Removing slide shoes from operating cars (formerly: "Reed switch"

RJR posted:

...a reed switch that will work with a USC track?

By "work with" do you mean you have a physical magnet, and you position it near the reed switch to switch the several Amps of current to the UCS electromagnet?  Are you also saying you want to use 12V DC as your UCS supply?

Working to eliminate pickup shoes on operating cars.  Put reed switch and relay into car to feed track power to solenoid.  Have tried dc into ucs track in case reed switches used magnets in reeds

Gregg, you got me pegged

So you can put DC into the UCS electromagnet.  That's handy as reed switches are very fast and will buzz.

You are using a relay in the car to feed track voltage to the solenoid?  That is, the reed switch itself does not have to switch/carry Amps of solenoid current.  So you can use a low-cost 12V DC relay with a simple circuit as has been discussed in this forum for signal relays triggered by AC insulated rail sections.

In which case, I must be missing something since I'd think any 50 cent reed switch ought to do the trick.  UCS electromagnets indeed put out a huge magnetic field which will easily trip a reed switch from 1" (or whatever) away. 

Yes, I am using a 12VDC relay, 400 ohm coil, fed through a full-wave bridge rectifier with 450mfd cap on output.  Don't bet on the 1" "range" of the UCS electromagnet.  Your use of the term "ought" is recognition of the fact things may not quite work out.  The glass reed switches from RS actually seem to be directional, and well as fragile.  The windows switches have square cross-section, eliminating a directional issue, and come reasonably close to the UCS magnet, but as stated above, they freeze.

FYI, the trucks have been replaced so there is a center rail pickup and no shoes.  Relay points supply power from the roller to the solenoid.  Until the window switch stuck, system was working as planned.  Also worked great when I brought a magnetic pickup tool near it.

Therefore, I started this thread to find out if perhaps someone else has used reed switches with a UCS.

Stan, I only used 12VDC into the UCS electromagnet, from a DC power supply I have on the workbench.  It did seem to get warm more rapidly than with AC, and I suspect that, due to lack of self-inductance, it is dissipating more wattage.

Rjr,

Trying to understand what your trying accomplish here.

 Is it that your Relay Coil will only works with 12 volt  DC current ? 

Also what is it your trying get working here ?

Good luck, John 

Kb2agpjohn@aol.com or on my cell phone 848-992-2157

John, operating car pickup shoes have a nasty habit of fouling Gargraves and other switches.  I've even had them operate the cars where they shouldn't.   I trying to eliminate them.  On the test dump car, I've replaced the shoe-type trucks with trucks with a center rail pickup roller. I'm testing using a reed switch to operate a relay, to close a circuit to feed track power from the rollers to the solenoid.  Power for the relay comes from the rollers, goes through the reed/magnetic switch, through a full-wave rectifier with 450mfd cap, to the coil of a 12-volt relay with 16-amp point rating.  Works as planned, except for the problems noted above with regard to the two types of reed switches I've tried.

AC vs. DC presents no issues

I'm just trying to find if anyone has used a particular switch so I can save the money in buying many different types.  When perfected, I have about 10 cars that I intend to convert.

In another posting, a wireless remote is suggested for controlling accessories remotely. I have been using the single- and multiple channel versions of these for a few years. 

The single-channel receiver will fit in most operating cars.  You need to rectify track voltage to DC to operate the receivers. Add a relay if the contacts are not big enough for your load. Measure with an ammeter to determine if you need an additional relay.

The pushbuttons can be used instead of the UCS, and another advantage is that you can operate the car at any point along the ROW.

https://ogrforum.ogaugerr.com/t...ccessories#lastReply

"What'd making RJR?"  

Please explain this question.

