Thoughts on grade crossing detectors

[SP-Wolf]

So, I'm looking into a grade crossing detector. Any thoughts or advice on what company to go with-? I am running DCC. This will be for a double track crossing. I've looked into Circuitron, Azatrac, Signalogic and Logicrail. So, I'm just looking for a little direction on this. I have the N J Int'l crossing gates with LEDs. So, I will need circuitry for raising the gates and a flasher. Perhaps even a bell.

Any assistance is greatly appreciated,
Wolf

[carmelmodelrr]

I run DC, but I've used Circuitron's Tortoise machine along with their DF-1 and ER-2.  However, I also later looked into Boulder Creek Engineering [www.bouldercreekengineering.com]Nightscope Infrared Detector and ended up with their detectors primarily because their detectors will work in low-light or no-light situations.  They work in tandem with the Circuitron Tortoise Machine and DF-1 and ER-2.  Boulder Creek and Circuitron were both very helpful in coordinating their systems so that my NJ Intl. crossing gates and LED's would work.

Carmelmodelrr
Richard Wroblewski

[randgust]

Remember, this is from the guy that is stubbornly DC and makes no apologies about wheels OR C80 on the main.....

I looked at what I was going to have to do for a double-track main line with a passing siding (3-track crossing), bidirectional direction of travel on both mains, and a mainline crossover in the middle - to do what you're talking about - trigger the LED flasher on two gate sets, two Tortoise switch machines, and the sound unit - which in my case does both horns and bells off of a momentary pushbutton on a Radio Shack digital recorder unit. The toggle basically triggers the Tortoises, and those have rocker switches on them to hit the lights and horns.   The sound system is a latch, so there's also a second button to hit the sound in advance of dropping the gates on a faster train.

AND, I do switching in this area, so I'm back and forth across it within what would be the island circuits.

I enjoy a challenge, but this got nuts and was starting to look expensive and complicated.   I use a single DPDT toggle switch as in 'up/down' and everything triggers off of that.   When I have visiting kids, I put one on that toggle, gives them something to do.  That's actually been a big hit; blow the horn and run the gates.

The fun part is that when they redid the real crossings at Flag, they ended up pulling out the mainline crossover and siding switch just to help simplify the mess.  So it's not just me.

[mplsjct]

I have to second the toggle switch method, the switch for young visitors is entirely true! (and not so young visitors too).

I also am DC, and do a lot of switching, I will consider automatic detection in the future, but for now the toggle works fine.

[SP-Wolf]

Good morning Guys,

Thanks for the food for thought. The more I think about it - the more I like the idea of a simple toggle to operate the crossing gates/signals. So, what are you guys using for a signal flashing circuit and for the bell - ?

Thanks again,
Wolf

[Philip H]

I have one that's a Dallee Electronics flasher circuit I bought in bed ford a couple of years ago.

[mmagliaro]

Why can't there just be multiple detectors on each track on each side of the crossing,
so that if any one of them is occluded, the gates and lights operate?

It requires a lot of detectors to really make it work right, and that could run into money.
But I'm confused about why people see this as complicated.  Is it the expense of all
the detectors?   Or am I missing something else here?

In answer to your question, I use a Berkshire Junction detector/flasher board. 
They make a 4-detector infrared board for $33.   http://www.berkshirejunction.com/

[randgust]

In my case, what made it really interesting was crossing from one track to another inside the detection area, PLUS bidirectional moves on both tracks PLUS stopping and reversing inside the detection circuit PLUS a passing siding.  Like I said, I could figure it out, and I might even do it yet, but when I can do the whole bit with a single toggle switch...    With motor-powered gates it's a bit more challenging than just flashers.    The real problem with a typical model design is you really need to advance the gates down early on approach and snap them up rapidly as soon as the train clears the crossing; a real circuit is a combination of approach circuits and the middle "island" detection circuit.  It's been greatly simplified on the prototype with the adaptation of "predictor" logic ircuits on approach rather than pure fixed relays.

