS.

A few extra relays and one can interlock the exit signals at one siding with the opposing signals at the next siding, so that the operator can't clear both directions at once, and it would be a fully functioning manual CTC system.

[Cajonpassfan]

Good to know Gary, thanks.
Otto K.

[rrjim1]

Another Option is to use Infrared Model Train Detection, this way you don't need any resistor wheel sets, but you do need a good light source.

[peteski]

Another Option is to use Infrared Model Train Detection, this way you don't need any resistor wheel sets, but you do need a good light source.

Well, if you go with infrared then you usually install its own light source (infrared LED) on the opposite side of the detector.  Or if you go with visible light detector (which is not as convenient), you depend on the ambient light.

Then of course you can have detectors which instead of relying of the rolling stock obstructing the light source, you can have detectors which reflect light off the rolling stock to sense their presence.

But optical detection (active or passive) is also not without problems.  Stray light can give false readings and as you said, you always need some source of light (whether it is ambient light or an LED on the layout).  Plus, there is no single universal location where the detector will work 100% of the time.  For instance, there is no single location vertically where the detector will work 100%.  You have to be able to detect depressed flats, regular flats, tank cars, and all sorts of boxcars and hoppers.   Then there is the gap between cars. Maybe using 3 or 4 detectors at a single location would give you 100% detection reliability, but that complicates the installation.

I'm not saying that optical is bad -- just that it is not perfect either.

[ChrisKLAS]

Don't forget that if you're using JMRI to drive the signals that you can (easily) specify debounce values for each block which will force a signalling system using resistored wheels for detection to work with a train that has only a few "detected" cars to work flawlessly. I've set all of my islands (the portion of track representing control points, including the switch, between the signals) to have a debounce of 25 seconds, meaning that they will only show that they're cleared up and allow the DS to throw the switch after they've been unoccupied for 25 seconds. My other blocks have a debounce of 3 seconds just to keep them from flickering on and off when wheels start to get dirty, etc. This allows me to run 20-30ft trains with only 4 or 5 or 6 "detected" cars as long as everything keeps moving over the control points at a reasonable speed.

All that said, making resistored wheelsets is incredibly easy once you've done it a few times. I normally make them in batches of 40-50 and, despite the safeguards I've put in place on the signal system, most of my trains are running with most cars detected.

[sp org div]

Every axle is the gold standard, but we've found using one 10k wheelset per car on the layout has had very good reliability for years.  Yes, I have seen an occasional miss on detection from single cars on pre-ops testing, but never had a DS complain that his OS sections were dropping out.  I would not recommend less than one wheelset per car (typically the shortest OS sections have 3-5 cars reporting anyhow).  Recommend cabeese to have one detected wheelset per truck.  In certain conditions, trains do sit over OS sections for extended periods... and the goal is to make Ops for the DS as realistic as what the crews see out on the field.  As Chris noted, a few seconds debounce is good for pushing reliability.

Jeff
http://espeeoregondivision.blogspot.com/

[rrjim1]

Yes, nothing is perfect. Just like most everything else, if installed correctly one removes a lot of potential problems.  The Infrared Model Train Detection system I used didn't have any problems detecting any cars on my Model Railroad. My lights are all mounted from the ceiling. No problems with gaps unless they are wider than the two detectors used at each end of every block. So I would say my system works 99% of the time as long as the power is on and a light doesn't burn out. Is this system for everyone, NO, but I'm sure glad I did some research before deciding which system to go with. 

[John]

Very good information here ..  http://rr-circuits.com/Clinics/PDX-2015-Signals.pdf

[nscaleSPF2]

Very good information here ..  http://rr-circuits.com/Clinics/PDX-2015-Signals.pdf

That makes my head hurt, John.  Here is another, hopefully simpler solution.  I am all for simpler.  This one does not need to be tied to the actual rail blocks, nor does it require a computer interface.

http://www.greensteamproducts.com/Pages/GSP-14.html

Anyone here had experience with Green Steam?

By the way, should this thread be moved to the Electronics section?