For some reason I can't view the picture I thought I posted in this thread but I think I found it on my PC and posted it below.

https://ogrforum.ogaugerr.com/t...operating-cars-onoff

I thought I posted a video too.  In any event, you can read the thread and see I was doing something similar.  In this case it was a 2 reed-switch implementation where one turned on a latching relay, the other turned off the relay.   That is, I wanted to turn on the accessory for long periods rather than your momentary pulsing application.  In my case it didn't matter if the reed switch momentarily buzzed (from AC) since redundent pulses to the latching (or unlatching) relay coil are benign.

The point is I don't recall any problems having enough magnetic power to trip the reed switches.  The specific reed switch I used was the RI-25AA which I got at DigiKey but I see it is no longer sold.  There are many equivalent parts.

Are you placing the reed switch INSIDE a car with a steel chassis?  Ferrous objects near or blocking air gap between the UCS electromagnet and the reed switch will effectively spread out or weaken the "cylindrical" force field going straight up from the UCS electromagnet.

Also, if the glass casing and mounting options of low-cost reed switches presents fragility/vibration issues, you might consider semiconductor hall-effect switches (no moving parts).  Same price - less than $1 and smaller than a reed-switch.  Needs a DC power supply but you have one on board already (to drive the DC relay coil).

Arthur, thanks for the post.  Having 10 cars to convert, I'm not going to rf means. 

Gregg's post probably had a typo:  I'm sure he was wondering what I was up to now, for he knows that over the years I've come up with a few specials circuits, like one to cut off power to a UCS magnet after a few seconds, to prevent burnouts when used, and others.

Stan, reed switch was under the car, the old Lionel dumping coal cars from the 1950's.  Given their consturction, that's where any reed switch would have to go.  I have to tackle the Digikey catalog, and try a few.  Other than having had a Hall effect distributor on a Ford 20+ years ago, I'm not familiar with the needed circuitry.

I'm attaching 2 pictures of the test car before wiring finished.

Are the "failing" reed switches sticking magnetically or are their contacts welding electrically? Was wondering if the DC field from the UCS is magnetizing nearby ferrous car parts and keeping the reed relay from dropping out. Seems there used to be a trick with a "bias magnet" when this was encountered, but I am not an expert by any means.

Do we know for a fact that the charging current of the 450 mfd capacitor is not welding the contacts? Depending on the relay coil, maybe put 10 ohms or so to limit the charging current.

Magnetic power to trip the reed switches.  The specific reed switch used uncouple one of your train cars ?

Yes ? 

John

John:  No.  Changing the trucks brought me into the second half of the 20th century, by eliminating the 1940s electromagnetic (coil) couplers.  Cars now have 1950' magnetic couplers.

PLCProf: No way to tell; the windows sensors are sealed.  The question of magnetizing cars is a good one, but removing the failed switches from the cars does not disconnect the contacts.  Question about the charging current is also good; worth a try adding the resistor.  Related question would be how large a capacitor is really needed to keep the 400-phm coil relay from vibrating in the 60-hz-pulsating DC????

Arthur P. Bloom posted:

"What'd making RJR?"  

Please explain this question.

It's kind of slang around  here  for.... what are you making or doing, Anyway I deleted it.

John:  No.  Changing the trucks brought me into the second half of the 20th century, by eliminating the 1940s electromagnetic (coil) couplers.  Cars now have 1950' magnetic couplers.  Its sound like electronic is not working out I thought I could help out But afraid

Sorry, I don't have answer here. I thought maybe you could work out something using a track relay's instead and perform the some thing your looking for here.

Good luck. Please keep us posted on your progress.John     

This is your lucky day.  Take a look at the sub-discussion around this recent posting:

https://ogrforum.ogaugerr.com/t...22#55630645898116322

PLCPROF shows a technique which might solve your problem and eliminate the capacitor (and the attendant huge inrush current which may be zapping your reed switch).  You only need 2 diodes, but since you already have the 4-diode bridge installed, I think you can go from where you are to the following diagram with minimal changes.

Thanks for the link, Stan.  Something to chew on.  I used diodes years ago to keep inductive loads from causing excessive arcing at relay points. 

