tom s

That is a neat setup! I didn't know about worm transmission losses being high, which is unfortunate with the space in a 2mm model. If it doesn't work in practice then I will have a look at bevel gears.

https://www.aliexpress.com/item/1005006462873434.html?spm=a2g0o.order_list.order_list_main.4.32cb1802ZtNRW1#nav-specification Unsure! Just the cheap as chips one. It just feels like it lacks the push to get the motor moving. I did a quick recording of the motor and on the 3v pwm its cut in speed is about 1200rpm at the motor shaft (32000rpm max according to the specs), whereas the 12v combi controller can do it as low as 65rpm. Still, even if it is a magical model with no transmission losses, 1200rpm through a 1:11 is still a scale 2.7mph cut in speed.

Unconventional as it is, currently 3d printed gears, bogies and the propshaft parts seem to be working well enough in mock ups. To get a realistic top speed would need 10-12 tooth gears however, and I've yet to try if that works.

The 30Ohm resistor doesn't have much of an effect, but after I ordered it I read into the specs of the motor and the rated free turning amps are just 25mA, raising to 100-200 at stall (conflicts between sheets.) So that's not even 1 volt it is taking down. I suppose I won't know the ballpark value to use until I can give it a test in the full chassis with some representative drivetrain resistance and some load. Also picked up a pwm speed controller and it seems to work well, but the lowest motor speeds aren't as slow as the gaugemaster can do. Again, test chassis needed to see if that is still true when it matters I suppose!

After noticing how superglue retains the reflective finish of microscope glass on a failed project, I wonder if there is a potential to use them to create a glassy finish on clear prints. Sandwich a dot of resin between the print and a slide, cure it, then pop the glass slide off and clean the edges?

Vallejo Model Air Silver is another good real aluminium powder paint with good coverage and convincing shine to camouflage plastic rails among metal ones, or act as a base to add black or gold tones to, depending on your rail of choice.

I think I found why my chairs were not printing correctly. Doing a full cleanout I noticed a distinct residue impression of the trackwork on the FEP, and I am assuming that as it was clouding the surface, the exposure for the smaller details might be getting distorted and fuzzy. The impression took a microfibre cloth to remove, and seems to be appearing after every print, so multiple prints while simply stirring the vat of resin rather than filtering and cleaning seems to have bitten me. The resin is two years old now as it takes quite a while to get through 1kg of the stuff when working in 2mm scale, so perhaps it is the age? Either way, complacency!

Indeed, I have found a spare wheel needs to be used to feel out the chair collisions and a needle file needs running along the inside of the rail to knock them down a fraction.

Only issue I have had with the flat on the plate approach is chairs in the centre of the FEP not forming correctly, and the frequent irritation of having the whole piece scrapped when they snap off when trying to thread rail though, even when the outside chairs accepted rail correctly. Admittedly this was in the colder part of the year and it was so annoying I shelved the project for a bit, I am meaning to do a proper clean and relevelling, and try again now temps are back up in the 20s.

While adding the webbing I bevel the sleepers in blender to combat the elephant footing and give the spatula something to slip under.

Thankyou both! I will pick up some 3w resistors and a pwm controller to try both out. I would think that the 12v would be more reliable at traversing the track-wheel-pickup connections, but having the layout on a 3v rechargeable battery rather than the mains seems appealing too if there isn't much difference. Cheers, Tom.

I have been experimenting with some cheap tiny coreless motors for use in a 2mm scale scratchbuild, and so far it seems to be working decently with a gaugemaster combi controller and attached to a short gear train. They can crawl with the inbuilt 1:5 gear reduction and feel like they will be sufficient to shuffle around 3 coaches on an end-to-end layout. However, I would like to set up the locomotive to drop the 12v output down to the 3v that the motor uses. Is there an accepted way to do this? I have so far looked at a voltage divider but don't know if heat in the resistors will be an issue if I get the wrong values, even if the ratio of R1 to R2 is correct. Cheers! Tom

The bevel weight idea.

I have only used Blender so am unsure how other software use non destructive modifiers, but I would either use a Bevel modifier with the the edge's bevel weight value being changed from max at the top to the minimum at the corners. Or on the right making a simple version with a handful of vertices and use the subdivision surface modifier to fill in the overall curved shape, manipulating the vertices to get the shape to match the profile of the roof arcs behind, while using a supporting loop at the top and bottom of the bevel.

Hi Martin, that is correct, apologies for the potential confusion! The sleepers themselves are 95% the Templot .stl file sleepers in the default solid mode, the chairs scratch made and positioned using the .stl rail centrelines as a curve modifier, and the webbing was scratch made on a guess for what would be necessary. Cheers, Tom