Sunday, 21 January 2018

It's the "I don't have a name yet" railway

I have been modelling US outline HO for many years with only a couple of unsuccessful forays into British OO and OO9. However, I have got a bit depressed about the costs of buying stuff from the USA. Finally, I have decided a) that I want to do something where I can pop to my local model shop,that isn't going to cost an arm and a leg for postage/added VAT and handling charges and c) gives me more room for a railway.

Hence, I have taken down the HO layout. I am slowly putting all of the equipment up on EBay and am now building an Gauge layout. My basic area is one leg of 9'x 2' and a second leg of 6' 6" x 2'. In HO/OO this is quite tight but in N it is acres!

The problem is that I know little or nothing aboutBritish Railways having been brought up in South London (Streatham to be exact) and thus was only exposed to Southern Electric in my youth. I have decided that I will use basically LMS equipment as I can cope with 3F, 4F, etc. andI like Duchesses and Jubilees.

I went off down to my local model railway show - Scograil in Ipswich, which seems to have morphed into Orwell Model Railways and now run by Brett (of Scograil) and Kevin(of Coastal DCC fame!). I must say that we are very lucky to have both a top quality model shop AND a genius at DCC - in Kevin Dickerson - within 5 minutes drive from me. It seems that this results in daily visits!

Where am I then. I have designed the layout - basically a medium sized terminus feeding a small branch end of line and a fiddle yard. Location seems to be in the north west. I am thinking of making it a seaside resort where you can see the Lake District in the background feeding out to a small town with a Creamery as its biggest employer (Seascale way).

The terminus looks like this:



and the branch like this


Telling my wife that this would be cheaper than US HO, I went round to Orwell and spent £350.00 on some track, a loco (Ivatt 2MT), some passenger coaches and goods wagons. I then acquired a 4F. Having tried DCC Concepts DCC decoders (I am a Digitrax man) I swapped them out for some Zimo as Zimo seems to manage coreless motors better (thanks Kevin). This was when I was thinking of building just a small board for fun. Having turned it into something bigger and better(?) I then spent another £350.00 on the track for the terminus. Given that I am expecting to run a Jubilee or a Duchess into here with some "blood and custard" Mk1s and some extended coal trains, I haven't yet explained that I am not finished! Fortunately, I have all of my HO stuff to sell so that should offset the costs somewhat.

I am quite disabled with arthritis and can't stand up for too long so my model railway sits on top of some B&Q kitchen units that sit on the floor. This means that I can roll around on an office chair and do most things sitting down. I can't contemplate climbing underneath baseboards so I build everything out of 5mm foam core, extensively braced. This makes a 9' x 1' 4" baseboard light enough to pick up but strong enough when sitting on a stretch of kitchen worktop.

I use DCC extensively and thus have Cobalt IP motors and I drive them from JMRI buy setting up routes for specific trains. These sit underneath the layout but if there is any issue, I can just flip the board on its side and fix it. To make the shunting on the branch more interesting, I have reverted to wire and tube for the point operation but the fiddle yard (which will be behind the branch eventually), will be wired into the DCC bus and JMRI.

I have a good friend who is an artist (see his website at Brunswick Fine Art). He has painted some back scenes for me - and I built his web site for him. I am in the process of fitting these. As the layout may need to be lifted, the back scenes are all removable.

Anyway, that is the story so far. Here is where I am with the layout itself.

The main board.


and the branch terminus


So far, I have been unable to think of a name for this. I am thinking of something like"Grange-over-Sands" or "Wells-Next-The-Sea" - in other words, something that tells you it is a seaside location. So far, I have come up with nothing!

Friday, 12 January 2018

So, the world turns - AGAIN!

I was feeling a bit off about US outline modelling over the weekend. The cost of buying anything that isn't a box car, a passenger car or a loco is out of all proportion because of the costs for shipping, added VAT and the Royal Mail/Parcelforce £8.50 handling charge. My mind started to wander and I found myself wondering if I could knock together a small N Gauge British outline layout. By small, I meant around 3' 6"" X 12".

I didn't get far as I started to work out what was needed and suddenly the word "budget" loomed large. However, the advantages are many as everything is available locally here in Ipswich, or at least 90% of it as we have a good model shop here - was Scograil but more of that later. I gave it a bit more thought and then started to be a bit creative on the money side. Without going into details, I felt that I could justify the expense of a few items. This got a bit out of hand, as these things do.

I started playing around with the idea and, rapidly, found that 3' 6" was a bit tight so expanded it to 4' and then out to 4' 3". I have now designed quite a nice little terminus running into a "fiddle yard" - the UK name for "staging".


This has a single passenger platform, a goods line, three industries, a goods shed and an engine shed so quite a lot going on. I thought that I could run this on two locos and a few wagons. Well, that didn't last.

