Sunday, 30 December 2012

Christmas goal achieved!

I set myself the goal of having something working over the Christmas break and I achieved it - for all of 2 minutes!

Firstly I twisted lengths of wire together. Technically you only need to do this for long runs but I decided it would be beneficial on tracing faults to have teh wires woven together. With Jill's help we were able to wind together 15ft of wire at a time using the Dremel. I then soldered on 30 individual wires to the different sections and it looks like this.


The pair of wires that feeds a section I want to detect in is fed through a DBD22. The AIU is in place to take the output from the detector. The other piece of electronics is the UTP panel. The other board looks identical.

I then connected up the NCE station and it worked first time. The mainline points are very smooth. There was no jumping and despite not yet having wired the frogs the engines ran well.

Unfortunately I then shorted the layout and it all stopped. Most has come back to life but not everything. I suspect I may have damaged a decoder or I just don't know how to use the new equipment. The basics are definitely right. Now I need to go back to Kevin at CoastalDCC and purchase some short circuit protection.

Thursday, 27 December 2012

2 wires

It is often said that DCC can be run with just 2 wires. Now technically that is true if you have a simple circuit of track. The reality is significantly different. The recommendation is that you have droppers to each individual piece of track to minimise voltage drop. Therefore if your circuit contains just 4 pieces of track then you are immediately up to 8 wires. That is before you think about any accessories.

Here is the block diagram for the new layout.


It's really simple - honest! The pieces are:

  1. NCE PH-Pro - the main controller of the layout. This is the heart of DCC. 
  2. The handset (T shaped piece bottom left)
  3. Programming track (prog) for programming engines
  4. Track - which gets broken down into 3 districts each with their own short circuit protection
  5. UTP panel - central point for connecting the cab bus
  6. NCE AIU - auxiliary input unit for taking in random inputs - there will be 2 of these
  7. DBD22 - detects occupancy of a section of track - there will either be 3 or 6 of these
  8. Lenz LS150 - point controllers, there will be 2 of these
  9. Computer - any old laptop that can run JMRI
  10. USB to Serial converter - most DCC systems have serial connectors. Most laptops don't :-(
I can't be truly definitive as to how many wires there will be as the goods yard has not been designed yet. However, my initial calculation is that I will have over 200 wires to put on this layout.

If that wasn't bad enough I've worked out I need 3 different power supplies!

Wednesday, 26 December 2012

Fresh Start

It's been a while coming but finally decisions have been made and a start has been made. I've had plenty of ideas but have come down to a few core ideas that I want to achieve. In no particular order they are:

  1. Exhibitable - it would be good to be properly on the exhibition circuit
  2. Able to be left up at home
  3. A town scene where the trains go between the buildings as on the Welshpool and Llanfair Railway
  4. A continuous run - it is nice to just watch trains occasionally
  5. Computer controlled
I had some boards that I had built for my 3mm work which I increased in width and two of those are capable of being set up in the spare room at a push with us still able to put other junk in the room.

The Welshpool and Llanfair Railway is particularly interesting and the book by Glyn Williams has some superb pictures. There are many cameos that can be modeled especially where the railway runs over the brook between the buildings.

For computer control I have moved on from my NCE starter set and bought the professional version from Kevin at Coastal DCC. Kevin was really helpful. He happily had a phone call with me one evening and we discussed the different options and what would and wouldn't connect. It was a straight choice between switching over completely to Digitrax and LocoNet or staying with NCE and mix and matching different manufacturers offerings to make it all work. I stuck with NCE as I prefer their handset and most of the extra work involved is just common sense.

So it's Christmas and what else do you do except eat, drink and work on the railway. The track plan is a simple circle. There is a 4 road fiddle yard at the rear and a passing loop at the station at the front. This will allow 4 trains to be run in a sequence. There will also be a goods area for shunting and showing off wagons.

I have made use of the new Peco mainline 009 points. At 18" radius I am hoping for better running than with the standard 9" radius points. I have used 2 Y points in the fiddle yard as these are 12" radius. I've laid the outer track only and now am going to wire it up and prove to myself I can connect all the DCC electronics successfully.

 

Friday, 3 August 2012

Simplicity of DCC (not)

Two weeks ago I hosted a regular meeting of the Oxfordshire Narrow Gauge Modellers. We spent some time reviewing the successful Open Day the group had had in my absence (I'm sure sure there is no connection really). The rest of the time was spent discussing my limited experiences of DCC. Since my last post, I had added in fresh GaugeMaster point motors to replace the old Peco ones. I could have stuck with Peco but I would have had to glue on frog switches. It was easier to start afresh. I also installed some Digitrax DCC point controllers. These worked fine with the NCE PowerCab that I have. However, I didn't install any point motors on the sidings of the layout, intending to do them later.

What did I tell everyone at this event? Firstly, that running a layout on 2 wires is a complete fallacy. Look at the picture to see the current state of wiring and it is incomplete. There is no control of two points, there is no electronics to tell the occupancy of a section and there is no short circuit protection, something definitely needed with the PowerCab.

Secondly, I demonstrated that DCC layouts can run as badly as any other layout if you don't clean the tracks and you don't wire the points. I think I put them all off DCC on cost and quality grounds. However, I shall continue with it as I seek to add more automation. The real question is whether NCE is really the best platform for automation. The PowerCab is a great starter set but I am not convinced it is going to be cost effective in the long term.

Sunday, 1 April 2012

Unexpected progress

Yesterday the decoder was fitted and it worked, for a whole 60 seconds! Typical progress it felt like. Today, under a magnifier I realised that I could see more of the commutator than I should. I hadn't pushed home one of the brush clips as far as I should have. Once I solved that I was able to drive the loco round the layout. There's a couple of dodgy points but I think that's a left over from how they were previously wired. I can fix that another day.

