Jump to content
 

Platform 1

Members
 View Profile  See their activity

  • Posts

    182

  • Joined

  • Last visited

 Content Type 

Blog Entries posted by Platform 1

  1. Platform 1

    All Change...

    By Platform 1,


    Building work
    It probably looks like it's been rather quiet on Platform 1 since the Spring. Far from it!
     
    The almost empty space in the loft above our garage has been beckoning for some time - 16x14 ft seems such a waste! So after considerable um-ing and ah-ing, I took the plunge and had it converted into a proper room, girders, joists, insulation, walls, the whole nine yards.
     
    Work started in June, and by golly those guys got through urn-fuls of of tea! This is after the joists went in:
     

     
    With most of the insulation and plasterboard up:
     

     
    And here it is pretty much finished - quite a difference!
     

     
    But when the joiners, sparks and plasterer finally departed, it was time for the ladders, paint brushes and rollers. Quite a lot of surface area to cover with that high ceiling, and with other pressing tasks to do, painting took a few weeks. But the result is worth all the pain (I think ) though it feels a bit cold up there at the mo'.
     
    So that's where the summer went! Now to plan that layout to fit the new space...
     
     
  2. Platform 1

    Make a DC Controller: Part 2

    By Platform 1,


    Electrics
    The amended objective was to produce a simpler, no-frills DC controller that would meet the original 'anytime anywhere' requirement, working with either a battery, AC or DC power.
     
    Circuit Design
    Starting with the emf-sensing feedback circuit used in Part 1, I removed the integrator and schmitt trigger U1 leaving just U1b as a simple non-inverting amplifier. The output driver Q1 - a cheaper and more easily obtainable BD243C - is included in the gain loop to ensure good output voltage regulation. Resistor values were re-calculated so that RV1 would provide a full 12v range at the output. D3 is retained to clamp any back-emf or spikes from the track and/or loco motor. D2 protects the op amp from reverse voltages at power off by discharging C3 via the supply rail.
     
    I also took the opportunity to add a current limiter to protect Q1 from track short-circuits - more a case of 'when' than 'if'. R7 and Q2 work to 'steal' current from Q1 when more than about 500mA flows through R7 - adjust R7 for different limit-current values. Two 3R3 0.25W resistors in parallel worked well in my case - my N locos draw up to about 250mA. Note that Q1 will need a heat sink for higher current motors; use of a metal box might be the best approach.
     
    The unused half of the op amp is wired as a unity gain stage (pins 6 & 7 linked) with its input (pin 5) grounded so as to minimise stray pickup and noise.
     
    To aid testing, typical voltages are shown in green.
     

     
    Construction
    The circuit was built using the same approach as in Part 1 - perforated stripboard - and with the same size plastic box 80x60x40mm. The reduced component count meant an easier fit on the board, but my board layout shown below is rubbish - too many wire links. In case anyone else wants to build it, pinouts for active components are shown below.
     

     
    All wiring was again checked twice.
     

     
    Testing
    As this is plain old DC a voltmeter should suffice when checking - no wizzy pulses to worry about this time
     
    Power was first applied with no load. The output seemed to waggle happily up and down as RV1 was turned forward and back. With a loco on the track, smooth control was noted - very gratifying! As expected, creep control was not as good as the feedback controller. However, it was possible to move a loco very slowly with about 2.5v applied to clean track.
     
    One observation: "zero output" at the minimum setting of RV1 is actually 2mV. It may be just stray pickup in my test rig, and is unlikely to cause problems, but worth noting.
     
    Top speed using a 12v battery - giving about 10.5v to the track - was more than fast enough for realistic running, so maximum oomph will rarely be required when using a 13-15v power pack.
     
    Conclusion
    Sometimes simple (KISS) is best!
     
  3. Platform 1

    Make a DC Controller: Part 1

    By Platform 1,


    Electrics
    Circuit Design
     
    As described in "A Test Track in N - 4: electrics" I started with an emf-sensing feedback design by Jonathan Scott (www.scottpages.net/UniversalTrainController.html second circuit from top). It was intended that the controller would be powered from AC, DC or battery to enable my N test track to be used 'anytime anywhere'.
     
    Construction
     
    I made a few small changes to Mr Scott's original circuit:
    C3 to 3uF to reduce inertia on a small layout add 1000uF reservoir/smoothing capacitor across the supply rails (I'm never happy running op-amps from raw unsmoothed DC even if they will work that way) use a 3-pole direction switch with the 3rd pole switching a red LED to warn of reverse before the knob is turned. Having gathered all the parts together, the big question was, "Will it all fit in there?".
     

