Pneumatic Switch

[sjim]

I have long tried to make a mechanical switch for my Lego trains, but never got it to work properly. Part of this has been that the underside of the switches has been too weak. The resolution so far is pneumatic switches.

You can find a video here:

If you want to see some pictures and download the ldr, go here: http://greytag.se/lt2/pneumatic-switches/

[kieran]

Looks to work well

[Severus A]

Awesome idea. It seems that the needed space is not very big. If I've seen well, the assembly could be easy integrate into a building.

[baard]

Nice idea.

As far as I can see from your pictures the pneumatic cylinder turns a technic liftarm that moves an 8 tooth cogwheel on a rack that moves the switch?

I have found that motors do the same job as long as the rack can be held in place (it is merely resting on top of the switch) by the 8 tooth cogwheel. If you have space underneath your rail, the motors can be hidden. I guess you could do the same with the pneumatic cylinder. Have you tried using the small pneumatic cylinder?

See my Flickr for a motorized version (rails are mounted 2 bricks and 1 plate over the baseplate):

[HiFish]

Hi,

I like the idea of pneumatic switches because it offers a completely different way of integrating into the layout, with tubing instead of wiring, maybe getting a dedicated building for a compressor...

Although its way out of scale it looks a bit more natural like a hydraulic switch, compared to the thing i used with all kinds of gears visible. I used a motor based version in my last exhibition with very mixed results.

All my switches have different ages and "pressure points", after some testing only 2 of those worked somewhat reliably with the design i had which is a adaption of this:

I have a bunch of old 9V Technic motors (1990) with different levels of wear and performance, see here:

The 2 switches i used ended up working 2/3 of the time. The problem was not jamming but simply not enough power. According to this chart it only has roughly half the power of the motor i should have been using, but so far have never seen in sets here in germany (maybe missed them due to dark ages). Sometimes it would help to rapidly move sewsaw style to get it to switch to the other position. Sometimes the problem seemed to be not the motor but the clutch gear giving way to early. I settled for replacying the clutch gear with a regular gear on one of the switches, but that only helped a little.

I would like to test the pneumatic version, but the ldr file i got from the page has no cyiliders and so far i did not quite figuere out how its supposed to work. could you maybe post a couple of close up fotos from the side?

Thank you for your efforts in findig a workable solution.

baard, have you tried your solution "out in the open" or do you only use balasted track that always gives you enough room to hide the big gear and motor?

Edited February 16, 2016 by HiFish

[baard]

baard, have you tried your solution "out in the open" or do you only use balasted track that always gives you enough room to hide the big gear and motor?

Yes, but I also don't like the visible gears, so for a couple of my switches sits a house over the little motor. The obvious disadvantige of any solution that is built above landscape is that it looks bulky and has to be built quite close to the track, thus eliminating the possibility of overhanging trains and cars.

I have also tried the "spear" solution where you don't have to modify the switch at all, but use a minifig spear to push on the little moving piece of the switch (I don't have good name for the part on the switch), but again, the motor that you have to use will have to be hidden somewhere.

I will defenitely look into the pneumatic solution you have posted, as this for a small area can be a viable solution. If you elevate your track at least 1 brick, you will also have room for the tubing.

In conclusion, I think that my layouts in the future will always be built a couple of bricks above baseplate, at least where there are switches. This is of course brick demanding, but hey, then you can add small slopes and gradients that in the end will make your layout look more real. After all, earth is not flat

[sjim]

Thank you for your interest and feedback!

Nice idea.

As far as I can see from your pictures the pneumatic cylinder turns a technic liftarm that moves an 8 tooth cogwheel on a rack that moves the switch?

I have found that motors do the same job as long as the rack can be held in place (it is merely resting on top of the switch) by the 8 tooth cogwheel. If you have space underneath your rail, the motors can be hidden. I guess you could do the same with the pneumatic cylinder. Have you tried using the small pneumatic cylinder?

See my Flickr for a motorized version (rails are mounted 2 bricks and 1 plate over the baseplate):

That is correct. Pneumatic cylinder -> 2 x liftarm (arranged in an L-formation) -> 8 tooth gear -> gear rack

I tried the small cylinder, but failed. I can't remember why at the moment though... :) I'll probably have another go at it soon enough.

If I had seen your pictures before I might have gone that way. It looks brilliant! I'll have a try with pneumatics one day.

