Hiding the Uncoupling Magnet in Your Yard Tracks

Yep, not sure how your wife will take it.  I could probably get away with it, I'm the one that has to fix the sewing machine if it malfunctions.

Rather than a sewing machine, just use a drill motor, drill press or lathe to hold the bobbin. No need to count turns if you know the wire gauge and coil resistence. Get a spool or bobbin close to the UCS in dimension (OD, ID, Thickness) use a nut and bolt to secure the bobbin to your motor and fill it with wire. Number of turns or resistence is not critical. Close counts. More wire with added drive voltage will make a stronger magnet. All of the variables being equal the magnet's strength is approximately the current times the number of turns.

Anyone needing quantity this would be easier than cannibalizing UCS tracks.

Pete

Yes, had we thought of the bobbin idea, we'd have probably bought some.

Steve, your 97 looks fantastic.  I went to the old thread and there were a couple of really good looking camo jobs for the coal loader.

I made some measurements on a Realtrax UCS which might give some ballpark numbers.

Please check my math but here's my swag at # of turns:

The magnet wire is ~0.013".  At the core, the radius is 0.12".  At the outermost coil, the radius is 0.4".  So one "layer" of coil has (0.4 - 0.12) / 0.013 = 21 turns. The coil is 0.4" high so there 0.4 / 0.013 = 31 layers.  Hence about 600 turns.

I measured the coil at 5.3 ohms.  Let's assume 0.013" is AWG 28 or 65 ohms/1000 ft.  So that's about 80 feet of wire.

Try it the other way.  The average turn is at a radius of (0.4 + 0.12) / 2 = 0.26".  So the average turn has a circumference of 1.6" or requires 1.6" of wire.  80 feet = 960 inches.  960 inches at an average of 1.6" per turn = 960/1.6 = 600 turns. 

My estimate is 600 turns, requiring 80 feet of #28 magnet wire.

That eBay bobbin says: "Size: 12x10x2.4cm / 4.7x3.9x0.9inches".  4.7" seems like an enormous dimension for a UCS coil?

Keep the good ideas coming!

BTW, if the spools (e.g., sewing machine bobbins) were not the same size (diameter + depth) as the UCS spool, wouldn't a couple of the 'bobbin' coils stacked on a longer core give you more strength for the same voltage?  (Being sure, of course, you spliced the coils in series?) 

It would be interesting to have some experimentation with all of this to give some direction for the DIY-coil crowd.

Just saw an ad for a yarn bobbin.....3.5" in diameter!!  Fill that sucker with #28 magnet wire and your couplers will open for sure, for sure!

KD

That will work John.

I don't understand those bobbin dimensions.  These are from our Bernina sewing machine which I'm told would be similar to (though not the same) in dimension to other sewing machines.  Definitely not inches in diameter!  Would want a plastic or non-ferrous bobbin too.  I'd think two plastic disc could be drilled with the diameter of the steel pin core and glued into place eliminating the need for a bobbin.

John, I was blown away with all the wire in this coil.  

If you want to be precise with the length of the wire....I used to have my fishing reels wound at my local tackle shop. They had a counter with a spool on the end to measure the line as it went on the spool. Just wrap the wire around the spool before running it on to the bobbin. Would work perfectly for this project.

That looks like a handy tool but again there is no need for this level of precision. Consider that most of these USC tracks are powered by track voltage. They would work between 10 (maybe lower) and 20 volts. The magnetic strength will vary with voltage though usually not in a linear fashion due to saturation. The material you use for the core will have a much greater effect. The ideal material is soft iron (see ebay and science project suppliers). Other exotic materials are better but not worth the money and effort. Using a steel screw will reduce the field strength. Also extending the core much beyond the windings also reduces the magnetic force. How much to be determined.

I have a gaussmeter at work. If I get a chance I will try and make some measurements.

Pete

Like others have stated, the dimensions are bogus for the bobbin.   I took a Brother bobbin and measured it.

20mm in diameter, 12mm thick, and the inner core is 8mm in diameter.  This is very similar to the uncoupling magnet coil.