     
    As this is only a small, one-off circuit, I decided against making a printed circuit board. So a piece of copper SRBP stripboard (aka Veroboard) was cut to fit in one of the slots helpfully provided by the box manufacturer. Then components were placed around the op-amp IC according to the circuit diagram - a few wire links were required to hop between strips. Everything fitted except for the bridge rectifier which was bolted to the box base. Very careful measurement was necessary to ensure that no components would foul the speed and direction controls mounted on the lid of the box.
     

     
    Then it was a case of cabling the various bits together: rectifier to board, board to switch, speed control and LEDs, switch to output. All wiring was checked twice before proceeding...
     

     
    Testing
     
    A 12v battery was connected, quick voltmeter check on the output, place a loco on my test plank, and... movement! But only a crawl, even with the knob up to maximum. Luckily I have access to an oscilloscope - this revealed several interesting things that I probably should have realised before testing:
    the output waveform never achieves better than a 50:50 mark:space ratio hardly any back-emf in the 'off' pulses the op-amp peak output is ~1.2v less than the supply rail (data sheet confirms it!) voltage drops occur across the bridge rectifier (1.4v), D3 and Q1 (1.3v) Together, this meant the loco only ever received 8v peak and 4v average - no wonder it moved slowly! So I tried a 15v power pack instead of the battery. Things improved - top speed was faster, with great 'creep' control too, but it just didn't feel right.
     
    A trawl on t'internet revealed feedback controllers tend not to to work that well with many small motors such as those used in some N gauge locos. Indeed, some tend to get very hot and in a few cases burning out. This wasn't looking good...
     
    In a flash of inspiration, I tried a 00 loco on my test plank (it has two tracks). Wonder of wonders - it ran great! Superb creep control, smooth acceleration, reasonable top speed. So this box works after all. Curious, I checked waveforms with the 'scope - sharp 12v pulses, plenty of back-emf in between, and generally 'alive'.
     
    Conclusion
     
    So yes, a result, but not quite as intended. After considerable pondering, I've decided to keep this controller box for other uses, but make a simpler no-frills DC one for my N test track. Reasoning is two-fold:
     
    1) it doesn't meet my original requirements for multiple power sources (12v battery in particular), and
    2) it works really well with 00 locos so it'd be a shame to waste it.
     
    The next exciting instalment - part 2 - will cover the simpler DC Controller.
     
    Postscript
    In case it may be helpful to someone, here's what a typical waveform looks like (it varies with loco - I guess every motor is different):
     

     
    Having done further tests, some 'DCC Ready' locos perform less well, with jerkiness at low speed and flickery headlights. But my old Lima pancake motors seem to perform quite well, though sounding raspy. Interesting!
     
     
     
     
     
  4. Platform 1
    A tidy example of 47323 Rover Group was acquired some years ago. The spartan chassis of these Lima models means a decent amount of space to fit circuit boards, LEDs and a speaker. I finally got round to adding these goodies as the Winter evenings fizzled out.
     
    Although covered elsewhere, I felt there were enough detail differences here to make a new write-up worthwhile.
     
    Approach
    My choice of components was:
    Zimo sound decoder, following good experience with a class 25 - MX645 with sound project DigiDrive E Version 3 Speaker to fit loco - Soundtraxx 28x28mm Express Models lighting kit - L47W Before starting conversions, motor and gear train were inspected and found to be in excellent condition - it looked like this model had hardly been used! Gears were carefully cleaned of excess grease and given trial runs. I'd intended to replace the ringfield motor with a Modeltorque motor kit, but it wasn't needed - so a spare is now available for a future conversion.
     
    The speaker fits into the fuel tank/battery space in place of the large steel weight. The weight was removed, and small holes drilled in the base to form a grille. Plenty of sound escapes through these holes but enough plastic remains to protect the speaker against dirt and damage from below. The aim is to form an air-tight rear enclosure using 1.5mm thick plasticard suitably sealed round its edges with white-tack (other glues are available).
     

     
    When selecting a speaker, I try to find one with the lowest resonant frequency (fo) that fits the available space. My initial choice of a 31mm round QSI HB122R turned out to be very slightly too wide to fit the tanks, so instead I plumped for a Soundtraxx 810131 28x28mm (fo 310Hz). On this speaker, the circumference of the cone is almost flush with the frame, so to avoid distortion, thin strips of plasticard were glued onto the front edges of the frame to ensure the cone-edges cleared the bottom of the tanks.
     

     

     
    The MX645 sound decoder is fixed to chassis using a small piece of double-sided sticky tape.
     
    Lighting LEDs were then added to the body shell per Express Model's included instructions. Required modifications are straightforward; just a few cuts to remove parts of the cab plastic that would prevent a good fit. The lighting circuit board is fixed to the roof with a double-sided self-adhesive pad.
     