There is an error in your sketch.  The terminal of the relay you show connected to the wheel is actually connected to the negative output of the rectifier.  What would be needed is another diode across the relay terminals in lieu of the capacitor.  Using full-wave rather than half-wave rectification lessens the tendency to chatter.

I think the sketch is correct.  The method described uses half-wave rectification and NO capacitor is required.  Chatter is eliminated by addition of the 2nd diode D2 which PROFPLC calls the "flyback" diode in his postings.

In my sketch I was showing that since you already have the bridge installed, you can implement this no-capacitor anti-chatter technique using 2 of the 4 diodes in your bridge by just moving a couple wires around in your existing circuit (and removing the capacitor). 

However if you don't want to deal with the technical intricacies of this method and you're more comfortable with full-wave rectification with a capacitor, then so be it.  In which case 450uF is more than enough - I'd think 47uF would work.  And I'd think a current limiting resistor of, say, 100 ohms would be suitable since your coil resistance is fairly high at 400 ohms. 

I just located a "reed-based magnetic position sensor" with 0.5 amp contacts, and ordered 2 for testing.  That should obviate the possibility of contacts welding.  Meanwhile, I run some experiments to see how small a capacitor will keep the relay from chattering.

It's not a question of dealing with intricacies, for it's not very deep technology.  Just thinking that a full-wave rectifier with the diode in place of a capacitor might be even smoother.

I misunderstood; you're on board with the technique albeit a 5-diode approach.  Anyway, since the application is anti-chatter or protecting against dropouts measured in the few milliseconds, I think a 2 cent diode is preferable to a 5 cent capacitor since, (a) the capacitor still presents a low impedance or large inrush currents so you need to add another component resistor, and (b) over time electrolytics seem to fail more than diodes.

Check the power rating of your reed switch.  Even though it says 0.5 Amps, you generally measure the switching power capability with the open circuit voltage.  Small reed switches will be down there in the, say, 10 Watt region.  So if you don't put in a resistor, you can easily be switching more than 1 Amp inrush to the capacitor...and with a track voltage of 18V you can be switching more than 10 Watts even though you would be carrying much less.

The window reed switches don't show a rating.  The RS said 0.5 amp, but didn't mention whether switching or carrying; in any event, they snapped easily.

I understand your point about the switching.  Good point.  If I can't find any diodes in my junkbox, I'll have to pay a visit to RS.

The switches I ordered, Cherry part # AS201901, are rated to a 0.5 amp switching current and 100volts ACorDC switching voltage.

I am optimistic that this approach to operating cars can be made to work. 

Per datasheet, the Cherry part is rated for 10 Watts switching power.  So if using the capacitor, I'd put in the resistor.  Too bad the Cherry datasheet only specs the sensitivity (trip distance) to a specific Cherry magnetic actuator.  More surfing required to get specs to compare with other reed switches should you have problems triggering the reed switch from the UCS electromagnet.

Pre edit:  simple solution thought about while rambling on about other things below:  

Use a much, much smaller capacitor. and only a single diode, not a bridge, if you are planning to run at command level track voltage.  The full wave will produce over 25VDC from 18VAC track voltage, and bring your coil current up over 60mA.  using a single diode, for half wave, you'll put just under 13VDC on the relay coil and only draw a little over 30mA.  On the Cap, 470uF is perfect if you are building a 1amp power supply, but something around 33-47uF ought to be plenty to hold a 30mA relay coil closed. 

Back to the regularly scheduled meandering tangent post: 

Just something silly here... Have you tried just setting the relay in place on each of the 6 sides  over the electromagnet?  the UCS magnet may just be powerful enough to pull the relay's contacts without power on the relay's coil... If it is not quite enough, you could power the relay's coil constantly with not quite enough current to trip and try the magnet again.

There are cheap hall effect sensors out there Search the A3144 on the auction site ($1.59 for 10) these can only switch 25mA, but will do it well... and are much more robust than a glass reed.  They also run on 5-28VDC, so you don't need any 'fancy' regulator.  Add a 2n3904 transistor and you can easily drive a couple relays or a 2n2222 and drive a dozen relays.   

