coaster

Very good. Anyone not in favor of moving the switch plate?

It was pointed out to me that we have a teensy little issue when constructing parallel sidings: Solution: What is everyone's thoughts on switching the throw mechanism to the straight side? Unless you have your tracks right up against the edge of your display, I can't think of a scenario where this would cause an issue. I suppose it'd be a problem if you were building a circle with straights coming out every 22.5°, but that'd be really weird. Certainly less common that building a yard like this. And, for what it's worth, if/when I do a double crossover, it'll have to have the toggle on the straight sections as well, otherwise you'll be forced into dependent directions like the official Lego one.

Correct, the switches would come with the appropriate curves. Those curves will also have a symbol on the bottom to differentiate them. Sure, I'm guessing the new ones are a little different from the old ones? I don't have anything newer than probably 10-12 years, so I'm probably referencing dated parts. I haven't actually ordered the printed bases yet, so I'll check on the guards first and get those corrected. I've ordered some used ones off ebay. Should only take me a few minutes to update that once I get them.

Thanks! Sadly, since I'm not Lego making tens of thousands of these, I can't make and sell a pair of switches for $30. It's probably going to be around $75-100 per switch, so figure to get a switch pack with two switches and 4 curve pieces (parallel and concentric tracks require 2 different curves, unfortunately) will probably be around $250. It's a little higher than I'd like, but if you compare to some of the LGB stuff out there it's not too far out of line.

I'll sell them completed. The metal rails can't be printed; they're too small to do accurately. Plus, there's a spring beneath the point and some metal stamped pieces needed. I'll source everything, finish, and assemble, so they'll be complete. I've also had some decent quotes come in on tooling, so molding the bases may not be outside the realm of possibility. Still too early to call though.

Took a little longer to finish than I wanted, but it's off to the printers! EDIT: If anyone wants to take a closer look, check this out.

I would certainly be interested in subscribing, and once I get my tracks off the ground, advertising as well. I'd prefer print copies though, rather than strictly digital. And I'll echo some of the sentiments above; "how-to" is far more valuable and interesting than just pictures of MOCs. Show the MOCs, by all means, but tell me how to build that as well.

$134 and change, and that's also with it missing several pieces. What I was hoping to do was leave all the components as a single item, this way someone could just order a "R104 right switch assembly" from Shapeways and get everything they needed, minus a spring. And that spring then becomes an issue because I need to source one that is readily available to the public, so I can't customize one, or even specify a stock one from a manufacturer (most don't sell to non-industrial customers because they want you to buy 1000+) . The alternative becomes everything runs through me, and I'll finish and assemble. That was going to be the case for the 9V ones obviously, but was thinking I could defer the all plastic ones to Shapeways. I certainly don't mind doing it, but then I have to handle not only the parts, but orders, packaging, shipping, etc, so that just adds cost.

Nope, just fails the bounding box. I modeled up a dummy bounding box, and with a little massaging, I can get it to fit, so Shapeway's software just must not be able to fully orient it to get it to work. I'll try uploading a model where I reorient it on the xyz, see if I can force it to fit. If not, I suppose I could always just call them. EDIT: Well, I was able to cram it in, so the Metallic Plastic is an option, though it's almost twice as much as the regular white or black (which also opened up, by the way).

Hmm, seems I need to brush up on my earlier era trains. I didn't jump into trains till they came out with 9V, so I'm not too familiar with what came before. But I'll look into it. As for the print color, unfortunately Shapeways is going to limit us to just the white. Whatever machine they have that's big enough to print the switches only has the one color option. None of my printers can handle it either. Now, if I had $250k to burn, I'd get myself one of these bad boys and then we'd really be rockin'!

The spring is part of it, but the raw finish of the printed parts isn't that good, so there's quite a bit more friction than would be on a molded component. The printed parts are also not as strong, so I've had to thicken them up and then there's clearance issues... Nothing insurmountable, but will probably take more work than I want to put into it right now. Ah, that should be doable. I don't have any 12V stuff, but I could pick up a few pieces. You're referring to one of these, right? I don't see a reason why we couldn't even do that powered. The rails pieces I'm making for the 9V tracks could easily be adapted here. Might have to get a little creative at the ends to connect the power, but I can put it on the list to look at.

