I’m finally making some progress on a new shelf layout. It is an HO pike focused on local industrial switching of a paper mill, and cotton compress and warehouses in approx. 11’ x 26’ around the walls. Beginning laying some of the road bed and picking out likely spots for installing magnetic under track uncouplers. These are the flat magnets with metal backing by Kaydee. On sidings I was planning on just glueing them under the track, but was concerned to do this on a main because of unwanted uncouple. Was considering putting one of these magnets in a hinged fixture so I could control the uncouple with a pull knob from the facia. Do any of you guys do this or would it be better to use the electromagnet uncoulper on main line locations. For my sidings I used a roto zip to cut out pockets in 3/4" ply sub grade so magnets would be close to flush with cork road bed. I’m using N gauge cork for a shorter ballast profile, thus the need for cutting pockets. Will try to post a couple of pics… Any input would be appreciated. 
Bryan,The only time I’ve seen unwanted uncouplings was with herky jerky train movements or if one stops over a magnet with the slack ran in…
As long as the coupler are tight they will not uncouple…To uncouple them you need to stop over a magnet and back up a tad.
Hey Bryan! That’s an interesting concept you have and it might well work. I honestly don’t know if it’s been done. That said, if you have room to drop a permag out of the way ( eliminating any residual field) would you have room for an under-the-track electromagnet? The difference would be with a permag there’s always a field you are trying to move away, with the electromagnet, once unpowered, it’s OFF! The newer Kadee under the track are just the old thru-the-track with added side plates to intensify the field. I’ve been converting my old units by adding 16gauge steel ‘enhancement plates’. It seems to work fine. I’m building a modified “the Virginian” layout for my two daughters, and installed an uncoupler between the two Peco curved turnouts David Popp desiged in. By installing at a carefully adjusted angle to the track , it works fine for the appropriate length cars! (I don’t like uncouplers on curves, I think I’m allergic to that! Regardless, good luck and let us know how whatever approach seems to work!
Morgan
I’ve seen just the opposite. If the last few cars or caboose take a little slack, they uncouple at the most inopportune times. I would never put a permanent magnet on the main without some way of disabling it (slide or hinge).
I have to agree with you on this. I removed the magnetic uncouplers from my mainline for just that reason. All it takes is a momentary stall by the loco and enough slack is created to cause uncoupling. I have these uncouplers in my yards and industrial spurs. My dilemma is the A/D tracks. I like having the uncouplers for breaking down a train, but it is very annoying to have an unwanted uncoupling when a train is entering are exiting the A/D track.
Going back to the OP, I would at least look into KDs electro-magnetic uncoupler. I haven’t used these but others have recommended them to me. They seem like rather bulky contraptions as compared to the simple magents, but might be a better solution than the swinging gate idea. Why reinvent the wheel? The one question I have is how they could be mounted to foam subroadbed. I did one section of the layout with foam and had to mortise a section of plywood into the foam to give me something to screw my undermount switch machines to.
The other option is a manual uncoupling tool. That’s what I use for mainline uncoupling. KD makes a plastic one but you can buy a whole bag of bamboo skewers for about the same price and they work just as well. In a pinch, I’ve also unsed a small screwdriver.
This has been done for many years. Both by sliding to the side and by drpping it down.
Thanks for the replies all! Since my layout is primarily a switching pike, the main line is not a main in the traditional sense of zooming long trains going by. Rather, the short local setting out cars here and there. Starting and stopping a lot, occasionally leaving a cut of cars on the “main” to retrieve empties at factory etc… I have one magnet location hinged and was trying now to come up with an appropriate mechanism. Tried a choke cable but it was too stiff pushing “in”. So I decided to try a wheel that belong on a sliding closet door setup. Will try to tie it all together with dowel rod or small dia tubing. 
The trick is to use one of those long-tailed hinges (sometimes referred to as barn door hinges, but not that big.) One leg gets screwed to the underside of the layout. The other end will need a couple of bends to raise the magnet to the desired height and then drop the end of the tail well below the subgrade to clear the necessary screw eye.
Operation is dead simple. There is a length of heavy fishing line attached to the dropped end of the hinge. It goes through a screw eye close to where that end will be in the fully raised position. Then, arrange a clear run to the fascia, put the line through a hole in the fascia (possibly lined with a bit of brass tubing) and attach a pull handle (large bead, coat button, used 1/4 inch nut…) To activate the uncoupling feature, pull the handle. When loose, the weight of the magnet will drop the free hinge leaf below coupler activation range. An added benefit - the steel hinge will act as a field intensifier, so that separate plate can be left out of the assembly.
Chuck (Modeling Central Japan in September, 1964 - with Kadee couplers, some on steel cars)
I have used them for both HO and N scale. You can save some money by cutting them in half. Use a hack saw for the back plate and an X-acto knife for the magnetic material. I do that all the time. Since our train control systems are really good now days, you don’t need the full size magnets. You should also have no steel axles on any of your cars as the car will be pulled into the magnet and cause operating problems. I mark the magnets location by putting some ties next to them in a small pile.