It looks like all you have for parts now is 1. reed switch, 2. capacitor, 3. Bridge rectifier, 4. Relay.   So let me see how much I can complex it up...  I'm going to assume you're good with running the relay coil with the 70% or so of track voltage it will see through half wave: about 12.7VDC from 18VAC track power.  As a note feeding a 12VDC relay with full wave DC from 18VAC track is going to dump over 25VDC on the coil. 400 ohm, 25V:  62mA.  At 12.7 v it takes about 32mA.  If you need the circuit to trip the relay with voltage less than about 14VAC on the track, the full bridge is back in play in place of "diode A" below.  (this will vary based on what voltage is required to throw the relay.) I used the same 470uF cap you already have, but you can probably use something as small as 47uF.   Diode B is the so called 'flyback' diode, and in this case it is needed to protect the transistor from surges. (it provides the same function as a TVS on your track) the 2.2k Resistor limits the current draw of the transistor, preventing it from damaging the Hall effect sensor.  All in all the only parts needed are a couple pennies worth of diodes, resistors and a capacitor, a 16 cent sensor, and a 2 cent transistor. less than a quarter per unit.  

If you want to use a nice, low current reed switch, put it in place of the Hall Effect sensor, between ground and the 2.2k resistor(R1).  

 JGL 

Edit of edit, turns out i was correct the original way around.  

Wow!.  Thanks John and Stan for the effort you've put in.  I think first I'll try going to half-wave with a diode vice the capacitor, and then go the Hall effect route.

Can the Hall effect sensor be used with the UCS's magnet fed AC, or should it be full-wave DC with a capacitor?

The belly of the Lionel dump cars is metal, so I doubt that the UCS would activate a relay armature through it.

Allegro MicroSystems reports that the A3144 is no longer available and has been replaced by the A1101-1106 family.  I do see A3144 listed on Amazon from China.  Jameco has the replacement A1104

http://www.jameco.com/webapp/w...001_10001_1718701_-1

Would these be suitable in your circuit?

RJR, 

To be perfectly honest, I chose the sensor I did by typing 'Hall Effect Sensor" into the search box on the auction site and clicking the cheapest thing that popped up, then checking the data sheet so see if it would work.  The A1101-1106 appear to have the same specs as far as operating voltage and current capabilities, so should work fine.   I tend to work on a budget, so the 16 cent sensor is my choice over the buck fifty one, but either one will work equally well.  

As far as AC or DC on the UCS, I'm really not sure.  If AC causes the circuit to chatter you could place a capacitor, say 47uf or so, across the relay coil.  The transistor will have plenty of current to charge it, but if you're worried at all, and the cost difference doesn't matter to you, use a pN2222 transistor instead of the 3904, they handle 5 times the current.  

Because complexing things up is what I do,

As another option, if the car you are running will work on half wave power, is to use something like a TIP120 power transistor and run the load right through it without need of a relay.  The TIP120 will run a load up to 5 amps, though it will need a big heat sink to do so... I'd use the metal body of the car as the heat sink.  

or, if it needs full wave AC, a Triac could be used, but I'd have to do the homework on a suitable part, Lionel's favorite, the BTA08, will probably do the trick.  

JGL

The NPN transistor (2N3904) needs to be of the PNP ilk.  When the hall-sensor turns on, it pulls down its OUTPUT pin so you need a PNP to drive the relay coil as drawn.

If you're driving the UCS electromagnet with AC, the Hall sensor will turn on/off 60 times/sec.  So employ the flyback diode method, or use a capacitor across the relay coil as JGL suggests.

