PaulaDoesTrains

I think many people modify their 3D printers because they can rather than because they need to. Certainly for printing PLA just about every current printer will give decent results unmodified. The only significant upgrade I've done to my printer is to upgrade the bed levelling springs. I'd had the printer quite a while beforehand and I only did it so that I didn't need to re-level the bed so often. I've recently had to change the boden tube but I knew I'd have to do that eventually as it's a well-known design issue and I had some Capricorn tubing on hand ready for that eventuality. I got close to a year from the original boden tube so I'm not grumbling.

I've got and Ender 3 Pro. Is the best printer? Probably not. Is it a bad printer? No. Is is good value for the money? I think so. But in the same price bracket there are others such as the Voxelabs Aquilla with more up-to-date features.

I think a lot depends on the amount of stress the mountings will experience. If the stresses are very low or entirely compressive then resin might be strong enough. If you do decide to go for a filament printer then PLA should be OK. I've watched quite a lot of YouTube videos where various filaments have been tested for strength (CNC Kitchen is a good channel for that) and none of the ones you could use on a home printer seem particularly stronger than PLA overall. PLA is also the cheapest, is readily available and is the easiest to print with.

On my little BLT I started off changing the points by hand which worked OK but a) I'm very clumsy and any stock anywhere near the points was in mortal of being knocked over and b) I was forever forgetting to change both points on a crossover. Then I discovered the method you refer to in a YouTube video by Casterbridge OB. It works really well.

4GB should be enough to run FreeCAD on Linux. I sometimes run FreeCAD on my 4GB Chromebook using crouton and XFCE. It's a bit slower than on my desktop but that's because of the Pentium-class processor not the RAM.

There's lots of good tutorials on YouTube for FreeCAD. My personal favourite are Mangojelly and Joko Engineeringhelp. The latter covers a number of CAD applications (Solidworks, Fusion 360 and few others) in addition to FreeCAD. He does CAD work for a living so, IMHO, his views on the relative strengths and weakness of the various CAD programs are worth considering. Perhaps check out some of his other videos to see just what FreeCAD is capable of? Any modern CAD program will do what you want and you won't "outgrow" any of them.

If you truly believe the problem is caused by bugs in Blender then I'm sure the Blender devs would be happy to look at it.

No. Current hobbiest printers are either SLA, which uses a vat of resin, or FDM, which uses a spool of filament. Each technology has its pros and cons but for fine detail SLA wins hands down.

I must confess that I didn't go any further than the import so I've gone back, redone the import then in the Part Workbench did "Create shape from mesh" which ran for a good half hour and was only about 25% of the way through before I decided to kill it as unfortunately I just don't don't have the time ATM to pursue the process end-to-end. But the process should be "Create shape from mesh" followed by "Convert to solid".

How are they importing? I've just successfully imported the albion body using the Mesh Workbench->Import Mesh From File.

I used a 1mm end mill. It does leave a radius in any internal corners which to my eyes is barely noticeable. A needle file could be used to remove the radius and leave a sharp corner if so desired. I have experimented with a vinyl cutter (about £10 on Ebay) on paper with some success but to cut anything of any substance would require a better mounting arrangement than my 3d-printed one which has a bit of side flex in it. Perhaps I'll revisit it at some point. I think the plus point for these little 3018s is that you can use end mills, vinyl cutters, lasers or even attach a pen and use it for drawing. The minus point is that it's all a bit DIY. You have to spend a lot of time ensuring everything is square and the bed is level. Hope this helps.

I've successfully cut window frames from 20 thou plasticard on my 3018.

I'm in the same situation with my Triang 31 (a cheap Ebay non-runner). I've got the replacement bearings but haven't got round to attempting to replacing them yet so I'm watching this thread with interest. And learning a lot

Mia Culpa

I think you're a tiny bit out of date. PECO have been making track, including Setrack, at their factory in Beer, Somerset for a long time. They also have a second site at Buddly Salterton where they manufacture their Wills, Ratio and Parkside products. They make all their injection moulds in-house (something Hornby were incapable of doing) using two vertical CNCs - a Hurco and a Roeders - and an EDM machine. They're now starting to use robot automation to help with the assembly of track. All of this is well documented in various videos on YouTube.

@JohnGiI agree with what you say about cutting plastics - single flute cutters are the way to go there combined with a high feed rate.

One of my little projects which I completed quite a while ago now was for my circa 2001 Bachmann Class 25 which is a nice model but came with some horrid little tension lock couplings which don't play nice with the large D couplings on my stock. And no NEM pockets so I couldn't just swap them out. So what I've done is machine some adaptors out of 3mm HDPE which allow me to attach a Hornby D coupling. The project was not without its challenges due to the requirement of machining features on both sides. If people are interested I'll do a post about that that but for now here are some pictures of one of the adaptors prior to a bit of cleaning up and on the loco complete with D coupling.

All these different units can be so confusing In all the stuff I watch online they universally seem to use distance per minute for feed rates so I've simply fallen into the same habit. I think we're both using similar feed rates. I suspect that the limiting factor with these small diameter cutters is their fragility rather than the machine's rigidity.

Thanks for the info @Giles. I've been experimenting with machining brass on my 3018 and have so far broken four 1mm 2-flute end mills. Ouch! I found a video on YouTube of someone milling brass on a 3018 with a cheap 1mm "corn on the cob" end mill and he was using a feed rate of 25mm per minute and a 0.2mm step-down. So I broke out my packet of corn end mills (one of those "they'll come in handy some day" purchases which may have actually been worthwhile) and started experimenting. Whilst my machine will cut with those parameters, I was seeing rigidity/backlash/tool deflection issues with my 4mm diameter test holes which came out anything but round. At 0.1mm step-down I'm getting decent results though so I think I'm going to be able to machine fairly simple parts although nothing approaching the quality and detail that you're producing. As an aside, I was watching a YouTube video about PECO's CNC machines which they use to make all of their tooling and injection moulds. They routinely use cutters as small as 0.3mm diameter and sometimes 0.2mm!

That's very impressive @Giles. Can I be cheeky and ask what size end mills and what feeds and speeds you're using?

My copy arrived from Strathwood on Saturday (ordered on the 21st). Well worth the wait

OK, for the sake of argument, let's assume that you made the basic cylinder by drawing a circle on the XY plane then padding it. What you do then is create another sketch on the same plane representing one of the splines.Note that the sketch will have to a closed shape. You then pocket that sketch to create your 1st spline. Now make sure that the pocket is selected in the tree view and press the Polar Pattern icon. Fill in the number of splines and you're done.

Use a Polar Pattern. Design the feature in Sketcher, then Pad/Pocket it in Part Design then use Polar Pattern also in Part design.

I think you're going to have a challenge on your hands getting all that into 1100mm. My own little layout (PECO Setrack) has a run-around loop like yours and the distance across the outer 2 points is 840mm. You then need to add at least 200mm to the right for the loco release plus a bit for scenics/buffers. That's basically used up 1100mm. You can claw back some space by shortening the run-round loop. At the extreme you could get 170mm by having the points touching. But you're probably going to need about that length to the left of the loop for the canal and scenics. An alternative might be to forget the run-round loop and have the 08 pull the stock away to release the 26?

From what I've heard the retailers are getting deliveries of PECO track but they have lots of pre-orders so it effectively becomes instantly out of stock. So ordering now, well before you need it, seems a sensible idea.