Coupling magnets problem?

Is your design based on the discussion from this thread?

In that thread I ball-parked about 80 feet of #28 Magnet Wire - and about 600 turns on a bobbin with a steel core/rod.  Did you happen to count your number of turns? Or, as tediously detailed, you can do-the-math to back out what you have by measuring coil dimensions, inside/outside diameters, electrical resistance, Amp-Turn product, and so on.

As for the melting, note that an uncoupling coil dissipates some 30 Watts of power when active!  There are soldering irons that dissipate less power than that!


1. What is your dowel/rod diameter?  It doesn't look like it quite fills the bobbin hole.  All things being equal, if the rod fills the bobbin hole it will generate more magnetic pull at the coupler tack.  Also, you want the rod to be as high as possible; it looks like it might be a tad below the height of the center-rail?  It may be only a 5% effect but every bit helps.


2. Is your dowel steel?  The way the rod is bent in your video I wonder if it's some "softer" metal?  You need a ferrous rod like steel.  Brass, aluminum, etc. are obviously easier to work with but not suitable for the core of an electro-magnet.

dowel bends

3. In your video it sounds (the buzzing) like your working coil is driven by AC.  Why are you using DC for the home-built?  To be sure, there are good reasons to go DC... just want to be clear what your objectives are before diving into the details of driving a DC coil vs. an AC coil.

4. In the video, you question the need for the bobbin. The plastic bobbin serves no electrical or magnetic function.  It is for the convenience of neatly winding the coil.  Once the coil is wound, you can remove the bobbin if you can keep the windings in place; not sure how you can do this but perhaps some kind of glue that you drip onto the windings.  Different plastics have different heat resistance but I doubt sewing machine bobbins are designed to handle elevated temperatures!  


Photos (2)

Remember that the iron circuit also has to be nearly complete-- the gap between the pole piece that carrier the winding, and the coupler armature should be the only noticeble gap.  Also, the magnet will not accuate wound coil couplers.  Recall the 6o17 uncoupling track (0-27) is a sheet of steel with a dimple to hold the coupler coil, while the long edges are folded to form the two outside rails.  Non-magnetic materials cannot be in this mag circuit, which must form a nearly closed loop.

Common non-magnetic metals include soft, bendable stainless steel (austenitic, iirc} of Gargraves stainless (the tinplate is okay), Atlas O rails (aluminum), brass rail, zinc or zamak castings,  Common magnetic materials include tinplated iron or steel, soft iron, silicon-bearing iron, galvanized iron or steel, also chromed steel.  Cobalt is magnetic, often found in permanent magnets, as is the alloy Alnico.  German silver (cupronickel) rails are not magnetic.

Air is magnetic, however it is about a thousand times less so than the common materials used in making motors:  soft iron (DC); silicon-bearing iron (laminated sheets for AC),  In general air gaps must be kept as small as practical.  Copper windings are not magnetic, however when carrying a current the magnetic fields in air gaps exert a force on the moving electrons, which they in turn exert a force on the copper wire carrying the current.  Parallel wires carrying current in the same direction experience forces that repel the wires, hence coils are tightly wound and anchored.

It is fairly certain that with a plastic (non-magnetic) roadbed you are lacking a complete magnetic circuit,  This must be supplied.  A look at a factory-made uncoupling section may show how this can be done.  Overheating should be prevented by using  a  push-button switch and a brief contact only.


Pennsylvania & Ohio rr conway yard posted:


Reason for the DC  I couldn't get enough of a cold even drop the magnet even up close on the workbench...


I tried reading this several times and still not sure what you are saying - perhaps a typo in here? 

Did you try the same AC supply used in the video?

As pointed out above, there are many types of steel which are better/worse for use in an electromagnet.  I'm coming up empty with a practical method to confirm your particular material is suitable.    I was thinking maybe a small magnet brougt near your unpowered coil core vs. near a Lionel coil core and somehow confirming the magnetic attraction is comparable - very touchy-feely to say the least.  I have magnetic field strength probe which measures how strong a field is so I can compare different materials, different amounts of voltage, etc.. but that's why I'm saying practical.

In my opinion, 300 turns brings you in the ballpark but it sounds like you're not even close.  Note that simply adding more turns works for and against.  Without trying to confuse matters with too much math, but what you're after is the so-called Amp-Turns product... or # of Amps times # of Turns.  It's this multiplication that determines how hard the coil pulls.  Double the Amps doubles the pull.  Double the Turns doubles the pull.  Double the Amps but halve the Turns and it stays the same.

Point being, if you double the turns to 600, you are at least doubling the length of wire you need.  This increases the resistance of the coil so for the same applied voltage, the current drops in half.  So the final Amp-Turns product actually remains the same and you're no better off!  It can be worse than that since windings further from the center take more wire to make 1 other words it might take even more than double the wire to double the turns.  So the Amp-Turns product may actually decrease by adding turns.  Very confusing indeed.

Can your DC supply provide suitable current at more than 12V?  All things being equal, double the voltage doubles the current which doubles the magnetic pull. 




Pennsylvania & Ohio rr conway yard posted:

It's been several weeks so I figure these should have come in by now from Asia.  How did they work out? 

I notice they operate at only 3 Watts (12V at 1/4 Amp) whereas all the uncoupling coils I've messed with require about 10 times that (30 Watts or so). 

uncoupling coils


Photos (1)

I too am curious how these worked out? 

I just found this thread because I was thinking of this exact thing today, and figured to search to see if anyone else had been successful. 

CFO of the Progress and Linglestown Short Line RR. "Hauling nothing, nowhere, on time, everyday!" 

Unless you live in the land of no train shows, why fool around with this stuff?  You can buy old postwar O27 uncoupling tracks for $1 or so and harvest the electromagnets from them.  Solder on some extension wires and install in your track (I use Gargraves).  I have dozens and dozens of these on my layout.  I just bought 6 of the tracks (for a total of $5) at a show yesterday for a friend of mine.  You want the tracks without the postwar controllers, lots of times they are in under the table junk boxes.


Add a momentary pushbutton for control and uncouple away.  The buttons labelled "U" are the uncouplers.



Photos (2)

Add Reply

OGR Publishing, Inc. PO Box 218, Hilliard, OH 43026 330-757-3020