And from left-field...I realize you have the UCS in place.  And I realize you nixed the RF wireless approach.  But what if in place of the UCS electromagnet you simply activated a 10 cent LED aiming upward from the track.  Then on each of your ten cars, you simply put a 20 cent phototransistor switch aiming downward at the track.  Not that you are limited by power on the track side but you'd only need to drive the LED with, say, 20 mA of current rather than 2 Amps (or whatever) to whack the UCS electromagnet.  Phototransistors are not fragile like glass reed-switches.  LEDs are small so would be easy to embed into the track bed. And you could use Infrared LEDs and phototransistors if for some reason you don't want a flash of visible light when you activate the car.  Phototransistors can't drive a relay coil directly, so you'd have a circuit like above with the phototransistor in place of the Hall sensor IC chip.  This limits the car operation to just YOUR layout since no one else would have these 10 cent LEDs on their track.   Just a thought...

Stan, thank you.  That was the edited edit of an edit in my post, and turns out I got it wrong some more.  The pnp transistor will work as you show.  May also need a pull-up resistor between the base and vcc, but I'm unsure of the value on that. A pair of NPN transistors could also yield the desired results.  

Something fun I found today are the Latching type Hall Effect sensors.  These will turn on when passed by a south pole of a magnet, and stay on until they pass through a north pole of a magnet.  Might be fun to simply place some magnets under the track in various locations to make a car turn on and off as it passes those magnets.  

JGL

Thanks for the info on the transistor, Stan, and the verification, John.  The LED is an interesting alternative, and is in my price range.  Would ambient light have an effect?  I hate to have coal dumped anywhere but into the coal loader bin. 

I can see one big advantage to the LED route:  The led can be off center to avoid the center rail, and nearer one end of the car, so that if the car is in the wrong direction, it will not dump.  Also gets away from AC vibration.

Phototransistors is an area into which I have never gotten.  Looking at the Jameco catalog, p 17, I see that the term encompasses optoisolators and optocouplers.  I also see there slotted optical switches and optical tilt switches.  And they all have more than 3 legs.  Now I'm in over my head---way over.

Take a look at Jameco p.23 in the middle table.  The Fig. 2 and Fig. 3 show typical package types with the emitters (aka LED) and detectors (aka phototransistors) being available in either package type. 

Can you take a photo(s) of the bottom of your coal car or whatever accessory cars you're trying to modify?  I'd like to see the kinds of nooks and crannies that might be suitable for mounting the detector whether it be a reed switch, hall sensor, phototransistor, whatever.  With some common sense and planning you should be able to mount a phototransistor in a way that rejects any ambient light, but to your point perhaps it's just best to use infrared components.

I figure you're probably an old hand with LEDs so nothing to say there.  Think of a 2-legged phototransistor as a switch that turns on when light strikes it, and turns off when dark.  The reed switch and Hall sensor are the same except its magnetic field instead of light.  Since it's a semiconductor device there are some voltage/current constraints/considerations but as mentioned the circuit would be similar to what JGL has shown.  And of course the activation LED could be turned with DC current so it would be solid ON so you don't have the 60 Hz AC issues.

After all, it is YOUR railroad so what is kind of amusing about the LED approach is your can make another activation track for 10 cents to dump coal at yet another location...as opposed to buying another UCS or electromagnet which I'd think costs more than 10 cents?!

JohnGaltLine posted:

Something fun I found today are the Latching type Hall Effect sensors.  These will turn on when passed by a south pole of a magnet, and stay on until they pass through a north pole of a magnet.  Might be fun to simply place some magnets under the track in various locations to make a car turn on and off as it passes those magnets. 

Right!  My railroad empire is small, but I've always thought that it would be cool to have a electronic latch of some type that would turn on the lights in a passenger car set as it leaves the yard onto the mainline...then turns off when it returns to the yard.  This could be done with a latching hall sensor... or even a latching reed switch though they tend to be finicky creatures.

That's always struck me as a big deficiency in passenger operations.  You have these fancy command control engines that can startup and shutdown..but the passenger cars are permanently lit up unless you wire up and manage some kind of block power control.