Sure thing. I've been tinkering with the new mechanism, but it still isn't working to my satisfaction. I'm not giving up on it, but I'm going to table it for now in favor of the standard Lego mechanism. However, having only one moving point does introduce a handful of other problems. I have most of those worked out, but currently looks like this (less said moving point): As for open sourcing these later, I'm not sure how much good that will do you. STL files are sort of a one way street; they're really not usable from a CAD standpoint, even doing something as simple as removing a stud would be an arduous task. And I wasn't familiar with Openscad, but looking at it, I think you would want to kill yourself if you had to work with that. Plus, there's no "code" per se for you to edit, so not sure you could do anything with it anyway. Even if I were to release the step files for these (universal format that is editable), editing them to create a custom piece would be an absolute nightmare for someone who does this everyday, and likely insurmountable for the casual CAD user. Just to give you a sense of the complexity to this, here's a screenshot as I work on the switch base: I have a set of rules that I've also programmed into the model that aid in regenerating the curves and crossties if I change something. It's modeled parametrically as best I can, and all the dimensions are actually referenced off a master spreadsheet. I'm doing this as efficiently as I can, and there are still some 1200-odd parameters and 150+ work features, none of which would be available to you from the step file (think of it like a weird, insanely huge zip file). Your CAD software would have to have direct edit capabilities, which, as far as I know, is reserved for the higher end software like Solidworks or Inventor, and it would still make you want to kill yourself. Not trying to be too discouraging, just don't think it's something you're really going to want to take on.

I still have concerns about the metal rail solution. The 9V motors pick up the power from the wheel flange, not the rolling surface. The metal rail doesn't overhang the plastic one at all nor present a side surface for contact, so I'm concerned that conductivity may be intermittent. This is why for the curves and switches I'm working on I'm using a custom profile that matches the effective Lego geometry, like this: I also still don't understand why you are doing the rails and sleepers as separate pieces rather than single rail segments, especially now that you've departed from the all-metal rails and generated a brand new tool.

That is interesting. What did it look like? The one I ripped open is probably 25 years old, so wouldn't surprise me if the design's changed a bit since.

No problem, man. Looks like this: There's a tab on the throw arm that fits between the legs of the spring. That moves the point while still allowing it to "flick".

Either Nylon or Polyethylene seem most likely. And Steve is right, you want Ø1/8". 3.2mm is the plate thickness for Lego (and subsequently all rods), so 1/8" tubing is almost a perfect match. I'd suggest starting with this.

Ah, McMaster is my engineering bible. Unfortunately, their supply of springs is extremely limited. Also, those are compression springs you linked to, we need torsion springs, and those are a really narrow offering. The closest one they have is this. I may be able to make it work, but won't know till I try it. But I will.

Thanks everyone. There seems some interest in the turn-table mechanism I have shown, so I'll give a little more detail about that. That section is one brick tall, studs on top, but bottom is flat (only so much I can do and there's no room for a plate bottom). The axle hole go through the part though, so if you wanted to hide your motor underneath the table or layout you could do so. Range of motion is quarter-turn lock-to-lock, and it's spring loaded such that it simply toggles between the two positions. There's no weird in-between area. But it would be simpler to simply stick with the standard Lego design mechanism. I'll make both versions, and we'll see which is better. I was originally going to stick to strictly 9V tracks, but I don't hate the idea of modeling all-plastic track and posting it on Shapeways as an additional means to fund the 9V project. There are two issues though with just putting these on Shapeways. First, for the size of these, they can only print them in white. The gray is only available for smaller parts. Second, and more importantly, even the PF versions of switches still require a spring for the points. I might be able to get creative and add some tabs to the bottom of the points that could act as a spring, but no idea how well that would work or if it would hold up to repeated use. Of course, you could always pirate the one from the Lego switches if you don't mind sacrificing them, at which point you could steal the throw-arm as well. As long as these things aren't objectionable, I'll copy/modify the designs for PF and upload them to Shapeways.