Hi,

I like the idea of pneumatic switches because it offers a completely different way of integrating into the layout, with tubing instead of wiring, maybe getting a dedicated building for a compressor...

Although its way out of scale it looks a bit more natural like a hydraulic switch, compared to the thing i used with all kinds of gears visible. I used a motor based version in my last exhibition with very mixed results.

All my switches have different ages and "pressure points", after some testing only 2 of those worked somewhat reliably with the design i had which is a adaption of this:

The 2 switches i used ended up working 2/3 of the time. The problem was not jamming but simply not enough power. According to this chart it only has roughly half the power of the motor i should have been using, but so far have never seen in sets here in germany (maybe missed them due to dark ages). Sometimes it would help to rapidly move sewsaw style to get it to switch to the other position. Sometimes the problem seemed to be not the motor but the clutch gear giving way to early. I settled for replacying the clutch gear with a regular gear on one of the switches, but that only helped a little.

I would like to test the pneumatic version, but the ldr file i got from the page has no cyiliders and so far i did not quite figuere out how its supposed to work. could you maybe post a couple of close up fotos from the side?

Thank you for your efforts in findig a workable solution.

I found that the only way to get it done by motors was to build an enormous 'holding cage' to ensure that the power needed didn't break the construction. In the respect I have to say that baard put me to shame in a brilliant way. :)

In my experience there is no way to have reliable automatic switching without raising the track. The bottom is too weak.

The CAD part is not where I excel. I'll get some more photos posted in various stages of building, no problem. :)

Edited February 21, 2016 by sjim

[sjim]

I adjusted the switch to use a small pneumatic cylinder instead, and also I put pictures of the building process.

Check it out and let me know what you think: http://greytag.se/lt2/pneumatic-switches/

[ALCO]

I was going to create a new thread but then I found this one and decided to bump it as I feel I have new content/ideas to contribute to this topic. Here goes:

I have also long wanted to use pneumatics to operate my switches...mostly because I thought it would be the most realistic and pretty cool. I had a pneumatic crane set a while ago but my sister sold it on ebay while I was in the dark age. Most recently while placing a bricklink order the seller just happened to have some pneumatic parts, so I got them and the project started. I already had modified a couple switches to be a cross-over and thought that would be a good place to start since both switches had to be changed at once. I wanted a clean and realistic look and most importantly, the final result must allow for even the most demanding of clearance requirements. The best I came up with in the clearance department was the BMR PS-1 40' boxcar. The trucks and the bottom rung of the ladder are quite low and stick out pretty far, so that was my test.

My first plan was to have the piston under the switch so it would be completely hidden, but quickly found it was not strong enough to move the switch. This lead me to use the power of a lever and an off-set fulcrum. After many Thomas Edison moments (trial and error) I have found a method that works with actually few parts requirements.

The project: I thought I would start with a cross-over using MILS standard: Baseplate with Brick and one plate, track sits on two plates of ballast. So essentially the track sits on two bricks. I don't have all the parts to make this look pretty as this is a Proof of Concept project and I want you all to see the the details in how this is done. I feel this is knowledge worth sharing to better our community so this is why I am sharing. I did not build out the second switch as I will need more parts from Bricklink to finish that. As it is, I had to "borrow" a couple parts from a finished locomotive for this to work.

Edit: Just some notes on this project, the whole mechanism does come rise above the top of the rail, so any parts of any locomotive/rolling-stock will not come into contact with the mechanism. The 4L half-lever fits nice and snug over the switch lever. the 3mm tube is only 1L and that is the hinge between the 4L half-lever and the 2L half-lever. The pivot/fulcrum point is the axle/pin that connects the 2L half-lever with the 3L "L" shape half-lever. The Plate Modified 2x3's with Hole are arranged as indicated in the "parts" photo below to provide maximum stability as there is a fare amount of force exerted by the piston on the whole assembly. 

I think this captures the scope and requirements of the intent quite well. Clearance requirements are met, it looks and operates "realistically" and provides an extra bit of detail to a landscaped layout.

Below is the final photo:

The switch in each position:

 

Clearance:
 

 

Parts Required:

 

And the Video! https://www.flickr.com/photos/128636628@N08/47741308491/in/datetaken-public/

(Can't figure out how to embed it)

I'd love to hear (read) your thoughts, ideas, comments, questions. 