Tom Densel posted:

...Does the length of the core affect the strength of the magnet?  If not, I'm thinking of mounting the coils underneath the table top and extending the core up through the table top and roadbed.  This would only require cutting a small section of rail and drilling a 1/4" through hole.  This would make retrofitting them much easier as I wouldn't have to tear up a bunch of ballasted track.

How long would the "core" need to be to reach under your table?  As Norton points out, there are forces of nature involved.  The core is funneling or concentrating the field because the steel shank offers a magnetic path of lesser resistance than air.  But there are limits.

Anyway, some other thoughts about a DIY coil.  It appears on eBay it's about $1 for 100 feet of 28 magnet wire in small quantity (to wind a few coils).   But you might have or be able to find a different gauge cheaper.  As Norton also pointed out way back it starts with the so-called Amp-Turn product.  So if you hit a 5.3 ohm coil of 600 turns with 14V, that's about 2.6 Amps (=14/5.3) and you have an Amp-Turn product of 2.6A x 600 turns = 1500 Amp-Turns.  Within reason I could increase voltage 10% while decreasing the number of turns by 10% and achieve the same pulling force on the uncoupler.

Separately, there's the issue of over-heating or the dreaded melted uncoupler magnet.  This one also requires some DIY consideration.  You go directly to jail and don't collect $200 if you run 2 or 3 Amps of current through #28 wire.  I wonder why.  So heating power into a 5.3 ohm coil would be V x V / R = 14V x 14V / 5.3 ohms = 37 Watts.  OMG!   Is it any wonder you should only briefly hold down the Uncoupler button!

So back to original question on extending the core.  This would require some experimentation but safe to say you will need more Amp-Turns to achieve the same pulling power.  Practically speaking I think one would go to Home Depot and buy a 25 cent bolt of suitable dimension as opposed to finding some pin with a special alloy or whatever.  And you might only find, say, #26 wire at your store.  Then you can do the math with how to set the Amp-Turn product with different wire size which of course has different resistance.  The larger wire size means fewer Turns on a same-sized bobbin....offset by increased Amps because of the smaller per-foot wire resistance.

I don't know what all variations of bolts are out there but perhaps one could use the threaded portion in some clever way to assist mounting or to, say, screw on some plastic discs with threaded hole to serve as coil holders...

The cores are easy.  Home Depot.  Go to the metal section and buy a length of 1/4 rod.  You will get many cores out of the 6 foot length.

Gunrunner John, and everyone, these are great ideas, Thank You for this Thread...

I was just going to skip uncoupler magnets, but now this has got me thinking.  Perhaps it's time to research about making them and some experimentation.

You have seen a few variations in this post.  Pick what you like.  Not a lot of money involved.   I run a passenger only layout and back the trains into my 23 track terminal and hit the un coupler first to fire the car coupler then the engine electro coupler.  I then pull the engine out of the yard and back it down onto the turntable.  Whenever I want to bring another engine in the terminal and hook onto that train, the coupler on the car and engine are both open and everything (most of the time) comes together.  I have many couplers on my main line also.  I can leave a train on the main  and bring the engine  to the turntable.  I then bring out any engine I want to join that train.  

Having traveled to Pennsylvania and New York many times, my favorite operation was when we pulled into New Haven and the diesel left the train on the main.  The train went silent inside.  Next I would see a GG-1 sliding by us on the next track and it would hit the switches and back down and hook to our train.  After hookups were made we were off at instant speed.  I duplicate this often on the layout by using any one of my New Haven diesels or I-5 Hudson.   Steam/diesel to GG-1 or GG-1 to diesel/steam.  

If you have no need to uncouple cars, no big deal.  If you want to do any operation, give the un couplers a go.

Is there a better way we can do this?  With all the great minds we have on this site, I think we can.  Food for thought & discussion, could this be done inside the car by using track power to run the magnet and a sensor that could be turned on & off?  Just having some fun.

smilesrs posted:

...could this be done inside the car by using track power to run the magnet and a sensor that could be turned on & off? 

Not sure I understand your idea. 

Are you suggesting an active electro-mechanism in each car?  And the electro-mechanism is like an engine's operating coupler (spring-loaded solenoid)?  Do you see each car having powered-trucks to get track power or have some idea on how to daisy-chain track power from car-to-car such as disguised steam-brake lines or whatever?  What is the "sensor" sensing?