I went off to the local shop and found that all had changed over the new year. My friend Kevin of Coastal DCC had joined forces with Brett from Scograil to buy out (as I understand it - if I am wrong, then sorry) Neil Scoggins and run the UK model side under the name "Orwell Model Railways" (OMR)- nothing there yet on the Internet but there will be. Meanwhile Neil is taking what used to be ScogBahn and now has a dedicated European model railway shop, on the same premises.

Anyway, back to the story. I bought some fresh 5mm foam core from Hobbycraft (4 A1 sheets for £10 - bargain). I then hit OMR for the real stuff. Well, £355 later - 7 points, 4 yards of track, one loco (Graham Farish Ivatt 2-6-0), one passenger coach, three wagons, a brake van and a decoder for the loco. Oh, and I ordered a 3F shunter for the town pilot. This tiny stuff ain't cheap!

I then found a cheap, new, LMS 4F which will arrive by Saturday. I took it all home and started building the base. Knocking foam core together is easy - all you need is a sharp knife and a hot glue gun. I have a special foam core cutter which keeps the blade at 90 degrees for a better fit. I then spent a couple of days laying some track and here we are:


I have installed all the wiring to make it DCC friendly. Last night, I put the decoder into the loco and - nothing happened. In fact quite a lot happened but not what I wanted. Firstly, I was able to change the address using JMRI but then, when I took it off the programming track and placed on the N Gauge layout (connected by jumpers to the main layout) it instantly shorted! I spoke to Kevin but we didn't get very far. I took it into him today and it turned out that one of the tender wheel axles was shorting.Not enough to stop the programming but enough when full power came on to short. He fixed it but when I got home, I still couldn't run the loco. It didn't short out but just sat there. After a lot of fiddling and speaking, again, to Kevin, it finally jerked into action. I have absolutely no idea what I did but it runs now.

I found that it runs very intermittently so I looked into a rolling road, but at £60.00 that was too rich for something that has short term uses. I decided that I could quickly put together a test track as I had two yards of track sitting there ready for the fiddle yard. 

Good old foam core. In 10 minutes, I had a 24"square baseboard built and a circle of track installed. The Ivatt is now running around that and I am hoping that the jerkiness will come out now that I have oiled it and it has been running for a few hours.



Tomorrow, I get to run some trains and then install some wire and tube point changers. I am also building a Metcalfe Engine Shed. See you then.



Friday, 15 December 2017

Rebuilding 470Router to suit me

As I have mentioned, I have been working on a software project to complete my Open University degree. This project was named 470Router - 470 after the code for the OU course - TM470 - and Router as it was intended to be used to create freight car routings for my model railroad. Well, I finished the degree (got a 2.2 and was awarded a B.Sc. Hons) and went off to do other things (mostly building some plastic scale model cars - Gentle Scale Modelling).

Now, I am back onto my model railroad I thought that I would review what I was doing for the software now that I wasn't under time pressure and the need to meet the OU module project requirements. It turned out that, although the software worked, it didn't do a very good job of building way freights. This is quite a complex procedure which has to take account of industry requirements in the towns, collecting empties, delivering loads, gathering everything back up and delivering empty cars back to their home roads. The software that I had developed didn't even look at that last element.

I spent a few days pondering over the state of the project and decided that the logic was way too complex. I had to think of a solution and it came to me - Car Cards! What, you say, back to bits of paper? Not quite. What I had in mind was to develop a set of virtual car cards. The basic problem of the existing software was that everything was run off the car and the industry. There was nothing that summarised the situation for any one car shipment. This is what a car card does perfectly. Therefore, I expanded the database to include a set of car cards. They look like this:


carType boxcar
town Sunset
target Creamery
State e
waitTime 2
destination home
id 1
timeOnSite 0
currentCarID PRR:81146

All of the above data is static (doesn't change) except for the currentCarID which records which car, if any, is using that car card currently and the timeOnSite which counts down the holding time at the industry.

Changing to Car Cards meant that I could implement some of the logic for the routing in a simpler way. I no longer had to deduce what was going on by looking at the industry data, the car data, etc. The whole picture could be deduced from the card. Most of the fields are obvious. However, a few need some explanation. The destination being designated as "home" means - refer to the actual car to see where it gets sent after this card is complete. For an NYC box car, this would be Hartford - as this is our nearest access point theNew York Central (although you could argue for Boston, given that the B&A was wholly owned). A Canadian Pacific car could be sent to either Hartford or Boston as both would give access to the CP system. This field can also be used to hold the ID of another card. This provides for a double trip - a reefer goes to the ice house and then, in this field, it gets redirected to a user of a loaded refrigerator car.  We know that the car is currently at the Creamery because there is an entry in the currentCarID. PRR:81146 is a Pennsylvania Railroad boxcar currently being loaded (we know this because the State field is "e" indicating that the car events are started with a delivery of this car, empty to the Creamery).

