Three options for controlling uncoupler magnets... Update: Finally got a round TUIT ;-) New video added!

Question?  If you have the recessed buttons to avoid hitting them accidentally, what's the purpose of the rotary switch?  I'd think I'd rather have a button for each switch on a small map of the track section.  Also, unless you're actually running a train over them, accidentally hitting one isn't a big deal, is it?

John the rotary would select the magnet, and the button would be in series with it, to power the magnet. The one button could control up to 12 magnets with the rotary (one at a time). Leaning on the button isn't about accidental uncoupling, it's about protecting the magnets from over heating.

The small buttons I showed in method one, would kind of turn my fascia into Swiss cheese

By the way, none of the buttons will be recessed, that's why I was thinking of the pipe ring.

P.S. I haven't totally dropped the ball on my signal test car. My LHS parts guy is only there a few hours on the weekend (PITA). I have to go by there both Saturday and Sunday so I hope I can catch him. If not I'll just order the R2LC.

OK, personally I'm not a fan of a rotary switch in the mix, I'd just have a button for each track.  That makes the operation mnemonic, you just pick the track on the picture and press the button.  Obviously, this is just my opinion, but if I were doing it, that's how I'd do it.

Remember, you did ask for opinions.

As far as leaning on the buttons and frying the coils, you could recess the whole panel.

I totally get it John, there's no need to interpret a diagram to figure out which position to set the rotary. The one to one button method is just direct and intuitive.

I'm just thinking what will be involved in mounting all those buttons. Keep in mind that my fascia is 1/4" Masonite over the 1x4 benchwork. Every hole will have to be drilled out from behind without penetrating the Masonite, probably using a spade or Forstner bit. Then I have to drill a smaller hole , and hope I can get the retaining nut on (times 70+).

The rotary concept reduces that problem to 19 or 20 back drilled holes. The arcade buttons are easy, because they have over an inch of thread on their bodies, or can be glued in place with silicone, all from the front.

There's the trade off. Thanks for your opinion.

I do it this way, mount switches into a plexiglass panel. I wire a 1/2 amp hold/1 amp trip resettable fuse into the feed for the uncouplers, it trips in about 5 seconds so nothing burns out when I lean on the buttons. 

I too thought long and hard about the selector switch, but decided against it.

Nice job. 

I'm a fan of option 1 with individual pushbuttons near the site of uncoupling.  That's what I use for all of my uncouplers.  Each uncoupler button has a little "U" above it on the panels. 

With individual buttons, the operator can tell at a glance which button to press for a given uncoupler.  They can push the button with one hand while operating the Cab-1 knob with the other to get a good uncouple.  Option 2 requires the operator to look at the panel to properly aim the probe.  Moving a locomotive at the same time would be difficult.  Option 3 is cumbersome: determine which uncoupler is needed, turn the rotary switch, then press a button.  Classifying cars to different tracks would get old for me using this method.

By the way, since the above photo was taken I've gone to yellow buttons for uncouplers, yellow label U's above the button on the panels and a tall tuft of yellow grass next to the uncoupling magnet on the layout.  These are the only tufts of yellow grass.  So, yellow grass leads to yellow button with yellow label.

We've been operating regularly with exposed pushbuttons since 2004 and in that time two (2) uncoupling magnets have been cooked.  Given that there are over 60 uncoupling magnets on the layout, that's not too bad.

Gotta admit, this is an option I never thought of, Elliot.  Zero moving parts pretty much makes it indestructible.  Probably 9 out of 10 railroad guys would know what the panel was but wouldn't know how it works.

Bruce

Thanks guys, I see a pattern developing here. No takers on the hot probe huh?

PCLProf, I like the resettable fuse, that overcomes one of my fears of the push buttons.

Bob, I really value your opinion on this because you are actually doing the kind of operations I hope to be doing in the not too far off future. I finally got a chance to see your video. I can only hope my layout turns out half as nice.

Whatever I do, I am limited by the size of the fascia. It's about 8" wide on the upper, and varies on the lower, mostly in the 6" range with some spots wider.

FWIW, and from one with no switching experience, I think the pushbuttons sound easier than the rotary switch and pushbutton. Making the pipe extension for protection against unwanted presses sounds like a good idea to help prevent burning things up. PlcProf's suggestions of the resettable fuse sounds good as well, would probably remove the need for the pipe extension and be easier to install. I like the pushbuttons with the long threads, 1/4" material with most of the miniature pushbuttons having short threads is a pain.

Yeah Bruce, that was an old idea I picked up somewhere. Unfortunately, for this application, it's too clever by half. Now if you wanted to use it to throw switches, it might be better suited. I just thought I'd throw it into the mix here, because there was a time I was considering it.

It seems like the harder it is to install, the easier it is to use. When it comes to easy to install, the hot probe takes the cake.

Up to some more fun, Elliot?  