    The weight lost (260g) by the removal of the steel block is more or less replaced by the combined weight of the speaker, additional electronics and a chunk of brass cut to fit the remaining space above the speaker enclosure (38l x 26w x 12h). As shown below, the brass weight is actually two pieces of 25.4 x 6.2mm glued together.
     

     
    Wiring was completed using heat-shrink sleeving on all exposed joints. The motor RF suppression capacitor was also snipped per DCC best practice.
     

     
    Results
    Initial motor and sound tests were performed using the chassis alone - much easier to fault-find or modify without taking the body off again.
     
    Sound output was surprisingly good for this small speaker - with remarkable (and loud!) engine 'grunt' - but sealing that speaker enclosure is crucial.
     
    The lights can be programmed to work singly or in combination by setting the appropriate CV values. For my installation, changes from the default values are:
     
    CV Value Action 33 002 - Reverse direction of headlights - FWD 34 001 - Reverse direction of headlights - REV 42 012 - Map FO1 & FO2 tail lights to F8 key 60 015 - Dim head & tail lights to more realistic levels 61 097 - Enable the CV42 map 126 002 - FO1 on - FWD 127 001 - FO2 on - REV  
    Finally, the body was carefully replaced, checking clearances in case the new wiring fouled the motor bogie. And I must say it looks and sounds the bees knees!
  5. Platform 1

    What colour should the seats be?

    By Platform 1,


    Multiple Units
    Hi again...
     
    Sounds daft doesn't it? Of course they were all green/brown stripe in the '60s, weren't they? Or were they? Does it matter?
     
    Such a simple question, but oh so difficult to find the answer. Butterley has restored M79900, the sister unit latterly known as 'Iris', and there's a bunch of really great photos on the web (such as here and here). Current upholstery is blue, but long research found several useful pics of the exterior in colour, but the few interior shots were mostly in monchrome.
     
    As can be seen from the earlier post, I eventually concluded the consensus is maroon for 3rd class and blue for 1st. As M79901 had only one class, it was settled. Of course, if you know different...
     
    So as can be seen from the last 2 posts, there is a collection of parts to fix together - another pic below. This is going to be, uh, interesting!
     

     
     
  6. Platform 1

    I've started so I'll, er, flounder...

    By Platform 1,


    Multiple Units
    Hi!
    Last time I set out the stall so to speak. So I thought I'd reveal a little of the motor bogie troubles encountered along the way.
     
    In them olden days, MTK offered a motor bogie kit as an option to add to their DMU kits. It was set of plastic and copper parts which went together to make a 4-wheeled bogie with electrical pickups. In fact it was much easier to assemble than their coach or DMU kits! Add an X04 or a Romford motor and bingo! a pile of useless junk . This unit may have been OK for some kits, but in mine, the motor armature fouled parts of the chassis which were designed to hold the pivot clip. Also, when the clip was installed, the tension it placed on the bogie effectively squeezed the motor creating friction to the point of stalling! And this was before it was anywhere near the body it was designed to power!! Many hours were spent trying to resolve this, but eventually the brittle plastic (now 25 years old) snapped and I abandoned the whole unit.
     
    Here's the wreckage:
     
     
     

     
     
    So now I had a spare Romford motor to fit into another loco, and a problem to solve: how to power this thing?
     
    Much research and thinking ensued - more modern technology is available now, so several options, e.g. black beetle (expensive, does this kit warrant such a good motor?), other proprietary units. But I hit on a cunning plan - why not obtain a RTR power bogie as a spare part and just tweak the side-frames? To cut a long story short, I placed an order and waited nearly a year, then cancelled and tried another supplier, then just happened across something similar enough to do the job at a local railex. This hobby certainly needs patience .
     
    Here's the 'new' bogie - much nicer!
     

     
     
    Which brings us nearly up-to-date. Next - what colour should the seats be?
     
     
  7. Platform 1

    Where do I begin?

    By Platform 1,


    Multiple Units
    Returning to the hobby after, errrm, 25+ years, a rummage through old boxes revealed a few unmade kits. "What a great way to spend those dark winter evenings", I thought.
     
    I chose the MTK Derby Lightweight Single M79901 because I'd not attempted a DMU before and it would be summat small to fit into a simple end-to-end layout. Some construction progress was made, but nothing fits, the parts are poorly made, the air was often blue , and then the days lengthened - everything slowed to a halt (short platform, trains stop by request). Just very difficult to stay motivated...
     
    But that was last year! So this year, maybe a blog would provide a focus - a driver (5ft? please, no more silly puns) - to continue and actually get something built and working. Or not, and you can all have a good laugh . So first a little background...
     