The timeOnSite field gets set to the wait time on arrival at the Creamery. There is a cycle of trains representing a single day. At the end of this cycle, the timeOnSite fields for each car is reduced by1 until it reaches 0, in which case the car is regarded as ready for collection.

On collection, the car is assessed for its next move. I, currently,  make no provision for multiple destinations. Loaded cars are collected by the next way freight and shipped in the direction of the train. Hence, loaded cars can go to Hartford or Boston depending on the route of the freight train. Empty cars are assessed against their "home" designation. All SNE (home road cars) are placed in the Sunset yard for further allocation. All other empty cars are shipped to their "home" as described above.

So far, this is meeting my needs. I accept that there is much more that I can do with this but, having spent a lot of time on this, I want to run some trains before hitting the keyboard again.

Friday, 13 October 2017

Building JMRI Routes

As I now have a system for car routing and train management, I need to move on and get JMRI working for control of the track. Ideally, I would like have JMRI controlling all of the points and then, as I have a defined set of trains, a route set up for each of those trains.
First step is to layout the track on JMRI and hook it up to the Digitrax system. I have a Digitrax Evolution wireless set up with a Digitrax PR3 connecting it to my iMac. As the run is a quite long, I have a powered USB cable connecting the PR3 to the iMac, which is on the other side of the room.

This video gives a good idea of the layout (sorry that it is a bit out of focus but you can see the essentials).




I have laid out the track in JMRI and allocated Digitrax IDs to all of the switches. The PanelPro panel looks like this:


(Suggestions on making this prettier would be appreciated).

As part of my 470 Router project, I have a defined set of routes to which trains are allocated. This is the 470 Router output for those routes (still work in progress so needs prettying up).


I went through each of these routes and built a JMRI route for each one. As it happens, NH1 (a Budd RDC that shuttles between Hartford and Boston) has the same route as BMP2 (which is a B&M passenger train that does the same trip) so I only had to define eight routes through.  There are 20 trains in the schedule for a single day and each of them has one or other of these routes.

Remembering that I can't stand up for too long so operate by wheeling an office chair around the railroad room, this saves me having to check that everything is set up for each train. When I am doing switching, it doesn't matter if the Boston bound freight is switching in the middle of the run or I am just preparing the next S&NE way freight. As soon as a train has to move around the layout, I can reset the route and know that I haven't missed anything.

Monday, 9 October 2017

Major reconstruction on the staging

Having finished with the routing software, I found that single line staging at both ends, although it seemed to be OK, wasn't going to work. I decided to make both staging to be double track. This involved the following:

  1. Purchasing 2 new switches
  2. Purchasing 2 extra switch motors
  3. Taking down the old track
  4. Replacing the top layer of the boards with wider panels
  5. Relaying the track.
  6. Fitting the new switches
  7. Fitting the new switch motors
  8. Wiring up the switch and the indicator LEDs
My chosen switch motors are DCC Concepts Cobalt IP which I find work perfectly for me - plus they have route/frog switching built in.

Boston Staging

OK, so the first step was to take down what is known as the Boston Staging. This runs across the front of my scale model display cabinets so this has to be removable. I have had some trouble with the stability of this and even more trouble with the rail join across the gap, so something better had to be done.

First off, the staging needed to be lengthened as it was always going to be more stable when connected to the shelving unit beyond rather than stopping someway short. The base for the staging is a box girder made from 5mm foam core. I have a special foam core cutter that cuts a V-shaped channel which enables me to fold the material up and make a box girder.



I lengthened the box, fitted some foam core to sit the free end on the shelving unit and covered the top with a wider board so that I could now get two tracks along it.



The next step was to lay the track. Now, there is a problem fitting a switch onto the box girder - the switch motor hangs down (of course) but I can't get into the box to fit it or to wire it up. I have faced this problem before when having a switch motor issue on the normal layout part - again the base is foam core and I can't lift the layout to get underneath. What I do is make the switch and the motor into a single unit and wire the motor up from the switch rails. This way, if needed, I can lift the switch out at any time to work on the underneath. All I need to do is to pull the rail joiners back and free the switch. Here, you can see it in the process of being created.


The hole in the girder is large enough to take the switch motor. However, there was a problem when first installed - the switch is an "electro frog" and thus has insulated rail joiners attached. You can't slide these back so I had to increase the size of the hole to enable the switch to be slid backwards to free them up. Remember that this is on staging so the gaping hole doesn't matter.

the next task was to fit the indicator LED. The DCC Concepts motor has a built in connection for managing the polarity of the frog. This connection can be taken off and used to drive LEDs. I like to fit bi-coloured LEDS - red and green. Normally, these would be run out to a display panel such as this one:

Lights-on is a bit overwhelming but you can see how it works.