Out of the options proposed, I like the idea of the separate buttons for each magnet the best... added to the idea of the 'Fire' button from the rotary switch idea.  In each area place a "fire" button in addition to the various uncoupler buttons, so that you need to press both for the magnet to fire. This would limit accidental firing.  Optionally you could use a 'master power' switch so you can turn off the button control when you are not using the couplers to prevent accidental firing, and still only need one finger to run things. 

While the small buttons you pictured are cheap and easy, I'd also go with a more finished/flat style button so that they wont snag on  your clothing or things like that.  The button you pictured is also they type I see usually only rated for .5A current.  I'm unsure what the draw is on the magnets, but I know it is pretty hefty, so .5A may not be enough.  

I found these relatively expensive buttons that look really nice and come in a variety of anodized color finishes as well as the standard polished look: Item # 172027321643 on the auction site.  

While they are a bit pricy ($1.10-2.50 ea., depending on the finish color), , I like these because they would mount almost completely flush with the surface.  Also looks like they have screw terminals which might be easier than soldering all the wires... I've managed to melt more than one of the el-cheapo buttons before when soldering them.  they say they are rated for 3 Amps.  

I found a similar button in plastic for about 54 cents each here, in quantities of 20:  Item# 380885428252

I'm sure spending more than a few minutes looking will yield more options and lower prices, but the point is mostly to chose a low-profile button that would be difficult to press by mistake, and won't get snagged on folks passing by.  

Now to put on my official hat of making things complex...

Are you planning at all to be able to control the uncouplers from TMCC/JMRI/C-MRI/etc?  If so, you may find it easier to build a small box that speaks to the command system with maybe a small number keypad to punch in the number of the uncoupler and a fire button.  you could place these keypads about the layout, or build one with a radio that you could take with you like a remote.  You could also build a controller for the couplers as a stand alone unit from your other controller if you wanted.  Perhaps use a standard universal TV/VCR remote as a controller with IR sensors places about the layout, though I do tend to prefer using 2.4GHz radios over IR.  

One last fun and crazy idea... A variation on the 'hot probe' idea with a rather complex twist:  You could mount a tiny photo photo-transistor near each uncoupler on the layout, and then use a laser pointer aimed at them to trigger magnets.  The electronics gets a bit harry, but it would be neat.

Oh yes JGL, it never ends on this layout. Always something else to think of. Don't worry, all of this stuff is local manual control. No TMCC on this project. The fancy stuff is reserved for the mainline.

Those low profile buttons are pretty cool, but they seem like rather large diameter, and space on the fascia is at a premium with this project.

The small buttons are cheap, but they aren't easy. Just discussing this, I remembered that I have a set of brad point drill bits with depth collars. That may be the way to back drill for those buttons without punching through the Masonite.

There is also the form vs function to consider. I'm trying to keep everything literally black and white. While function has to take priority. I'm planning white lettering and line work, and I just realized that the white button won't show up on the line, but black might not work either as it would blend into the background. The red that Bob said he swapped out for yellow, does pop. Orange would work too. Those cheap small buttons come in every color except pink and purple.

Keeping it strictly old-school, then...

The buttons I pictured say they are 16mm through hole, which equates to a hair larger than 5/8 inch.  The ones you show are usually 1/4 inch, so about twice the size, over all...could be a problem with trying to stuff a whole mess of them in a small area...

Just how many magnets do you have in what amount of area?  I know your layout is HUGE, but it might be the couplers are not all so close that you have to place all the buttons super close together.  

Another option might be to build a control panel in the area of the yard that folds down, or slides in under the table when not in use?  

Just tossing things out there at this point.  

JGL

There are just a few spots that will have heavy density of magnets.

Places like the Ford plant, 8 tracks wide,

the Minnesota Commercial yard, also 8 tracks wide,

and Pigs Eye yard, 7 tracks wide, are the worst.

This is Red Wing, it has 10 magnets, but they are pretty spread out.

I'm really trying to get rid of control panels, if I can. All the mainline switch panels will be going when the C/MRI and JMRI are up and running. And that brings us back around to a discussion we've already had.

Hmm...  Well, I think you can get 8 buttons vertical in 8 inches of facia board even at 5/8 inch, but might opt for using a small piece of metal in that spot rather than just mounting them in the board.  Using something like a 3 or 4 outlet switch-plate cover seems like it would work, and you can get one with a nice finish if you like.  Otherwise, simply mounting the buttons slightly staggered will work out for you, ex 4 vertical, then a few inches to the side the other four offset half way between the height of the first row.  

If going for the look with a track plan under the buttons, I might opt for a piece of 1/4 inch plexiglass sized to blend in with the 8" or so height of your facia.  You can then place the track drawing under the plexi where it won't get damaged. 

JGL

Good morning 

In a yard you can always hook more than one uncoupler to a single button. I have done up up to six like that.