Time to try something new. Plug in the USB interface for the computer. Even that worked. I was able to read the decoder and store it's parameters on the laptop and also control the loco from the laptop. An amazing amount of success so the question is - should I push my luck and try something else new?


With JMRI there is a feature called the WiThrottle Server. This turns the laptop into a server to allow other devices to control the layout over the home wifi. Someone has written an application for the Android SmartPhone to do just that. I started the WiThrottle server and allowed it through the firewall when prompted. After that the app on the phone detected that a server was available and I just said 'connect' and this video was the result. Apologies for the quality but the light was not good. However, the numbers on the screen are the speed of the loco.

You cannot do this at an exhibition, I don't think, as it relies on a wifi connection and really it is a bit of a gimmick but it was surprisingly easy to set up.

Sunday, 25 March 2012

Murphy's Law

I made some progress yesterday in that I completed the rewiring of the layout. It was only 6 sections that needed to be brought to a central point and then the NCE PCP board had to be wired in.

Today's job was to rewire a Grafar loco to DCC. Fortunately the guys at DCC Supplies have manufactured something called a DiGi-HatTM which makes the job relatively easy.

So the loco was dismantled, the old brush housing removed and the DiGi-Hat inserted and it slops around. The instructions say " If it is a sloppy fit then it has either been drilled out or is a rare variant and you need a Phat-HatTM. As I've owned this loco since new I know I haven't drilled it out so I suppose I should be proud to have a rare item. However, I now have to wait 2 or 3 days for a Phat-Hat to arrive - have decided this is one job I would like not to be bodged.

Sunday, 4 March 2012

Calling it a day

I've had another couple of goes at making the chassis work but I can't make it run reliably. This is sooo frustrating that I have made the decision to stop modeling in 3mm. It's a great shame because 3mm is a fantastic scale and has a huge amount of potential. If I has modeled in 12mm instead of 14.2mm then it would have been significantly easier as old Triang ready to run chassis could have been used and also ready made track is available. As it was, I wanted to try finescale modeling and I have come up short.

The 3mm society has made great progress in making finescale less of a model engineering scale. There is the flexible track which is tremendous and components for making points are now becoming available. There are the square ended axles that should remove the quartering issues that beset my early efforts. I would recommend the scale to everyone. For me, I just don't seem to have developed the right skill set and I want to be able to see progress for the limited time I have available for the hobby.

So yesterday, at the Abingdon show, I handed over my 3mm modeling items to a fellow 3mm Society member for sale on the Society second-hand stall.

I shall take a little time to plan my next work but I know I will be going back to OO9. I have come to realise that I do like to see trains running round and round every so often and that won't happen in anything except narrow gauge. I also know I want to experiment with DCC and computer control. I have an NCE Powercab and so yesterday I bought their computer interface card and some point motor decoders. I intend to make use of Mospick Halt to try DCC on an existing layout before committing myself big time.

Yesterday I ripped up the fiddle yard of Mospick Halt and removed the electrics. The way I did the fiddle yard was not at all suitable for automated control so I will turn it into a traditional passing loop yard and see how easy it all is to control.

Here's to progress!

Sunday, 5 February 2012

Old loco chassis

I went to put the TCS MC2 DCC decoder into the Watford Tank but realised the decoder I had was about 1mm too wide. Thankfully TCS have a different decoder, the M1, that I will purchase which will be ideal.

Flushed with the success of having a working chassis I got out another loco that I had been working on - the Judith Edge LMS/BR Jackshaft drive 0-6-0DE. I knew I could put the existing TCS MC2 decoder in the body of the diesel. However, the chassis didn't unscrew from the body cleanly and then I realised the chassis was all pitted.

At this stage I am unsure how much it will clean up or whether the damage has gone too far. I shall take advice and decide what possible cleaning can take place.

Sunday, 29 January 2012

Oops - how did a year go by so quickly?

No the decorating didn't take a year. It did take longer than expected. I discovered the hard way that painting over blue with magnolia requires a LOT of coats unless you undercoat... Anyway, it is all finished and the room can be worked in again.

Enthusiasm waned because that $£#% chassis wouldn't work reliably. Whenever I thought it was quartered,a wheel would slip. Needless to say it ended up in a box in disgust.


Last year I did build some baseboards aimed at a new layout and then a eureka moment happened. A very clever man in the 3mm Society developed a method to punch the Society finescale wheels with a square hole so that quartering problems go away. It took me all of 5 minutes to find my wheels, pack them up, and send them off.

As soon as they were back, they were onto the engine and it still ran poorly. Very depressing and back in the box in more disgust.

New year, new resolution and I'm going to fix it. The bench was cleared tonight and the engine dismantled, various alignments checked, decisions made where washers should go to space the wheels, pick-ups adjusted and.... it just staggered along.

I've been doing all this work on the Bachrus Saddles as I don't have a layout but I popped the engine directly onto the track and it just ran straight off the end.

Out with the vernier and a closer examination of the saddles showed there's a difference between the top and bottom measurements. I had assumed the dimensions would be the same so when it was tight on the track it would be spot on for the wheels. No, the gap is wider at the top so when I set them tight on the track they were so wide there was very little contact with the loco wheels. Set them in by 0.5mm and it just works...

Maybe now we will see some proper progress.

Now it runs

Well it ran for a while, all of probably 10 minutes before the gremlins started appearing. On the right hand bend trains were slowing far mo...