Happy Valentine's Day, everyone! To show my love for you all, here's a progress update! I've adopted Steve Barile's idea and been running with that. However, I'm doing it a bit differently and having a special profile rail made that mimics the overall exposed area of a standard Lego track. Here's a test piece I did: As you may have noticed, this also give me the opportunity to wire the tracks without having to use the Lego connector, making for a more realistic looking layout (that connector still works with this, by the way, just showing an alternative). With a bogie on it, it looks satisfactory. Curves are a piece of cake though, so figured I'd share some progress on something a little more interesting: Finished, it should look like this: I was going to wait till I was finished with it before posting, but my printer is down at the moment and will take me a couple weeks to get back up and running (low on material + broken tray; it's a crappy little printer). Once I am though, I may make some additional modifications. Originally, I wanted to have 2 moving points like a traditional switch does, but that required a new switch mechanism. I wasn't opposed to this as I felt I could make it a little better. After playing with this base a bit, it would be simpler to go back to the single point Lego version. Simplifies the assembly quite a bit. So I'll likely make a new base reflecting that change. Otherwise, contacts are all good and solid, power across the frog is fine, either way, should be good to go when finished. Now, let's get down to brass tacks. Investing in a mold to make standard angle curves isn't a big deal, should be able to make enough pieces to justify it. Something like this switch though, probably not. However, that doesn't mean we don't have some options. The rail cross-sectional profile is the same across the board, and I will be having 2 tons (literally) of that made. So, what we could do is instead simply 3D print, sand, paint, and assemble the switches. They'll look and feel a little different from the molded ones, but should be close enough no one will notice. The downside to this is, you won't be getting a switches set (left and right + curves) for $30. Most likely, a switch like the above is going to have to be somewhere around $80-100 by itself. That's just the nature of this beast. This approach is also a bit interesting because it allows us to made some of the more exotic track pieces without having to worry about making 70,000 of them. Double outside slips, double crossovers, Wyes, 22.5° crosses, 45° crosses, odd-length straights, you name it. If I can model it, I can build it.

While I understand the hollow stud is more money, there's a good reason Lego did it. It cost them more money as well, so they wouldn't have done it if it weren't functionally necessary. Without it, you're going to have some serious sink marks in the stud, and not just the face, the sides as well. The shrinkage is not going to be even either, so you can't just say, "well, we'll mold it at 5mm and let it shrink to size." It's not a controlled deformation, and I suspect your mold maker doesn't realize the stud diameter needs to be held to ±.01mm. Even aside from that, you'll still have a parting line along horizontally across the stud, which will also mess with the fit size. If you're serious about this, I'd strongly suggest you reconsider the side-action for the tooling. If nothing else, ask him. Unless he's also an attorney, conversations are free.

I'm fine about the Kickstarter bit to get going. I'll likely do one myself for the different 9V track geometries once I have everything buttoned up. It takes capital to start a business. and I agree, that is the purpose of Kickstarter. No one expects you to be a charitable organization making monorail track out of goodwill. I do have some concerns about the manufacturing though. First, it sounds like you have a good contact for doing the molds, but you're going to have to have side action for the studs, no? I noticed in your picture you don't have the small holes in the studs, which concerns me for 1) shrinkage, and 2) that you're going to clamshell along the stud, thus having a parting line there. Stud tolerance is very tight, and I find it unlikely you'll be able to hold it properly here. Adding side-action to a mold is not a trivial cost either, and I'm not sure you've taken it into account. Additionally, maybe you guys have a great digitizer, but the ones I've used (albeit very few) have had crap for resolution/accuracy. Best I could get was around ±2mm, which for the studs and toothed rail is horrendous. I'd reconsider simply dimensioning the part and modeling it up separately, then use the model to fab the molds. The part's not that complicated, it shouldn't be difficult to pull dimensions off it. Good Luck!

I've never used the 12V tracks, but we have a similar problem with the 9V tracks. What's happening is you get a voltage drop across the rails, so on that far side, you're not really seeing 12V, probably more like 3 or 4V if that. It's not a cleanliness issue; it's simply the nature of the rail material. We just add additional transformers on the far side, and that solve the problem.

I'm glad it's not just me. My current backlog:

I've talked to Steve separately on this, and I'm going to take this approach as well, but keep it strictly to a single complete piece (no assembly required on your part). This is a really clean solution, and if designed right, should consist of a single piece base into which the rails snap into place, so the minimum 3 pieces. I've started sketching it out, and it should lend itself to nice clean switch design too. Steve doesn't seem to mind who does these, and his fee is more than reasonable. It also doesn't require a huge change to what I've been working on so far, so I think I could have prototypes right after the new year.

I have considered it, but given the amount that I'd need to build these, the cost would be insane for me to just buy stock material. For example, Walthers in that link you provided offers 8x 6' sections for $132. The equivalent amount from the mill will cost me about $40 whether I use the same Code 250 profile or one specially for the Lego profile, at which point, I may as well make exactly what I need. Of course, I need to buy about 1500 lbs of the stuff, but for production purposes, that actually isn't that much. That said, if I do run into alignment problems with the next prototype curves, using and modifying stock rail joiners may be a necessity. I'm really hoping to avoid that though.