Clem k 

Looks like the rotary switch wasn't a popular idea, I know I didn't like it.  

I still like the mnemonic that has the buttons on the map of the section so that it's intuitive what button you should use.  I think with all the sections with coupling tracks, you'll want the ease of use factor.  If not for you, when you have visitors, it's frustrating when the controls are non-intuitive.

My final word?

The opposite of what Clem said is true.  You can connect more than one button in parallel with one uncoupler.  This would allow you to fire the uncoupler from different locations, i.e., a central panel and a local panel or JMRI.

Jan

If it were me, I  would want to have a button for each uncoupler. As long as the fascia is not glued to the layout, you can remove it to install the buttons. Before removing, drill a pilot hole for each button, through both fascia and benchwork. With the fascia separated from the benchwork, both materials can easily be drilled out to different sizes. Since you had a pilot hole, they will line up when you put it back together. I have had to do minor "deconstruction" to get at things, which is never fun. But if it allows me to do a good job - rather than a botch-job, done leaning over at some weird angle, or with the tool held someplace I can't quite see into - then it's worth it.

Any thoughts about using lower voltage for magnets?  Just enough voltage to reliably trip the uncoupler plate of the worst car.  Lower voltage would let magnet be on longer before burnout.  What about capacitor discharge circuit for magnet, or would the pulse be too short for reliable tripping (need to determine joules and time length for operation, and might be unweildly).  I don't have uncouplers on my layout it all five finger operation, for the rare occasion I do some switching.

Good afternoon guys.

JGL, I'm trying to keep the fascia smooth with no overlays. There are already some things attached to it though. Things like car card pockets, and knobs for some of the manual switches. There will be LED's mounted directly to it as signal repeaters, not sure if there will be turnout indicators, but LED mounting is really easy.

Clem, while that does reduce buttons, I'm not sure I want to go that route as it opens up different issues.

John, you and the others have convinced me that there is only one proper course of action here. Just a matter of working out the specific details on how to accomplish it. I'll probably just order a 5 pack of the push buttons so I can see what I'm working with, and do a test mock up, before I take it to production.

Jan, for this project, JMRI isn't part of the solution. Uncoupling is only done locally by the person doing the switching, so no need for a second button in another location.

Nick, unfortunately, the fascia is permanent, but your method  would make the drilling far less tricky. I had planned to use the small diameter bit to mark the locations first. You do have me thinking though. Maybe with a Japanese hand saw I could remove a small segment of the fascia and use your method. Still a lot of work, but much safer.

RRman,I had always planned to use a lower voltage to power these. I have easy access to it as it is available on my power supplies. All I need to do is run a bus wire around. That along with the resettable fuse should be all the protection this project needs.

I'm all ears guys, I love hearing your ideas, then sifting through them to figure out what will work best. Thanks.

I've added some photos to show you more of what I'm working with here.

This is a close up of the Ford plant, note the blank area between the switch knobs.

Here's a broader shot.

This is what it looks like from the back side. A fair number of obstacles to be dodged.

This is a small piece of Hiawatha. The small yard in the foreground will probably be hand uncouple. The two tracks in the back will need magnets. As long as I stay with that mixed strategy the buttons shouldn't pose too much of a problem.

This would be the brad point drill bit I plan to use.

If you're going to use the capacitive touch buttons, you can just drill right through the whole surface, that would make it easier.   The idea of some sort of time-delay on the coil power is a good one as well, that would remove the issue of cooking them.  I'm guessing that having a limit of a single uncoupling track in one group at a time probably isn't much of a factor, so only one delay circuit for a group.

John, the idea of simple install excites me. What are these capacitive touch buttons of which you speak? And where do I get them?

Here's a 5-channel one that outputs a logic signal.  The $2 relay modules readily available on eBay can be used on the output to drive individual coupling tracks.  For $3-4 a channel, you can have one where you just drive a tack in to the track diagram, touch the tack and it energizes the coupling track.

Standalone 5-Pad Capacitive Touch Sensor Breakout - AT42QT1070

Elliot,

If you are switch from either side of a peninsula you might want to have two buttons.  The idea of having two buttons came from having a central panel and a local panel for switching.

Jan

gunrunnerjohn posted:

Here's a 5-channel one that outputs a logic signal.  The $2 relay modules readily available on eBay can be used on the output to drive individual coupling tracks.  For $3-4 a channel, you can have one where you just drive a tack in to the track diagram, touch the tack and it energizes the coupling track.

It's back!

And you can have fun watching your guests looking all over for the probe!

Seriously, though, this is a really nice idea.

Jan, you'd have to see the track plan to get a better feel for the peninsulas. There's a backdrop that runs down the middle of them, so the two sides are completely separate. Great idea, but it doesn't fit the situation here.