    Back in the early 1980s, I had a 10x8 ft insulated shed and a never-finished roundy layout in OO. Whatever I build now will re-use some of the old track and rolling stock - primarily because of cost. But that nice shed was 5 house moves ago, so now very little room for track or wiring. DCC seems to be the way to go, and early trials are very promising for realistic slow running in confined spaces. Anyway, back to M79901...
     
    Last season, the shell was painted, the driving ends were glued and painted, and the seats were cut, stuck and painted. A start was made on the motor bogie, but that was a disaster (more about this in a future post). So a major rethink was needed on bogies and motor arrangements.
     
    I realised early on that the finished article wasn't going to be an accurate model, because there are too many issues, e.g. the engine castings are wrong, the cab panels are for a different unit, the motor bogie is a joke, but the overall 'feel' is reasonable. So this will be a valuable learning exercise - maybe tackle something newer and better later?
     
    So this is where we are at the moment:
     
    The body shell:
     

     
    The seats and partitions:
     

     
    One side underframe:
     

     
     
    Next step is to decide how it'll all fit together in a way that enables maintenance, since the instructions have some funny ideas about how to fasten things .
     
  8. Platform 1

    00 Track Comparison

    By Platform 1,


    Track
    It's been a while! Amazed how long it takes to turn a new room into a usable space - painting, flooring, shelves, heating, workbench, lighting - together with the usual demands of family life...
     
    So here we are with a bit of new activity before the layout proper gets under way. The shots below show three types of 00 track: Peco streamline code 75 flat bottom, the new Peco code 75 bullhead, and C&L flexi track.
     
    The ballast looks over scale to me. Which do you think looks best?
     

     

     

     

  9. Platform 1

    Bits and Pieces

    By Platform 1,


    Baseboard
    Desperately slow progress, what with family commitments and other life stuff...
     
    However, the dropped (removable) board now has its elevated sections for the bridge under the mainline and a shorter bit for the head-shunt.  Once the alignment was drawn in to fit the plan, the track board supports were glued (with screws in some cases) and various fillets added for lateral strength.  Thought I'd try some laminate flooring underlay foam on this layout - seems to stick down OK with PVA glue, but let's see how it takes paint and Copydex for the track side.  Early tests reveal it's quite quiet, so here's hoping...
     
    Oh, and the track and points have arrived - just need to order some microswitches for the frog switching.
     

     
  10. Platform 1

    Activity resumes - track!

    By Platform 1,


    Track
    So there's the summer been and gone... wasn't it a good one (in the West Country anyway)?
     
    Now the evenings are drawing in, thoughts have returned to the loft layout.  Track has been acquired, and some progress made in laying it.  Each board has to be laid and wired separately in case it has to be removed some time in future.  The dropped board is easily (!) removable to access an under-eaves cupboard.
     
    The fiddle yard area is a bit of a squeeze to get to (see photos) so hunched-up working means limited spells at that end.  The five tracks were straightforward, but the removable join was, er, rather tricky to work on.  But that end board is wired now, so things should get a little easier to access.
     

     
     
    The board joiner strips are made from pieces of copper-clad board, intended for printed circuits - the technique is described in many RMweb posts.  It's a very fiddly job, as each piece of rail has to clipped in place while it's soldered - I used a couple of small letter/bulldog clips.  So far, wagons run pretty well across the joints, but we'll see what happens as the temperature and humidity changes...!
     

     
     
    Here's a pic looking in the other direction.  The dropped board will have a bridge under the main line and a head-shunt, both providing some scenic interest.  There will also be a tunnel mouth as scenic break just beyond the coal wagon, so that the points to the fiddle yard are all 'behind scenes' yet accessible.  Only six inches or so of height for a back-scene though!
     

     
    Next steps are to finish track laying and wire up the other three boards so that testing can start.  At the current rate of progress, that'll be slower than Crossrail !
     
     
  11. Platform 1

    At last, servos and wiring...

    By Platform 1,


    Electrics
    Over a year without posting anything!  Disgraceful - dunce's cap firmly on.
     
    Despite the pandemic, lock-downs and other distractions, things haven't developed very much recently.  The first lock-down was taken up with long summer walks and home-based IT work for a local charity.  By mid-summer it was much too hot to work for long in the loft - 30degC plus some days.  Then suddenly it was cold again!
     
    Work restarted before Christmas (2020!) looking at the design of servo point motors.  Having previously successfully used solenoid motors, they didn't look or sound very realistic.  Small servos are (were?) inexpensive so how to make a compact but adaptable assembly to put under the layout?  It needed to be cheap, easy to make with limited tools and easily repeatable - thirteen are required.  Various scribblings and mock-ups were tried, eventually producing this not-very-original design.
     