As it is, I will just place the single LED at one end of the staging. I have problems walking for any length of time so I spend my railroad time wheeling about on an office chair so using these display panels avoids me having stand up to check the switch settings.

I made up a little circuit board for the LED as follows:


Here it all is, ready to go:


The last thing to do was to fit a clip so that the right hand end couldn't come detached from the main railroad board. I have used foam core clips for these when putting the railroad together but these would normally be temporary. This needed to be more robust to hold firmly and be able to be taken off and put back on. The only suitable material that I had handy was a sheet of PCB. Here is the way that that works (and you can see the LED indicator in place as well):


So, what does it all look like when complete?

This image is made up of about 6. The camera was fitted to the tripod and a stream of images was made with the focus moved along the stretch of track. I then used Helicon Focus to patch these images into one with a constant focus across the whole length of the board.


Once this was complete, I started out on the staging at the other end - Hartford, CT.  This staging runs along the back of my trusty iMac. The process was exactly the same, but went quicker of course. I made one mistake. That was to put the LED indicator at the same relative spot as with the Boston staging - wrong! This ended up behind the iMac. I had to extend the wires by about 36" and mount this under the girder, using masking tape. I then ran them out where I could see them.

This is the Hartford staging finished.



One last little detail. Again, as I can't get inside the box girder, the wiring for the two staging tracks is managed by using some self-adhesive copper tape across the top of the box. 

That's about it. Now, I can get back to sorting out industries, configuring up my car routing software for the new industry plan and then, hopefully, running some trains!





 

Sunday, 23 July 2017

Back to setting scenery

I have been very busy with my Open University project (470Routing) but I have managed to keep going on the home railroad. I have now ballasted the yard (with Woodland Scenics (WS) cinders) and the rest of the main town (Sunset) with WS N scale fine ballast. I have also set up the two-LED routing indicators. There hasn't been much running but the scenic work has been fun.

I have also managed to build two laser cut industries. One is the Branchline Trains Creamery whilst the other is the final part of the Ice House from GCLaser.

Using some WS lightweight Hydrocal I have put down some basic ground cover around the Creamery and the Ice House complex. It isn't finished yet but is starting to look quite good.

Here is the current situation:


(Click on the image for a full sized version)

Shortly, I will be completing the ground cover and the roads, along with the grade crossing.

Thursday, 8 June 2017

SNE Wired up and ready to go

I finally have an SNE that I like and that will probably work! I spent a lot of time drawing layouts and  posted quite a few iterations on the forums. The plan went through a lot of changes. Not only that but the boards have been built with extra bracing so they shouldn't suffer the dips that appeared in the last one.

The layout is at a much lower level so that I can get to everywhere whilst sitting in my chair.The switches are all controlled by DCCConcepts Cobalt Digital switch motors so all of the live frogs are wired up. I have, also, installed two-colour LEDs on every point so that I can see which way they are set without standing up.

The boards were built in sections but they are now held together with strong tape so that there is no movement between them. This tape will be covered by the scenery as it goes down.

Here are some images of the current state of play.






I have staging on both ends. This staging is build on foam core that has been turned into box girders. I have invested in a set of foam core tools and one of them is a v-cutter. This runs down a channel on the special FoamWerks straight edge and cuts a v-channel whereby the board can be folded with a 90 degree angle. This is extremely string and can span at least 5 feet without any flex. here is a shot of the end of one of the girders. as you can see, the v-cutter makes an excellent job of the cuts (such that you can hardly discern the "V" cut at all.


I have fitted a small level on the end and you can see the screw for the levelling foot, which is an M6 bolt with some foam core glued to it screwing into a T-nut. The other end is fixed to the base board. An added interest is that the desk is electric (in that it has a power lift and fall) which I use because of my arthritis - if I sit at it for any length of time my joints seize so I change the level on long sessions. This means that I then have to adjust the staging to get it back  to level as I never get the electric lift back to exactly where it started. The other end of the railroad is the same except that I have more space soI have fitted a wide base with two adjusters to get the levels right in both directions.


All joints across all boards are fixed with PCB sleepers and the track soldered in place. The sleepers are then glued to the foam board and we have joints with zero movement.

As you can see from the images of the layout, I have filled all of the spurs to get some idea of the capacity of each. With this information, I can incorporate this layout into the test phase of my university freight car routing project.

I am still getting used to my new DCC controller - I have one of the new Digitrax Evolution Duplex Starter sets (with the wireless controller). So far I am very happy with the set up. I have also got it integrated with JMRI. The eventual hope is to have the trains run from my phone and the routing software on my tablet. We shall see.



Next step is to mess about with the resistors on the LEDs as they are very bright and need toning down.