Harlequin

What era are you going for? In the pre-1930 's trackplan of Plymouth the middle line was simply a very long trailing crossover. I only mention that because it's simpler than your plan, which might be beneficial and would eliminate some facing pointwork. I think the double track at the left should feed to the two middle platforms lines, as David has sketched. (It looks a bit odd that the outer track only feeds the top, outer platform.) What about curving the whole thing gently towards the viewer and the operating well? That should fit better in the room, ease some of the corner curves and give you curved platforms, which always look good.

Hi Neal, 94XX were first built in 1947 so a bit out of period for your 1930s timeframe...? Slightly closer than a Rocket, but still... If you're interested in sound it is possible to fit quite a big speaker in the Parcels Railcar without it showing obviously through the windows and without having to do any cutting or carving - just two soldered connections. (Dapol have modelled some basic shelving/racking inside the body.) Having said that, I'm having some sound quality issues with mine, which seems to be related to the decoder heating up. I'm working on it.

@The Johnster

Can't tell what's weathering and what's been "photoshopped"...

It's an awkward space. The 20ft long bit isn't really wide enough to do anything scenic (only maybe single and double track running along it) and the 4ft wide end boards aren't long enough for stations (probably). Some triangular fillets joining the end board to the long board might help. Edit: Something like this, perhaps: I'm not sure it really works because the roof braces still get in the way... Could you post a photo of the loft hatch side of the space so we can understand it better?

I see... Nothing in life is easy, is it? One day you'll be able to hand this sort of problem off to an AI - let the technology sort itself out!

Hi Andy, You're in a really difficult position, I realise... It sounds like the trekkies are trying to make the sevice work within limited resources but isn't the problem, in fact, the limits themselves? Would more memory, faster server, more bandwidth, etc. help? I know that would cost a bit but imagine the hassle-free, sunny world you would live in!

Ideally: Plan, Design baseboards to fit plan, Build baseboards, Buy track. There is Mac-specific track planning software. It popped up on here recently but I can't remember where. If you can't find it or don't want to use it, the next alternatives are: online track planning, installing a Windows VM and running one of the Windows programs, using a drawing program, using pencil and paper, planning it full size on the floor using paper templates. I use a drawng program (on Windows). It does take a bit lot more work than dedicated software but the end results are better, IMHO.

This: You could probably smooth out that crossover by starting it in the curve like this: If this were a OO design and you mentioned "1st radius curves" alarm bells would be ringing now! I don't know if the same is true in N but it still sounds like a bad idea to use the tightest set radius parts. I have a couple of thoughts: Since you've got a baseboard why not just buy some settrack and plug it together to try out different (flat!) combinations. That would be great fun and a great way to start to get an idea of what will and won't work. I think you are making things difficult for yourself: 1220*610 is almost the minimum viable size for an N gauge layout and you're trying to cram a lot into it. Judging from the photo above you could do something slightly bigger (say, 1525*762?) and that would open up the design possibilities hugely. (But maybe still not track rising on a gradient to cross other track - I think that's not really sensible in such a small space. Much easier to grade the landscape and keep the track level!)

Using track designing apps, or any other suitable computer program,is a perfectly valid method of planning a layout if you are comfortable doing it that way.

Sorry to open old wounds but the forum has been sporadically slow again for the past few days. I had Cloudflare kick in a few times and serve cached pages. I didn't report before through laziness and not wanting to be a constant whinger but it gets a bit frustrating and I wish it could be fixed. Whenever I catch a slow page in the Chrome debugger it's always the same pattern: the first http request is slow and all the subsequent requests for graphics, style sheets and everything else arrive quite fast.

You will need to space the two lines further apart to give room for a retaining wall. Will that look realistic? A lot of the inner/lower track work will be obscured by the rising track and the high level terminus in the centre. P.S. The latest plan is almost end to end... You have to take a sharp crossover, a deviation, to complete a circuit on the lower level.

Nicholas Parsons has joined the team of erudite and talented former players of the great game in the sky. At this very moment he is probably giving one of them the benefit of the doubt...

Near the bottom of this page http://www.cornwallrailwaysociety.org.uk/kingsbridge-branch.html# is a photo captioned ‘down train on the approach to Sorley tunnel’ which matches 2234 very well. In fact, 99.9%!

3D Printing? That moves the problem of rope weaving into the computer - but I'm not sure it would be any easier there! Maybe someone has already done the hard work?

I'm not seeing that specific problem but the site has been generaly slow for me for the past few days and posting a reply often results in the "Saving..." message being on screen for a long time. In that case I do something else until the post appears in the thread. Edit: I think it's the same general slow down that has previously been reported in the "grindingly slow" thread.

The three seperate baseboards mean that the tracks will require many more power feeds than those, I'm afraid. Almost as many as a best-practice DCC setup!

Hi Marcus, Best not to have turnouts butting up against baseboard joints for three reasons: It’s a bit more difficult to align track across the join because you can’t use the fix and cut method, you can’t get such solid fixing at the edge ( such as copper clad or brass screws) without modifying the turnout and if the rail gets knocked or damaged you might have to replace the whole turnout rather than a simple bit of track. Having said that some people lay a turnout right across a joint and then cut it in half! Could the two turnouts above the refuel point be replaced by a double slip? That would give the stabling area a direct connection to the main line. What radius is the curve? It looks tight. Do you need to get wagons in and out of the refuel point and engine shed areas? If so is the headshunt bottom left long enough?

Homer also said, "In case you didn't notice, I was being sarcastic!"

Here's a further revised version which I hope, after reading SRS Paper No.8, is closer to the 1920s style with later ammendments. The signal and ground disc symbols have been changed to be more correct for the period. Colours revised to match the period colours better (but there's still some work to do on that). Lots of detail changes.

Yes, that makes sense - and by that logic you could say that a banner, being essentially a "black armed signal", would have a black 'ball'. That feels right - but what feels right and what the Western actually did are two different things.

Not much difference between masking the flashing and doing the green or doing the green and repainting the flashing! You have to build these effects up in stages... And talking of effects in stages I have one word for you: Lichen

Wonderful model of Kingswear, including unloading of coal from ship into wagons here: http://www.gwr.org.uk/layoutskingswear1.html

Have you seen this topic?: @M.I.B has been piecing the evidence together.

Atkins suggests that before 1914 coal was sent by rail from South Wales to Kingswear because Dartmouth/ Kingswear was the principal South Devon coaling port for merchant shipping. After that date the story might be different.