 

 

 

Edited April 30, 2019 by ALCO

[Hod Carrier]

Interesting. I had considered using pneumatics also but the cost and availability of parts makes it prohibitive. I hadn't intended to use the solution you've designed but rather to have the cylinder operating the point/switch blades directly instead of operating the standard drive-bar mechanism. This would have given a cleaner installation and given me the option to remove the drive-bar altogether if I wished. Admittedly this idea is untried as I don't have the parts necessary to test it.

Have you removed the lugs inside the points/switches that lock the drive-bar in each position? The advantage with pneumatics is that, provided your cylinders are supplied with a constant air pressure in each position, it should hold the point/switch blades in place without any other assistance. It will also make the mechanism easier to drive.

[Roadmonkeytj]

On 4/30/2019 at 9:23 AM, ALCO said:

final photo:

Parts Required:

I'd love to hear (read) your thoughts, ideas, comments, questions. 

I love it ... I am unfamiliar with lego pneumatics ... (Not of the existance but of function as it was come and gone in my Dark Ages)

I know these have lego pumps, but could it be run off say a fish tank airerator? Also do you have to protect from over pressure if the switch is left thrown as you do with electric?

I love the way this looks and could easily hide beneath the landscaping for my switching layout.  Can the cylinder be set below the track level to just have the levers exposed above? Similar to the old manual thrown switch houses that were set back from the track?

 

[ALCO]

On 5/2/2019 at 12:33 PM, Hod Carrier said:

Interesting. I had considered using pneumatics also but the cost and availability of parts makes it prohibitive. I hadn't intended to use the solution you've designed but rather to have the cylinder operating the point/switch blades directly instead of operating the standard drive-bar mechanism. This would have given a cleaner installation and given me the option to remove the drive-bar altogether if I wished. Admittedly this idea is untried as I don't have the parts necessary to test it.

Have you removed the lugs inside the points/switches that lock the drive-bar in each position? The advantage with pneumatics is that, provided your cylinders are supplied with a constant air pressure in each position, it should hold the point/switch blades in place without any other assistance. It will also make the mechanism easier to drive.

Thank you@Hod Carrier! My cost for the piston was $6.50 USD. The hose I got from my local hardware store for $.17 USD per foot. I think the Pneumatic Switch (to control the direction of the air) was much less and the pump was about 3-4...ish. In all it was about $13ish USD with 3ft of hose and a couple T-pieces so less than the cost of a motor as it seems most people use when remotely controlling their switches. I did not remove the lugs inside the switch. The lever give the piston more power much like a crow-bar. Once the switch has been changed I can completely de-pressurize the system and the switch will not move back. the piston only moves the drive bar so the switch operates normally as if the "hand of God" changed it like when you change it manually with the yellow throw. Also, when the piston is not pressurized, it is still possible to manually change the switch. 

23 hours ago, Roadmonkeytj said:

I love it ... I am unfamiliar with lego pneumatics ... (Not of the existance but of function as it was come and gone in my Dark Ages)

I know these have lego pumps, but could it be run off say a fish tank airerator? Also do you have to protect from over pressure if the switch is left thrown as you do with electric?

I love the way this looks and could easily hide beneath the landscaping for my switching layout.  Can the cylinder be set below the track level to just have the levers exposed above? Similar to the old manual thrown switch houses that were set back from the track?

 

Thank you @Roadmonkeytj! I don't know if it can run off a fish take airerator... I have been thinking about that as a far-down-the-road addition. I do not know the max PSI the pneumatic system can handle but I think the hoses would pop off before any real damage can be done. I am planning to get a air tank and regulator in combination with an additional pneumatic switch so when I am finished moving the switches I can shut off the air supply and de-pressurize the system at the same time as a fail-safe to protect the equipment. Maybe over-kill on my part but I think it would be good in the long run.

Using MILS, it would be very easy to hid the piston under the landscaping. The reason I wanted such a simple solution is that I think having the piston exposed adds some cool, realistic, and interesting detail to the layout. Some of my first and working attempts at this had the whole mechanism hidden except for the 4L half-lever. It was then I decided I wanted the piston exposed. 

[Roadmonkeytj]

1 hour ago, ALCO said:

Using MILS, it would be very easy to hid the piston under the landscaping. The reason I wanted such a simple solution is that I think having the piston exposed adds some cool, realistic, and interesting detail to the layout. Some of my first and working attempts at this had the whole mechanism hidden except for the 4L half-lever. It was then I decided I wanted the piston exposed. 