     
    The servo is a tight fit into the aluminium U-channel, and butts up against a short piece of balsa or similar softwood, on which the frog polarity micro-switch is mounted.  My baseboard is just 3mm thick ply, so the pivot is mounted a few mm away in order to provide enough throw at the point blades.  The bracket can also be mounted on its side if needed.  There are screw mounting holes in the U-channel, but I found that double-sided black sticky pads normally used for affixing car number plates worked really well.  A thin coat of PVA or paint on bare wood assists the grab of the pad.
     
    The material used in this  U-channel is tough and quite difficult to cut, so I also made a bracket using PVC U-channel.  It isn't quite as rigid, but seems firm enough to resist side-movement from the servo.  Here is one such:
     

     
     
    The wire pivot bracket (out of focus at the back) still needs to be placed and glued onto this version.
     
    These servos will be controlled by an Arduino and two LED/servo driver boards connected via an I2C bus.  This has the advantage of few wires as mentioned in an earlier post - two data wires and two power wires in a screened cable.  So happy times were spent during Christmas restrictions programming an Arduino Nano to move the servo by about 90 degrees - enough to move the point blades and reliably operate the micro-switch.  The prototype mock-up works, so things a re looking hopeful.
     
    Wiring track has also been proceeding as a background task, one board at a time.  The corner board is now being worked on, with only a few droppers to fit.  This is what it looks like so far:
     

     
     
    So once the track power feeds are all soldered in, it will be time to re-assemble everything and do some test running - I can't wait!
     
    The next few stages will include:
    make more servo brackets fit servos to baseboards fit the control electronics complete I2C wiring.  
    Watch this space!
     
  12. Platform 1
    This week I have mostly been... messing with paint!
     
    Context
    On my long list of modelling-things-to-do is the assembly of a rake of coaches to make a 4-TC set. This will involve a cut and shut for the driving ends, plus two of the excellent Bachmann Mk1 coaches to form the TBS and TFK.
     
    Whilst gathering the various bits together, my mind turned to how I would paint the driving ends, and importantly achieve a decent colour match to the two middle coaches. Whilst many colour variations exist in the prototypes, 4-TC units mostly stayed in their sets, so their colours were well matched.
     
    Background
    My modelling box contains a range of ~35 year-old Humbrol enamels, including some of the old 'authentic' range. Amazingly, almost all of these still flow and cover well after a thorough stir. However, the old HR135 (blue) and HR136 (grey) are no longer good matches - maybe paints have improved over the years, or perhaps the pigments have aged. Anyway, time for more paint!
     
    I'd prefer to spray my new coaches, so a number of options exist:
    Railmatch or Precision spray enamel - no local stockist, expensive to courier since the 'mail of hazardous chemicals' rules, cans have a tendency to splatter Acrylic - less odour, easy clean, need to acquire airbrush (might be good to have for other jobs) Plentiful choice of brand now - best to keep to one if possible for consistency and compatibility Vallejo are getting good press, increasing availability, including a full rack locally - let's try some.

    Colour Trials
    What follows is not a particularly scientific approach and may not be repeatable - but it is the best I could do with what I had available. First I set up my TFT display to be accurate as possible: colour temperature (6500K), grey scale, brightness and contrast using Windows 7 'Calibrate colour' - other tools exist. Importantly, colour rendering varies by screen type and model, and even in JPEG compression, so the pics below may look quite different at your end
     
    Here's some sample Vallejo shades available:

     
    And here are some samples obtained from good photos of prototypes:

     
    The upper group are from a coach and have measured RGB values overlaid. The lower group are from a class 37 and also have three Vallejo samples (with 9xx numbers) for reference.
    On my screen, all these colours look a little too green to me. By the way, printing these was useless as my Epson SX colour rendering was appalling even using different colour profiles. It may be worthwhile if you have a super wizzy professional hi-grade colour printer.
     
    Things looked promising, so it was off to the shops to obtain a few of the colours that seemed to be in the right range.
     
    Once the paints were obtained - they are in eye-dropper style bottles which facilitate easy mixing and should reduce waste - the grand test began. For each colour, the bottle was given a thorough shake and roll, then two coats of each colour was painted on white card - the lines are useful to indicate the degree of coverage. The brush was cleaned and dried after each application. Here are the results after about 12 hours drying time:
     

     
    This pic reveals several interesting things:
    the camera has rendered these colours quite accurately in daylight (no flash), everything being slightly darker 'in the flesh' some colours are quite close to the colour chart but some are miles off Dark Blue is lighter than Medium Blue - have I a duff sample? my old Humbrol authentic colours look quite wrong Bachmann's rail blue seems slightly darker than the prototype

    Some of my older Bachmann Mk2 coaches have a slightly duller shade of blue, and most of my rail blue diesel locos (various makes) are a little lighter - somewhere between 962 and 963. So that's an area for further trials.
     