I misinterpreted your photos in the OP and thought you had pink foam for a base. I see now you have plywood. If you are as far along as you seem to be, I would keep at it until it works or you figure out it won’t. When Edison was talking about his early efforts to invent the incandescent light bulb, he said those early experiments weren’t failures. He learned about hundreds of things that didn’t work.
Chuck, the hinge length is something I had considered but started my first one with a short style hinge. I will look at the hardware store to find some of suitable size. This makes more sense using a longer moving piece to remove the magnet further away from the uncouple area. This will give the moving apparatus more “throw”. On the first go at producing a working hinged section, I first cut out a pocket from the sub grade 3/4 pky with he roto zip. I built a small frame to act as a cutting guide so the pocket would have straight edges. Set the depth so enough plywas cut out so the magnet would be flush with cork roadbed. Make 1 pass around the guide then start working back and forth mowing strips away until I had a clean picket in which to place the magnet. I then set the roto zip deep enough to cut the rest of the way through the ply board. This thinned plug is used as the base of the swinging plate on which the magnet will mount. Now I will try to employ Chucks suggestion by using a longer hinge. This is starting to slowly take shape. I think this is going to work! 
My earlier response seems to have disappeared into the black hole bit bucket. The gist was -
My preference is for pairs of 1/8" x 3/8" rare earth magnets mounted vertically (axially polarized), with tops of opposite polarity. These are inserted just inside the rail, or under the rail. These are much easier to install after laying the track than are the large flat under the track magnets. The tops are just below the tops of the ties so that ballast or dirt can hide them. Of course, some method to mark the magnet location is needed.
On my HOn3 rolling stock, I prefer the MT 1015/1016 couplers as closer to scale size and appearance for my non-Colorado prototype. But I find my eyes and hands are not good enough for using skewers on the smaller couplers as the long term solution. And reaching into the scene with arms and shirt sleeves is not always a great idea as the scenery is developed and detailed.
A single rare earth magnet pair has a limited longitudinal uncoupling window - less than an inch under normal conditions. I find this to be acceptable for my limited main line use. On spurs and in yards where a larger uncoupling window is wanted adding pairs of magnets spaced at about 3/4" - 1" does the job. If there is too much unwanted uncoupling on the main, then I plan to revert to hinging or sliding the large bar magnets (haven’t had this happen yet). I don’t plan to go down the electromagnet path.
my thoughts and experiences, your choices
Fred W
I’ve been a “hands off the layout” guy since I was a teenager. I’ve still got a bunch of horn-hook uncouplers from that layout of 50 years ago, somewhere, including an electromagnet version that uncouples horn-hooks. When I built my current layout, I installed Kadee permanent magnets on sidings and I’ve got 3 electromagnets for my main lines placed at strategic spots. They work reasonably well.
But I am going to make a confession. I was wrong. I now keep a bag of bamboo skewers in my trainroom, and I prefer to uncouple my cars with those. In some cases, the magnets still must do the job as some spots are kind of awkward to get to, but I’ve been converted to skewer uncoupling.
Something I saw some time ago, could be an option.
Rob
For me and my layout its still hands off uncoupling simply because I don’t care to poke a stick between cars…
I found a small flat tip screw driver will do the same work though as the round stick…In fact that’s how I uncoupled X2F couplers many years ago…
I’ve discovered the same thing and the small screwdrivers are much easier to carry around with you as you follow your train around the layout. Unfortunately, I have a propensity for setting things down and hiding them from myself, so the bamboo skewers are a little easier to locate when I do that.
I’ve experimented with a similar contraption which I discussed in the thread “passenger car uncoupling”. The results were not very good. I mounted two KD undertrack uncoupling manets to a similar shaped piece of wood but couldn’t get it to work. Both pins tended to move in the same direction, rather than opposite each other. I haven’t given up on the idea. Maybe using magnets like those in a tongs like device to get the magnets under the car to pull the pins apart is the answer.
And here I thought I was the only one to lay things down and hide them in plain view.[:O]
After hiding three in pain view I started sliding them in the pocket of the car card.
This evening I was able to resume work on my manual uncoulper control. After making a couple of different mock ups of a mechanism to swing the magnet into and out of position, I cobbled together the following out of scrap wood pieces. A 1x2 with a sliding door track wheel on one end and a pivot point with metal bushing on the other end. 
Attached the swing arm to some scrap ply pieces with a 1/4" carriage bolt. Fashioned a push rod and located a hole for it to run through the facia. 
It works! This thing was giving me fits. I just could not get it right in my mind how this mechanism was going to be made. I went back to the keep it simple method…which most times is the best approach. The only thing left is to add stops for the push rod travel so the wheel does not over run the swing plate on both directions. I hope you guys find this interesting and or useful as a method of controlling uncoupling with the under track magnets. Thanks to all who replied for your valuable input! If anyone has additional suggestions on any tweaks / refinements I would be glad to hear them. Thanks again!
Looks very good and reliable, I just used some fishing line and some screw eyes to make it pull in the right direction. I have to hold the knob that sticks out of the facia while uncoupling, but that’s not a problem.