Skip to main content

There is a posting of someone wanting to use IR detection and types.

Z Stuff, Lionel, MTH etc do not tell how their detectors work ie pulsed LEDs, steady beam.  They all seem to have sensitity controls which means it is adjusting sensor analog output but sensor still suffers from big ambient light changes.

 

If i were designing a sensor I would do pulse emitter.  My way would be turn on LED and sensor for an interval then off.  Read sensor analog value, convert to digital value and store.  Repeat cycle and over period of time obtain a baseline value average which would include the ambient lighting, or a traveling average.   If the value suddenly dropped below my average after so many blinks, then I assume path is blocked and trip output.  Now my baseline average falls to some value around ambient lighting value.  When value suddenly jumps after so many blinks then I assume path has cleared and open output.

 

Are any vendors doing just this?  I am thinking of using some IRs at several places but reading posts seems none really are reliable under all conditions.

 

Would be interested in discussing this topic technically. 

Original Post

Replies sorted oldest to newest

Perhaps there are some using your method, but the more common method to reject ambient light is to pulse modulate an IR LED and use an IR detector tuned to the modulation frequency.  The following was the first Google hit searching for "IR model train detector".  

 

http://www.azatrax.com/ir-model-train-detector.html

 

If you like to tinker, the IR detectors are less than $1 each in hobby quantities and are factory-set for frequencies around 40 kHz.  So all you then need is an IR LED that pulses at that frequency which would run less than $1 in parts. Sensitivity would be set by adjusting LED current.

"If i were designing a sensor I would do pulse emitter.  My way would be turn on LED and sensor for an interval then off. "

 

You're on the right track rrman. There's a method called correlation. In it's simplest form you would modulate the transmitted beam at say 1 KHz. The receiver is designed to accept only inputs modulated for that frequency and would then reject all other input levels.

Seems ZStuff stand alone sensor with led and detector built in claims to be imune to ambient lighting.  Also AZtrax (sp?) also claims immunity but using separate led and senors that user positions.  Googled IR train sensors and got alot of hits, and these two seem to be robust.  I don't mind spending money just want to "plug and play"

 

Anyone using these and how well do they work? 

I have had pretty good luck with ZStuff trackside sensors, after getting the spacing and initial setup right. they have to be very close to the outside rail to work I have found.

The best so far are the Lionel 153IR. They are quite sensitive and have a time delay pot, which is really handy. They are much more tolerant regards spacing to the track. They are not cheap though.

I have not had any issues with different ambient light levels.

 

Rod

Been e-mailing John Parsons at Aztrax about IR sensor immunity.  His respone is as follows:

Our IR train detectors do a very good job of
rejecting ambient light levels, as well as incidental IR light from
other nearby detectors.  You can place two of our sensor pairs right
next to each other and they will not interfere.

The only vulnerability is to ambient IR light that is intense enough
to saturate the receiver.  This is more likely to happen with
across-the-track installations and is remedied by creating shade with
scenery or structures, or by swapping source/receiver positions so the
receiver points away from the offending source.

With sensors embedded in the roadbed the receiver can be pushed
farther down into the roadbed to shield it from ambient light.  A
light source directly over the track is not a problem because a train
provides shade.  If a train is not on the detector, then in reflective
mode a no-signal condition represents a no-train condition anyway.

***************

 

So this may be the way to go to replace my hall effect sensing in tracks.  Plus the Aztrax carries the NJI "A" arm style black strip crossing gates that was protypical in my layouts 1940-50 time era.  And the $60 Aztrax module supports the crossing gates and light flashers.

 

Now to save my pennies.

Hi,

 

Just to offer some clarification, all the Z-Stuff sensors use pulsed light to minimize the ambient light issue. Our DZ-1070, DZ-1075 series work in bright sunlight to total darkness with no problem. This is the same sensing system used in all our trackside signals with the exception of the DZ-1011. It uses 60 cycle pulsing for the IRLED which is OK, but the sensor is not as well shielded as the others.

Wanted to update this thread with latest goings on.

 

I bought the Aztrax MRX2 flasher board with six sensors.  John Parson instructions says to calculate for worst car height so sensor beams hit it. I calculated that I needed 65 degree inclination to hit the Lionel tank car bottom.  Fashioned a drill guide from scrap wood block and two 3/16 metal cable clamps (picture on request) and used a 3/16 long aircraft drill bit to place holes in roadbed, one hole on each outside rail so blacken center rail blocked any IR energy. Leds on board shows when sensor is active incase there was light leakage.  After placing the five sensors I pushed the tank car over sensors and board worked as advertised. 

 
Acid test was an Addladin caboose lamp (incandescent bulb) on wall that was over a sensor pair. No problem when train rolled by and exited with caboose lamp on or off.
 
I am using the flashers from my original setup.   Board gives lamps a fade in/out effect though not quite aqs pronounced as I had on my design.  Later I will install the NJI crossing gates and tortise motor setup.
 
Overall very pleased with this setup and would recommend Aztrax to others.
Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Ste 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
Link copied to your clipboard.
×
×