Might as well just equip every car with a TMCC dummy locomotive module and electrocouplers, problem solved.  It's probably just as cheap, and I'm 100% sure that it'll work.

I read all this and ran out and bought me an uncoupling track and a operating/uncoupling track.     Wait.   I don't have any operating cars?  I thought I had an ACME gondola with the coyote and roadrunner ..   I got it out of the box and it works off of gears driven by the axle.  ..   HA   

Jim

The "operating" (4th/5th rail) part of operating/uncoupling track sections reminds me of a recent thread where a guy proposed activating his operating cars using the uncoupler magnet.  IIRC there were clearance issues with the slider pickups to the 4th/5th rails.  The idea was to place a magnetic sensor in the car, position car over the uncoupler coil, then activate car by pressing "uncouple" instead of "unload". 

Perhaps the less obtrusive appearance of these coils whether modified or DIY would make it more palatable to put more of them around a layout.  For example, the inconveniently located on/off switch (under the chassis) for passenger car lights could be replaced with a magnetic sensor.  Position car over an uncoupler coil to toggle lights on/off. 

Steve-short answer not worth it.  Do what John and others outlined and be done.   Why fix something that is not broken.  Anything can be done limited by your wallet.  I have had my uncoupling magnets in the layout from around 1984.  They all work perfect.

I "rolled my own" many years ago to use with Gargraves track.  Just make sure you use the same gauge electromagnet wire and number of turns that Lionel used originally or the flux generated will not do the job.

I was about to use Atlas "O" uncouplers this time, but this thread has inspired me to do it the "old" way again.

Thank you Gunrunner John!

Hal

We just used Gargraves uncoupler sections, having way too many other issues to resolve, and not intending to build a "scale perfect" layout, we opted for the functionality and ease of installing the "stock" uncouplers.  Have to say your method sure looks good hidden.

stan2004 posted:

Tom Densel posted:

...Does the length of the core affect the strength of the magnet?  If not, I'm thinking of mounting the coils underneath the table top and extending the core up through the table top and roadbed.  This would only require cutting a small section of rail and drilling a 1/4" through hole.  This would make retrofitting them much easier as I wouldn't have to tear up a bunch of ballasted track.

How long would the "core" need to be to reach under your table?  As Norton points out, there are forces of nature involved.  The core is funneling or concentrating the field because the steel shank offers a magnetic path of lesser resistance than air.  But there are limits.

Anyway, some other thoughts about a DIY coil.  It appears on eBay it's about $1 for 100 feet of 28 magnet wire in small quantity (to wind a few coils).   But you might have or be able to find a different gauge cheaper.  As Norton also pointed out way back it starts with the so-called Amp-Turn product.  So if you hit a 5.3 ohm coil of 600 turns with 14V, that's about 2.6 Amps (=14/5.3) and you have an Amp-Turn product of 2.6A x 600 turns = 1500 Amp-Turns.  Within reason I could increase voltage 10% while decreasing the number of turns by 10% and achieve the same pulling force on the uncoupler.

Separately, there's the issue of over-heating or the dreaded melted uncoupler magnet.  This one also requires some DIY consideration.  You go directly to jail and don't collect $200 if you run 2 or 3 Amps of current through #28 wire.  I wonder why.  So heating power into a 5.3 ohm coil would be V x V / R = 14V x 14V / 5.3 ohms = 37 Watts.  OMG!   Is it any wonder you should only briefly hold down the Uncoupler button!

So back to original question on extending the core.  This would require some experimentation but safe to say you will need more Amp-Turns to achieve the same pulling power.  Practically speaking I think one would go to Home Depot and buy a 25 cent bolt of suitable dimension as opposed to finding some pin with a special alloy or whatever.  And you might only find, say, #26 wire at your store.  Then you can do the math with how to set the Amp-Turn product with different wire size which of course has different resistance.  The larger wire size means fewer Turns on a same-sized bobbin....offset by increased Amps because of the smaller per-foot wire resistance.

I don't know what all variations of bolts are out there but perhaps one could use the threaded portion in some clever way to assist mounting or to, say, screw on some plastic discs with threaded hole to serve as coil holders...