    Conclusions
    As if I didn't already know, don't trust the colour charts nor the bottles in the shop! From these trials, I found the best reasonable matches to a recent Bachmann Mk1 coach are:
    Rail blue - 70.963 Medium Blue (may benefit from a little 962 mixed in) Rail grey - 70.883 Silver Grey

    And I need to splash out and buy an airbrush soon.
     
    I'd be interested to know whether anyone else has tried this, whether your 'dark blue' looks lighter, and what you find works for you.
     
    Postscript
    These 'matches' only get close on a white undercoat/base. Further tests using a grey primer base coat revealed some issues - see next entry.
  13. Platform 1

    Clearances

    By Platform 1,

    The double-track main line has now been properly trimmed and laid.  Cutting and re-spacing sleepers is very time consuming!  I decided that the concrete sleepers on this section would benefit from being opened out to 8mm centre-centre.  After gluing and adjusting - 5mm coffee stirrers are perfect - the track does look better, even though it's all a compromise in 00.  The station area will be bullhead, so no chopping and spacing will be needed there.
     
    The other time-consuming element was aligning track at board joins.  Because the boards are supported on shelf brackets, and the need to remove the  elevated section for occasional cupboard access, getting the horizontal and vertical alignment spot-on was, er, challenging!
     
    So then on to the crossover at the station throat.  Even with the cut-and-solder mods for DCC, pointwork was fairly straightforward to lay - despite my continued tendency to prune rail lengths just slightly short.  But it looks reasonable and hopefully will look much better when painted and ballasted.  Before going much further, a check was made for clearances on the curve:
     

     
    This is worst case, assuming I've not overlooked any longer rolling stock.  Not too bad is it?
     
    Now to lay track in the station area... and then under-board wiring... and point servos... and... eventually see trains run!
     
  14. Platform 1

    Revised Track Plan

    By Platform 1,

    So here's the next development - having realised the similarity to Minories, it is now a Minories with sidings.  This one addresses the issue of the headshunt not being accessible from the inbound route.  The right-hand part with the fiddle yard is unchanged from the initial plan.
     

     
     
    A variant I tried when drawing this is to have only a single siding to the bridge, which then becomes a headshunt to two splayed sidings in the top-left-hand corner.  This arrangement might be better for goods and/or simmering engines awaiting turns.
     

     
    Any more thoughts anyone?
     
  15. Platform 1

    Initial Track Plan

    By Platform 1,

    The freezing weather has enabled me to spend a few happy hours fiddling with track components in AnyRail.  This is my initial attempt - it seems to satisfy a number of criteria:
    a mix of urban and semi-rural double-track somewhere undulating terrain with e.g. a bridge interesting operation somewhere to shunt
    Any thoughts anyone?
     
  16. Platform 1
    Why is nothing simple?
     
    Further test samples viewed in various types of light revealed some issues. In particular, the use of a grey primer undercoat produced a bluer cast than a white base, despite applying two top coats. As noted in the last post, I'm keen to achieve a close match across my rake of 4 coaches, so can we do better?
     
    A few more shades were obtained (this is becoming expensive) and more trials carried out. To cut to the chase, better colour matches to my two Bachmann Mk1 coaches of recent (Autumn 2013) purchase are:
    Rail blue - 50-50 mix of 70.965 Prussian Blue and 70.966 Turquoise (this mix can be tweaked; 60/40 looks marginally better in daylight)
    Rail grey - 70.883 Silver Grey or 70.971 Green Grey (better in daylight)

    Following Niles' kind suggestion, I also tried Citadel's 'Stegadon Scale Green' - Zak in the GW shop was most helpful, ferreting out an open bottle to inspect near the window. A test sample looks a greeny-grey with a hint of blue - in fact it's a really close match to my older Bachmann Mk2 coaches!
     
    For reference, the various 'official' shades of BR rail blue and grey are shown below, together with RGB swatches. As noted, they differ from my actual colour-matching results.
     
    10-Feb-14: updated for RGB values & closest match refs
    25-Feb-14: extended grey info, add grey scale

     
    Maybe this research will help someone else - if so, please would you leave a comment?
  17. Platform 1
    Controller Construction and Test
     
    This turned out to be, er, interesting! The details became rather complicated so rather than get bogged down here, I've put all the electronics geekery in a separate blog entry "Make a DC Controller".
     
    In summary, I ended up with two controllers, using a simplified non-feedback design for this project.
     

     
     
    Costs
     
    Case wood and hardware - about £17 (carry handle was £5 of that ).
     
    Track - about £35. Even discount prices seem to have gone up in recent years...
     
