Protecting uncoupling track from the AIU

RJR, you should be able to get your 24 VDC very easily with a bridge rectifier and capacitor from 18 VAC.   it may end up more like 28VDC so you may need to step it down from there.  I think there are cheap modules on eBay, and the like, that will do this for you.  

The current detector model, a 5-volt type, came in today.  Unfortunately, it comes with no documentation and I have no idea what each  the 3 adjustable pots control.  The underside of the unit is marked "Logo-I-sensor V1.1   www.auto-ctrl.com"  If I go to that site, it's all written in Chinese.  The 5 pins are labeled Voc, Gnd, Gnd, Aout and Vout.  I assume Voc is 5VDC+ and Gnd is 5VDC-.  The relay output terminals are clear.

 

 

How about a link to the exact model or version you ordered?  That website is cluttered and in some strange language with funny characters

I'd forget about that, just get the tig or mig welder out and weld the track together, more secure than uncoupling track.

A printout is attached.

Yep, everyone has a TIG welder to do that job...

 

http://www.aliexpress.com/stor...7356_1183478243.html

 

This may get you a little closer as I can at least read the characters!

 

 

Stan, thanks very much.  I appreciate your effort.  My searching on Aliexpress did not disclose that

A question for the knowledgeable electronics experts please.  

 

I rigged up a test bed.  One lead of the feed to the uncoupling track goes through the hole in the sensor.  5VDC+ is fed through VCC; 5VDC- is fed through the adjacent GND.  Pilot light is on. 

 

Nothing happens when I send 15-volt AC power to the uncoupling track; the magnet does get activated, so I know AC is flowing. 

 

Which of the three adjustment pots, in your opinion, would set the amperage at which the relay will pull in? I've turned the level signal adjusting knob and the sampling signal adjusting knob up & down, with no effect.

You might try looping more than one turn of wire through the donut to increase the signal.

Gentlemen,

   Right now I have zero uncoupling track on my layout because of this problem, I was hoping to see an inexpensive fix here in the OGR, so far I do not see a good cost effective solution.

PCRR/Dave

Define cost effective.  There are many ways to do this, as discussed above.  The method I would chose has a cost of about (quick guess) $35 for up to 16 uncoupling tracks. 

Dave, as of now, I have about $15 invested, and if I ever figure out the adjustments, this will cover all uncouplers.  In the 65 or so years I have been running trains, using standard push buttons, I have lost one uncoupling track.  A small price to pay to be able to break down and make up trains.  

 

The reason I am working on this system is so I can also use AIUs, in case I accidentally hit wrong AIU button.

SUCCESS!

First of all, thank you Dale.  Your idea made this work.  Instead of merely running the uncoupling track feed through the sensorhole I looked it around through the hole three times.

Attached is a video showing this system in operation on a test board.  On the left is a Z500 transformer, supplying 14 V.  It’s leads are the black and red alligator clips.  

Moving to the right, is a pushbutton to operate the uncoupling trac.  Next, screwed to the wood, are the current detection board, a power supply feeding 5VDC to the current detection board, A power supply feeding 12 VDC to the time delay relay, the time delay relay, and two leads heading off to the sounding device.

Attached to the red leads of the transformer are: (1) a wire that loops through the current detection board sensor three times and goes to a push button to supply power to the uncoupling track; (2) a wire to feed power to the power supply for the current detection board; a wire to feed a contact of the relay on the current selection board.  Attached to the black lead are grounds for the uncoupling track, the coupling detection board, the two power supplies, and the time delay relay.

When the current detector board senses a current, an LED on it lights and it closes a relay which feeds 14VAC to the power supply for the time delay relay.  When the time delay relay starts receiving power, an LED lights and about 10 seconds later its relay closes and sounds the alarm, and another led lights.

You’ll see this elapse when the button is pressed.

 

Note that this is a test setup.  For an actual unit, I's place them in a compact cabinet and the wire that here runs to the uncoupling track, would in fact be the feed to the uncoupling track control panel buttons and to the AIU ACC terminal feeds.

 

The adjustable time delay relay, adjustable power supplies, and the current detector sensor boards were all obtained from China through Aliexpress.

I think I'd set the delay to a two-three seconds, you don't normally need 10 seconds to decouple a car.

User's choice---user adjustable by turning the pot on the time delay relay.  I'm seeking protection from burnout, not saving power.  When moving slowly on sidings, I usually have to press the button for longer than 3 seconds.  Also, if using an AIU, there is reaction time.

 

Note that there are identically appearing 5-volt and 24-volt current detection boards available.  I chose the 5-volt so I could use a power supply from my stock.

 

A Gargraves uncoupler draws about 2.1 amps, an old UCS about 1.2.  As received, the current detector would not "see" the UCS current.  The pot nearest the relay has to be turned clockwise a considerable number of turns for it to recognize the UCS's draw.  If turned too far, the detector tends to stick.