Well right now the track is elevated 4 plates so the manual decouplers fit ... Perhaps i could convert them to air

Any pictures of your fully hidden setup?

[baard]

Nice setup, I like it!

Will definitely try this sometime. Thanks for sharing!

[Hod Carrier]

On ‎5‎/‎2‎/‎2019 at 7:19 PM, Roadmonkeytj said:

Also do you have to protect from over pressure if the switch is left thrown as you do with electric?

For me, this is the beauty of pneumatics. The piston will only extend a certain amount by design and no further. The ones used by @ALCO have a 2L throw, so you simply design your mechanism accordingly. Once it's reached that extension it can't exert any more pressure on the mechanism.

14 hours ago, ALCO said:

Thank you@Hod Carrier! My cost for the piston was $6.50 USD. The hose I got from my local hardware store for $.17 USD per foot. I think the Pneumatic Switch (to control the direction of the air) was much less and the pump was about 3-4...ish. In all it was about $13ish USD with 3ft of hose and a couple T-pieces so less than the cost of a motor as it seems most people use when remotely controlling their switches. I did not remove the lugs inside the switch. The lever give the piston more power much like a crow-bar. Once the switch has been changed I can completely de-pressurize the system and the switch will not move back. the piston only moves the drive bar so the switch operates normally as if the "hand of God" changed it like when you change it manually with the yellow throw. Also, when the piston is not pressurized, it is still possible to manually change the switch. 

Fair enough. It sounds as though you have a different design brief to the one I had come up with. For an individual installation the cost isn't too bad, but I was envisaging multiple installations. You're right that it's less than the cost of an electric motor, but electric motors would not have suited my intended usage anyway.

It would be possible to control the pressure within the system to prevent over-pressure. I've seen quite a few automatic LEGO compressors where the switch for the motor is controlled using a piston acting against a spring or rubber band. As the pressure increases, the piston overcomes the spring and moves the switch to "off". By tweaking the amount of force in the spring you can raise or lower the pressure in your system. And I'd feel fairly confident that a constant moderate pressure would not damage any of the parts.

[ALCO]

2 hours ago, Hod Carrier said:

I've seen quite a few automatic LEGO compressors where the switch for the motor is controlled using a piston acting against a spring or rubber band. As the pressure increases, the piston overcomes the spring and moves the switch to "off". By tweaking the amount of force in the spring you can raise or lower the pressure in your system. And I'd feel fairly confident that a constant moderate pressure would not damage any of the parts.

@Hod Carrier this is a great idea! I had only vaguely thought about a compressor, but to make one with an automatic shut-off is a great idea! Now, how to get it to turn on again when the pressure falls below a certain point... I'll have to check out the Technic threads for anyone talking about pneumatics.

Thank you @baard!

[Hod Carrier]

@ALCO You’re welcome, but the idea is not my own.

The compressor switch is balanced by the piston and the spring. The spring holds the switch on while the piston pushes against it to switch it off. As the pressure builds the piston overcomes the spring, but as it falls again the spring takes over. Therefore the compressor runs automatically and maintains a set pressure within the system without the need for operator input.

Here is one possible design. You can see the pressure switch in the last photo of the original post. It’s the yellow piston attached to a PF polarity switch acting against a red rubber band.

Edited May 4, 2019 by Hod Carrier

[ALCO]

19 hours ago, Roadmonkeytj said:

Well right now the track is elevated 4 plates so the manual decouplers fit ... Perhaps i could convert them to air

Any pictures of your fully hidden setup?

I will work on a hidden setup for. I will also play with your 4plate elevation too. It might take some time though.

3 hours ago, Hod Carrier said:

@ALCO You’re welcome, but the idea is not my own.

The compressor switch is balanced by the piston and the spring. The spring holds the switch on while the piston pushes against it to switch it off. As the pressure builds the piston overcomes the spring, but as it falls again the spring takes over. Therefore the compressor runs automatically and maintains a set pressure within the system without the need for operator input.

Here is one possible design. You can see the pressure switch in the last photo of the original post. It’s the yellow piston attached to a PF polarity switch acting against a red rubber band.

Again, thank you so much! This is awesome! What a great lead! Now I have some more shopping to do.

[Roadmonkeytj]

12 hours ago, Hod Carrier said:

@ALCO You’re welcome, but the idea is not my own.