    Controller parts - under £15 (+ p&p) but will vary with supplier and what's already in hand. Much cheaper than commercial offerings, though soldering skills are needed, and of course there's no warranty or support. If it breaks fix it!
     
    Final thoughts
     
    This project did seem to be a lot of effort for a relatively simple, small result. But in the absence of anywhere else to run trains, it has proved really worthwhile. I'd be interested in your comments - it seems very quiet in this corner of the blogosphere...
     
    I may add a limited amount of scenic treatment in due course, maybe in one half of the case. First step in that direction will be to make a building to cover the central hinge and provide a sense of scale and context.
     
    An open and shut case?
     

  18. Platform 1
    A simple track plan like this needs only basic wiring:
     

    track power feeds on each side of the box
    connector across the hinge join
    jumpers to isolated rails
    jumpers to the toe of sidings.

    But all wiring is on view when the case is open, so a neat job is called for.
     
    Wires were routed under sleepers and soldered to outer edges of rails - with hindsight I should have used something smaller than the standard 2.3mm bit on my X25 iron, and perhaps attached certain wires before fixing track in position. The connector is a 4-way PCB plug/socket with just two pins used; always useful to have spare ways! The longest wire runs are enclosed with U-channel polystyrene strip stuck with double-sided tape.
     

     
    And so to testing. The amazing thing was - it all worked first time! Only powered from a 12v battery with a series 50 ohm resistor, but locos trundled round without de-railing or stopping. Woo!
     
    For loco control, it's intended to fix in one corner a box of home-brew electronic gubbins, so that either a battery or a small mains adaptor ('wall-wart') can be plugged in for instant running "any time, any where".
     
    So now I'm building an analogue controller more-or-less to a circuit by Jonathan Scott (second circuit from top) using copper stripboard, LEDs, control knob and a polarity switch enclosed in a 80 x 60mm plastic box.
     
    So all we need now is to finish that and get it working...
  19. Platform 1

    Baseboards!

    By Platform 1,

    Baseboards constructed at last! Here they are:
     

     
    The key question is what to build here? Almost certainly this will be 00, because I've quite a lot of stock gathered over the years - though some is tatty and somewhat below par so will have to go!
     
    So here's the overall plan:
     

     
    (Red blobs are wall brackets)
     
    No real track design yet, just a few placements to get some idea of what might fit in this space. Some sort of bridge structure on the dropped board, maybe built-up on the left, and more rural on the right leading to a fiddle yard? It would be nice to fit in a raised track at the back left to add vertical interest and scope for extra running.
     
    All constructive thoughts welcome!
     

  20. Platform 1

    Controlling Points

    By Platform 1,

    Whatever my track layout will be (still thinking about that ), it'll need some points. But this time, I've decided I shall NOT have a tangle of wires under the board! Even on a small layout, tracing wires/faults/feeds etc can be a right pain in the neck!
     
    So the plan is to have one pair of wires to feed DCC everywhere, and another 4-wire cable to feed digital point control using i2c or similar control. Although I've a collection of old Peco point motors, I can't say I've ever been keen on their 'snap-snap' whether using AC pulses or DC capacitor discharge. Reading through many threads and blogs here, it seems servos are a good way to go.
     
    Suitably encouraged, I sent off for some servos, an Arduino nano and some I/O expanders. Actually, they are cheap clones because I've no idea whether it'll work out or not. Then I threw the lot together on a breadboard, cobbled some program code together, and this is what I have today:
     

     
    Now I should admit here that I have a background in electronics and IT, so doing this wasn't at all fearsome! However, I am amazed how easy it is to put the pieces together. The Arduino libraries cover a huge wealth of functionality, so it's really a case of integrating the parts with some C code statements and testing. The nano is particularly easy to use because the USB interface makes uploading binaries a doddle. What have I been missing?!
     
    The outline design is thus:
     

     
    Not very elegant, but I hope it gets the idea across.
     
    So far, 15 switch inputs are working on the nano - the maximum that can be configured without input multiplexers or shift registers. My expanders are actually 16-way, but they were cheap and had easy-to-use libraries available. So here's hoping these boards will fit quietly under the baseboards and 'just work'!
     
    P.S. I'll post the code when I get chance.
  21. Platform 1
    So now we have a wooden box which needs to be turned into a model railway.
     
    It's very tempting to start laying track at this stage. But some things need to be done now to properly prepare:
    apply 'knotting' to prevent knots in the wood from oozing and/or weeping resin later (but omit on parts that will be varnished)
    paint inside while it's still clean and bare
    two coats of varnish outside to help protect from inevitable knocks and scrapes (and to "bring out the natural beauty of your wood" )
    add catches and a carrying handle.