John, I saw that picture where the leads were plugged into apples (fruit not electronics) very cool. So that unit drives 5 relays  which handle the AC magnet power. PERFECT! I'll have to set up a 5 VDC supply and distribution bus.

Nick, that's BRILLIANT!! I knew there was a reason I shared that goofy thing with you guys. It was staring me in the face and I didn't see it, but you did. Maybe John did too.

Couldn't you use some kind of voice/audio recognition into a blue tooth or an arduino board and then into the TMCC or Legacy system?  You wouldn't need to be near the control panel.

aussteve posted:

Couldn't you use some kind of voice/audio recognition into a blue tooth or an arduino board and then into the TMCC or Legacy system?  You wouldn't need to be near the control panel.

Interesting Steve. Are we operating under the theory that if a little tech is good, more is better?

I still think it is much better to keep uncoupling functions local, and away from the larger systems. Voice recognition might be difficult and probably expensive compared to what we've come up with. There's also the issue of multiple operators and multiple commands. That could get confusing.

It's still a cool idea, and someone else may want to try it.

Big_Boy_4005 posted:

John, I saw that picture where the leads were plugged into apples (fruit not electronics) very cool. So that unit drives 5 relays  which handle the AC magnet power. PERFECT! I'll have to set up a 5 VDC supply and distribution bus.

Yep, seems like the ideal answer to the question.

What "we" (mostly you) have come up with is a totally elegant and reasonable costing solution to a kind of tricky problem. Granted, I presented a few extra hoops beyond what most people would encounter, but this has been a fun process.

In terms of cost, I probably have something like $3 each in the magnets, $1.50 for the sensor and just under $1 per unit relay, misc hardware under 50¢, total ~$6. Nice bang for the buck!!

Try buying a fastrack uncoupler.

The touch sensors are really neat, and I'm surprised those modules are so inexpensive.  I remember building circuits for these back in the early 90's with my father, and having a heck of a time using... if I remember correctly... 741 op amps  to get them to work. It made for a really neat button at the time.  We used them on a 'button' system sort of like a game-show buzzer where you buzzed in, and the first person to touch their "button' would lock the others out and turn on their light.   I recall everyone being really amazed with a button with no moving parts.  

Elliot, On the relays, you can also save a few cents by using boards with more than 1 relay.  They also come in 2, 4, 8, and 16 relay versions.  16 relay boards are nice and compact, and run about $12 shipped, the 8 relay boards are the most cost effective right now, however at five bucks.  You'll also save a little bit of hassle, as you'll have a little less wiring.  

JGL

JGL, I knew about the 2's, 4's and 8's. I've had an 8 sitting in my watch list, since John mentioned them a couple years ago, just waiting for the right project project to come along. FOUND IT!!! $5.95 from China for an 8.

It'll take a little planning and juggling to match the switches and relays to the uncouplers in the field without being too wasteful of any one component. That's just a simple puzzle.

As cheap as the parts are, waste a few.   I think this will be super-slick.

gunrunnerjohn posted:

As cheap as the parts are, waste a few.   I think this will be super-slick.

I agree with this.  Maybe pick up a couple 4 relay boards if you want, they don't cost much more that way.  I've got a couple of the 16 relay boards I picked up about a year ago, and paid $20 for them then, from china.  I don't know how these boards can be made so inexpensive... cheaper to buy them then to buy 1/4 of the parts  on them.  

JGL

Actually John, I may have thought of another use for this switch. Because of the way I power my layout, I am unable to use the HALT button. It has no effect. It will probably take one of my 24VDC relays with the heavy contacts to handle the track power. If we can latch and unlatch that relay, the power can be shut down for a single power district. This would be especially handy in yards and industrial areas, where most derailments occur. The operator could probably get to it before the polyfuse kicked in, and could then rerail cars with the power off, then turn it back on without ever disrupting the rest of the layout.

Just thinking out loud.

Update: OK guys, I'm committed to this project. I just pulled the trigger on 20 of those boards. Quantity price break too, down from $7.50 to $6.75. Still working on the final count and positions for the magnets, I'll order the relay blocks when I figure that out.

Well, I probably would have ordered one and tested it, but what the hey, might as well jump in with both feet!

Hey gang, it's been almost exactly a year since we had this discussion, and I thought I would bring you up to date.

The uncoupler project is well under way, with more than half the 74 magnets now embedded in the track.

I thought it would be fun to to start the assembly of the control electronics. There are five uncouplers at St Paul. The small card on the left, is GRJ's capacitive touch sensor. Inputs from the bolt heads on the fascia will be connected to the row of pins on the left. The outputs on the right will trigger the relays, which will handle the magnet power. Those black pieces are female Arduino connectors. I get them in rows of 40 conductors, and just clip them to the desired length, then clean up the ends on my sander.

I just mount the components to some scraps poplar. Actual wiring will be coming soon.