Here is one possible design. You can see the pressure switch in the last photo of the original post. It’s the yellow piston attached to a PF polarity switch acting against a red rubber band.

oohhhh this is briliant ... Im seriously leaning towards making the whole layout pneumatic now.

8 hours ago, ALCO said:

I will work on a hidden setup for. I will also play with your 4plate elevation too. It might take some time though.

Thanks ... I was thinking to adapt your design to my needs off your pictures but i wont complain if you want to come up with a design as well lol 

[ALCO]

On 5/4/2019 at 7:54 PM, Roadmonkeytj said:

Thanks ... I was thinking to adapt your design to my needs off your pictures but i wont complain if you want to come up with a design as well lol 

I have not checked with the 4 plate height yet but should be fine. I changed out the axle-pin used as the pivot to a 3L axle as it provides more stability for busier switching operations. I also added a couple levers to connect to a position indicator - through or diverging. Just need to move it to accommodate my clearance requirements. Hopefully I can get some photos this weekend and upload them on Monday.

[Roadmonkeytj]

@ALCO 

So perhaps our setups are slightly different.  I built the switch as you have shown.  However I used a 3L bar in place of your tubing.  I can not get the small cylinder to throw the switch. Perhaps it is the air pressure? Or the fact that the switches are brand new and stiff?

[ALCO]

10 hours ago, Roadmonkeytj said:

@ALCO 

So perhaps our setups are slightly different.  I built the switch as you have shown.  However I used a 3L bar in place of your tubing.  I can not get the small cylinder to throw the switch. Perhaps it is the air pressure? Or the fact that the switches are brand new and stiff?

I don't think a new switch should matter... I'd like to see a photo and maybe I can identify what the issue might be.  While I have been working on ballasting this in, I have made some minor adjustments. I changed from the 2x2 tile with pin for mounting the base of the cylinder to two 1x5 technic plates one plate off the baseplate and a 3L axle I have a 1L bush between the cylinder and the plates. This moves the base of the cylinder a half-stud away from the switch on both the X and Y axis and seems to give the cylinder a little better leverage. You might try this for the issue you have. Another thought is that the 3L bar might be REALLY snug in the 2L and 4L half levers, if you have a 3m tube handy, try that and see how it goes. I will be out of town for the holiday so I will put up some more photos next week when I get back. I will also test with a brand-new switch as well and report back.

Also, a new update on this project: I have a working position indicator attached to this mechanism. Green for through and red for diverging. Seems to be a common scheme on American railroads. I'll post a photo and video of it working next week. Also, I've got some of the cross-over ballasted as I'm trying to make it look all "pretty".

[powerwindows83]

is there a reason you went with your current lever setup and not something like this one...

https://www.flickr.com/photos/168003174@N07/48233368557/in/dateposted-public/

the 4s half beam on the switch like you have, then a 5s liftarm connected to the landscape with a pin like your cylinder is, then connect the cylinder directly to the end of the liftarm? the throw for the cylinder side of the lever would be two studs and the switch side would be one. i scratched this together quickly in my lunch break so its not at all polished, but it shows the basic idea of what would be going on in both positions. i feel like this setup would be stronger and less flexible than your current one. just giving ideas! :)

[ALCO]

10 hours ago, powerwindows83 said:

is there a reason you went with your current lever setup and not something like this one...

https://www.flickr.com/photos/168003174@N07/48233368557/in/dateposted-public/

the 4s half beam on the switch like you have, then a 5s liftarm connected to the landscape with a pin like your cylinder is, then connect the cylinder directly to the end of the liftarm? the throw for the cylinder side of the lever would be two studs and the switch side would be one. i scratched this together quickly in my lunch break so its not at all polished, but it shows the basic idea of what would be going on in both positions. i feel like this setup would be stronger and less flexible than your current one. just giving ideas! :)

Yes, there is a reason. I actually tried that setup. The reason I went with the longer lift-arms is that they provide more leverage for the small pneumatic cylinder. It's a simple lever... Like a crow-bar. Small force on the long end yields greater force on the short end. So long as the cylinder is at least 5L away from the pivot or fulcrum there will be enough leverage for the small cylinder to move a stiff switch. 

I'm working on some more photos, but I have this landscaped a bit better now. The basic construction is the same, I have just moved some of the parts around to make it look better and hide the air hoses. Hopefully I'll have some photos later this week.

[powerwindows83]

that makes sense. the real question is why didnt they make the slider a normal lego connection of some sort... blown opportunity imo :P