To add to this thought - Home Depot item - #10-(7/32") diameter center hole nylon washers -  for the threaded portion of the bolt for the core winding- twist it on

edited for #10  washer which should screw on to a 1/4" bolt

Look up Janome plastic bobbins,

Is anyone seriously trying to roll-their-own uncoupling coil?  If so, post where you are and we can compare notes.  My objective is to completely fit the electromagnet in the cavity of a Realtrax section so you don't have to drill into the base of the layout to extend the core length.  There are many electrical and mechanical considerations but I assumed I was the only one pondering this (the thread went cold) so haven't bothered to post any additional thoughts...

The Home Depot nylon washer is a great find...but is 1/8" thick which takes precious height away if trying to fit the entire "bobbin" in the ~1/4" available height under a section of Realtrax.  Same applies to sewing machine bobbins.

Stan,

I am building a layout now with a friend. We need about 8-10 uncouplers under or in Gargraves track. I wanted to experiment under table core mounting with the extended core. We have 1/2" OSB and 1/2" ceiling tiles + the 7/16" or railhead height.

I don't anticipate it to work. So, rolling our own and installing using NPOG's method may be the answer.

I was concerned about plastic melting easily.

I'll speak to him about his desired mounting method. I may be able to experiment with your objective in mind as it would provide a cleaner install on the Gargraves track.

Trust me, the extended core didn't affect the uncoupling strength, those uncouplers work great in the yards.   The only messy part is having the coil below the level of the table, but drill a hole and you're all set.

stan2004 posted:

Is anyone seriously trying to roll-their-own uncoupling coil?  If so, post where you are and we can compare notes.  My objective is to completely fit the electromagnet in the cavity of a Realtrax section so you don't have to drill into the base of the layout to extend the core length.  There are many electrical and mechanical considerations but I assumed I was the only one pondering this (the thread went cold) so haven't bothered to post any additional thoughts...

The Home Depot nylon washer is a great find...but is 1/8" thick which takes precious height away if trying to fit the entire "bobbin" in the ~1/4" available height under a section of Realtrax.  Same applies to sewing machine bobbins.

Stan,

the washer is .032 thick, so it's only a 1/16th for 2 of them. I changed the washer to a 7/32 hole version . 3/8" dia.

You have to order for store pickup.

I had a piece or two of RealTrax. I'll have to find it.

Right.  The Realtrax scenario is definitely a "size matters" situation.  After mess around with this a bit, my next step was going to be to wind the coil on a bobbin that had remove-able end-caps (or washers in this case) and figure out a way to apply glue or whatever.  Thus the coil would have no thickness from the bobbin end-caps.

Anyway, in your scenario what is the length/diameter of the steel rod core and how tall/thick is your coil (placed at the far end of the rod)?  I'm sure there are fancy electromagnetic simulation programs which can figure these things out but for us mere mortals I might be able to jury up a simple test to measure the loss of magnetic pulling force with an extended core.  I have a magnetic probe that can measure field strength so I can cobble together a simple fixture to confirm you're in the realm of reality.

stan2004 posted:

Right.  The Realtrax scenario is definitely a "size matters" situation.  After mess around with this a bit, my next step was going to be to wind the coil on a bobbin that had remove-able end-caps (or washers in this case) and figure out a way to apply glue or whatever.  Thus the coil would have no thickness from the bobbin end-caps.

Anyway, in your scenario what is the length/diameter of the steel rod core and how tall/thick is your coil (placed at the far end of the rod)?  I'm sure there are fancy electromagnetic simulation programs which can figure these things out but for us mere mortals I might be able to jury up a simple test to measure the loss of magnetic pulling force with an extended core.  I have a magnetic probe that can measure field strength so I can cobble together a simple fixture to confirm you're in the realm of reality.

The longest would be - 1 7/16" extended out of the core + the .4 of a 600 wrap core. That like an 1 7/8" total, yes?

it also look like the insulating varnish is mail order only.

What were you going to do to reduce the depth? Make a larger diameter?

Moonman posted:

...What were you going to do to reduce the depth? Make a larger diameter?