I've used old Radio Shack digital recorders (which give you up to 30 seconds) as self-contained sound units.  You can record sound right off of a videotape.   No moving parts.   The nice part is that they momentary-contact on, so they'll play the entire sound byte AND STOP with one momentary hit of a button.    I got recorded sound from the exact location  I'm modeling, including the echos, so the effect is really nice.
I move out the speaker, actuating pushbutton, and put them on a 9v wall-wart power system, but yeah, for $12 is hard to go wrong:  http://www.radioshack.com/9v-recording-module/2761323.html

I've put audio over top of this, but here's the current installation with the new gates: 

/>On this one you can hear the bells and horn: 

/>
The only 'oh CRAP' I ran into on my latest upgrade was changing from NJ bulbs to NJ LED's on scale-sized gates.  The ancient Walthers DC flasher circuit couldn't be used because it was a common wire to the GOW bulbs that was opposite the LED's - the common-anode vs. common-cathode LED problem, so you really have to make sure your LED's on the flasher agree with the wiring plan for the flasher circuit.   I think I ended up using a Dalee for the NJI gates.  Not all circuits will work with all LED gates, beware.

To me, the right system would have about four detectors per track to be bidirectional ; a 'distant' detector to drop the gates (effectively a predictor), which would latch two center island detectors on both sides of the crossing - both of which would have to [AND] clear detection to lift the gates back up.   Remember on real gate behavior, if you creep up to a crossing while switching, the predictor circuit may lift the gates, but once you hit the island circuit, down them come until the island clears completely.

[SP-Wolf]

Why can't there just be multiple detectors on each track on each side of the crossing,
so that if any one of them is occluded, the gates and lights operate?

It requires a lot of detectors to really make it work right, and that could run into money.
But I'm confused about why people see this as complicated.  Is it the expense of all
the detectors?   Or am I missing something else here?

In answer to your question, I use a Berkshire Junction detector/flasher board. 
They make a 4-detector infrared board for $33.   http://www.berkshirejunction.com/

Hi Max,

Thanks, for your input. Berkshire Junction looks good.  One question about their unit (since you have one) --I have a double track crossing -- would I need two of their detectors - ?

Thanks,
Wolf

[Philip H]

You could alwasy roll your own:

Detector - http://model-railroad-hobbyist.com/node/23535

Flasher - http://model-railroad-hobbyist.com/node/23504

The parts for the flasher (excluding the bread board) are about $3.75 through Digi Key

[SP-Wolf]

Thanks @randgust and @Philip H -- you guys have really given me some awesome options.

Wolf

[mmagliaro]

Hi Max,

Thanks, for your input. Berkshire Junction looks good.  One question about their unit (since you have one) --I have a double track crossing -- would I need two of their detectors - ?

Thanks,
Wolf

First off, remember, I have the earlier product that used photocells, and the newer infrared detection is much better
for low-light operation.    Having said that,
with their new system, no, you should only need one detector because you set up an IR emitter on one
side shooting across both tracks, and the receiver on the far side.  So if a train breaks through the beam
on either track, the circuit will trigger.  They say it's good for up to 8" of "shoot" across the tracks, so it should
be fine for 2 tracks.

You may end up wanting 2 boards anyway.  The 1 board can have 4 detectors.  So you can put 2 on each
side of your crossing.    But if you really want it to always work perfectly for any length of train, even a
switcher with no cars, you might want more than 2 on each side.  Imagine having one detector
about 2" from the crossing, and another about 18" back from the crossing.  A short train could
trip the the first one as it approaches, then roll past that detector, causing the gates/lights to clear,
and then roll over the next one right as it gets to the crossing.  So the lights would come on, go out, and come on
again as the train approaches.  It would work fine for any train that spans across that 16" space (18 - 2), but
not for anything shorter.   In my case, I really don't care.   But if it bothers you, you would need
two detector boards so you could put 4 detectors on each side of the tracks.
You would hook both output relays up in parallel, so if either relay closes, your flashes and gates come on.