Requested photo attached.  The white strip is double sided tape that held the last magnetic sensor. This is the coal dump car.  I also have 3 log cars, 3 milk cars, and an MYH coal dumper.  All differ.  Pictures from top are above.

Thanks, John.  I look for lowest cost options also.  Picked up some diodes today to try the first option.  Unfortunately, a client sent me an email and now I'm tied up for a considerable period.

Old Lionel searchlight cars had a plunger in the center, to working an o-off latching switch.  Activated by UCS magnet.

Well, IIRC there was a small "border skirmish" about the design of a certain un-named LED controller module.  I wanted a certain un-named individual to leave an unused SMD resistor spot to put a LM317 into voltage regulator (vs. current regulator) mode.  Anyway, these low-cost eBay 12V wireless receiver modules could then be powered by the LED controller module in voltage mode.  The remote modules seem to work fine at 9V or 10V too so you'd still have intensity control.

I can't remember where our discussion left off about "learning code" modules and how it might work or be applied for addressing in your MP3 sound/accessory module, but I've been messing with this and they are really cool for the price.  For example:

So in addition to "learning" the address without having to flip address switches or solder jumpers you could obviously set each passenger car in the consist to the same address so that all car lighting turns on/off together.  Additionally, the module "learns" the relay mode of either momentary or toggle/latching.  So if you can set it to latching mode, turn relay on with one button press, turn relay off with the next button press.  In the OP's coal-dumper you'd learn the module to operate in momentary mode - fire solenoid only when remote button is pressed.

I was thinking of writing this up, but there's the pesky issue of where to get 12V DC in a piece of rolling stock without having to mess with those relatively large AC-DC converter modules...or having to roll-your-own with DCS inductor, bridge, cap, regulator, etc.  Hence full-circle back to that yet to be "invented" compact voltage output module.

Stan, I have turned out over a dozen small (not miniature) AC-DC converters/voltage regulators for LED lighting in passenger cars and stations, which would easily fit into the coal dump car, but probably not the log cars.

What triggers the Chinese gizwack you show, and can you provide a link?

Maybe it's time for a compact track powered DC converter.  The hot ticket would be a switcher so  you could draw reasonable power from it without large heatsinks.

RJR posted:

Stan, I have turned out over a dozen small (not miniature) AC-DC converters/voltage regulators for LED lighting in passenger cars and stations, which would easily fit into the coal dump car, but probably not the log cars.

What triggers the Chinese gizwack you show, and can you provide a link?

There are hundreds of listings for 1-button, 2-button, 4-button, 8-button fob-style remotes for a few bucks.  For example:

There are some i's to dot and t's to cross such as choosing the frequency (315 MHz or 433 MHz or something else).  I thought you had nix'd the idea of RF so let this drop.  But if you've reconsidered or are just curious for another project I can provide more details if you have a specific application in mind.

So here's a photo showing a 4-button remote with the previously shown "learning" relay receiver.  You teach the receiver to respond to the "A" button using a sequence of button presses.  You teach the receiver to only trigger the relay when the "A" button is held down...or on when pressed once, off when pressed again later.  Or even to turn on when "A" is pressed and off when "B" is pressed.

After seeing your coal car underbelly, I'd think you'd want to use the so-called "sidelook" package of LEDs and phototransistors if going the optical trigger route.  You can see how flat they are and as shown would shoot the LED beam straight up (clear package is the Infrared LED).  As shown the phototransistor (dark package) would detect an Infrared beam coming straight down so if mounted under the coal car would be flipped over stuck to your foam tape.

Sounds like you have been called off onto another project but whichever approach you want to pursue we can get into specifics when you have more time...

Thanks, Stan.  Just curiosity.  Always interested in learning something.

Take a look at Jameco p.23 in the middle table.  The Fig. 2 and Fig. 3 show typical package types with the emitters (aka LED) and detectors (aka phototransistors) being available in either package type. 

Stan, are you looking at the March 2016 catalog?  I don't see any "packages."