This is definitely a work-in-progress.  The mechanical/space fit is non-negotiable in my mind.  I'm of the mind there's a rapid drop in effectiveness of the outer windings as diameter increases.  I'm thinking it will involve some electronic intervention...possibly involving a DC supply voltage, possibly some type of capacitive charging, who knows.  I know there are a few guys that use DC but they are few and far between.  But the point is the existing AC coils created a half-wave rectified sine wave pulling field.  This is where I think there's opportunity for fiddling.

The existing Realtrax coil is about 0.4" high.  If I only have 0.25" in the cavity, that's quite the deficit but it's less than a doubling.  So between picking up, say 40% from capturing the peak voltage (using some kind of capacitor charging), and requiring the uncoupling supply to run up to 20V AC rather than, say, 14V AC accessory voltage, I think I can get it back.  If operating at higher voltages, the electronic solution would include power limiting to prevent over-heating.  So there are a variety of ideas that are not expensive - especially considering the price of an off-the-shelf Realtrax UCS with the unsightly 4th/5th rails for the Unload function which I'm not interested in.  After all, per the theme of this thread, what's the point of hiding the uncoupling magnet only to make the 4th/5th Unload rails that much more obvious!

I even stocked up on rare-earth Nd magnet cylinder/rods ~1/4" diameter in a probably mis-guided attempt to create a pre-biased core.  In practice this magnet cylinder would be placed "in series" with the extended core.  In other words there's a constant DC field of, say, 25% of the force required to uncouple at all times.  Then when activated, 25% fewer Amp-Turns (in my naive way of thinking) is required to generate the required uncoupling force.

Moonman posted:

...

The longest would be - 1 7/16" extended out of the core + the .4 of a 600 wrap core. That like an 1 7/8" total, yes?

I was thinking about your situation so here's a bolt that extends out about 1-7/16" as you specified.  I measured the field strength above the top of the bolt.  The probe was placed 1/8" above the bolt to simulate the distance from the top of the center rail to a car's uncoupler disc.  I slid the coil up and down the bolt to see how the pulling force varied. Relative to the coil being flush to the top of the bolt, the field strength dropped in a roughly straight line down to about 25% when the coil was at the bottom.  This could be a problem! 

Stan, your bolt looks a lot longer than our design.  In the one I presented, the length past the coil is less than 1/2".  I can readily see that 1.5" would be pushing things a lot.  As Marty previously posted, there's a lot of wire on those coils, is your 300 turns an accurate representation of the actual coil?  Perhaps with a long bolt, you need more turns and/or heavier wire to give you more ampere turns for more magnetic field strength.

The issue I was addressing was Moonman's previously stated idea of mounting the coil completely under the layout and just running the narrower diameter core up to the track.  In his words, "We have 1/2" OSB and 1/2" ceiling tiles + the 7/16" or railhead height."  I would think it much easier to drill just a 1/4" hole for a steel rod...vs. 1" holes or cavities to place the coil higher/closer to the rail top.

The other take-away, perhaps obvious, is you want to place the top of the core as high as possible.  The drop off in magnetic field strength through air is profound for this application.  As mentioned I measured the field strength about 1/8" above the bolt.  If you double the gap to 1/4", that extra 1/8" cuts the strength about in half! 

As to the coil configuration, I realize whatever Moonman does will be different from what I am messing with for a DIY Realtrax section where the coil height is constrained to completely fit in the cavity under a track section.  OK, perhaps there is nothing new here...I thought I was doing my good deed for the day. 

That's why we put the core level with the center rail.  I always hated the projecting coil on many decoupling tracks, having it level with the rails is almost as close and works well.

As for your Realtrax idea, seeing the ugly one that MTH makes, I can understand why you'd want to improve on it.

Thank you, Stan.

I have been lumberjacking and finishing yard work. There were two big wind storms in our area , one in the fall and one in the spring that created a tree mess. I finally decided to get it done. Not fun stuff, you know.

Anyway, I didn't get further than acquiring the materials.

Your testing tells me at least I can burying the coil in the ceiling tile which is easy carving and have the bolt only up through the track.

You have a nice looking track now.

Our coil is just below the ties on the Gargraves or Atlas track, and it works very well, so I suspect you should have no issues with the same configuration.