    When all this was done and thoroughly dry - over a week in recent cold and damp weather - at last it was time to lay track. I favour a bed under the track to help deaden noise and provide an even running surface. As it was to hand, 1/16th cork sheet was suitably hacked and trimmed, but foam or felt could be used.
     
    I decided at the planning stage to use set-track for the main curves - both for assembly speed and to help maintain correct geometry in a tight space. On the track plan, pre-formed track parts are depicted in grey, flex-track in blue. All track is therefore PECO code 80 - although code 55 looks much better, it doesn't mate well with the set-track. Other manufacturers' track could be used if it'll fit and meets the need.
     
    Note the short extra piece of flex-track between the two points; this is to avoid cutting into points along the hinge line.
     
    For this small layout, I pre-assembled the main oval, positioned it more or less centrally, then marked the rail at the board joints with permanent marker. The tricky bit was actually cutting track in the 3 marked places. I wanted close rail joints across the box halves, but there just wasn't room to use the razor saw in situ. So the pieces to be cut had to be extracted, sawed in a bench clamp and put back in place. It's quite a fiddly job in N - most ended up within 1 mm of where they should be!
     
    To fix track to the cork, I used Copydex. In days of yore I always used track pins, so for me this was a first, as many people recommend it for track laying. But be warned - it has a powerful piscine pong!
     
    Having roughly laid track on top of the cork, I worked round in sections, gluing track to cork and cork to box, holding it down with heavy items until the glue set (about 20 mins each section).
     

     
    The remaining point and sidings were added later.
     

     
    I had assumed some form of rigid anchor would be required to minimise movement of rail ends when the case is closed. However, the glue seems to hold the track very well, so we'll see how it goes...
     
    So all we have to do next is the wiring...
  22. Platform 1
    Despite wanting this thing to 'be quick to make' I hit an early snag - all is revealed below...
     
    One of the key features of this design is that the track will go inside the frame, rather than sitting on top of the frame per the conventional approach. This means:
    no underboard wiring, point motors etc (a false bottom could be added - but this is meant to be quick and simple)
    no track pins - unless it doubles as a bed of nails!

    Assembly tolerances must be quite tight, as even small errors will cause running problems across board joins in N. My joinery is not the best, so this aspect was daunting
     
    Here are some of the box parts carefully cut to size and smoothed down:
     

     
    I did consider halving or even dovetailing the joints, but 14mm isn't much to work with, so I chickened out. A skilled carpenter would do it properly.
     
    One half of the box glued and pinned and screwed:
     

     
    On starting the second half, it was clear something was wrong. The longest side piece was bowed by about 2mm at the centre - despite carefully selecting timber when buying, rejecting badly knotted pieces and so on. The remaining offcuts I had weren't suitable, so how to un-bow a piece of timber?
     
    In fact, it's pretty straightforward for short-ish pieces. Add moisture on the concave side of the bow, then clamp so that an opposite bow is applied. I used a kettle (steam, woo!), a piece of kitchen towel, and a folded plastic bag to retain the moisture, and stuck the whole lot in a vice between two rigid battens for about a day. Then let it all slowly dry out - hopefully straight! If this seems a right ol' palaver, well yes it was, but those odd mm might compromise the result - derailing isn't the sort of test I had in mind!
     
    Whilst waiting - two days - there was time to obtain suitable hinges and catches. And also decided to add triangular fillets to strengthen the middle hinge supports.
     
    The second half of the box was then finally glued, pinned and screwed. When dry, the hinges were fitted - proving quite fiddly to align and fix all the pieces:
     

     
    - these pics highlight inaccurate chiselling! But at last things are taking shape:
     

     
    Of course, when the box is closed, there is a large gap at the hinge end. This must be covered so that rail ends etc are not damaged in store or transit, but easily removed to enable the two halves to butt together when open. My intended solution is to use a piece of ply/mdf/sheet stuff attached with velcro - with perhaps a couple of cross bearers to act as feet. I'll work out that detail later...
  23. Platform 1

    A Test Track in N

    By Platform 1,

    I think I've mentioned somewhere on RMWeb that, like many others, space is an issue for me. So now that N gauge RTR models have improved enormously, that might be the way to go in order to get something running! Plenty of small-space ideas on here and elsewhere for inspiration...
     
    But first, new locos would need to be run-in - 'run for half-an-hour in each direction' say the manufacturers' blurb - so a test track would be a good idea.
     
    After scribbling with pencil and paper for a while, this plan was hatched:
    a small 'roundy' to enable hands-off running-in
    maybe a siding or two to add interest
    portable and easily stored
    initially DC, able to handle DCC
    quick to make and commission

    Calculations based on dimensions of 2nd radius curves revealed that a nominal 3ft x 2ft oval would fit on a board that could fold in half to make a compact 24x18 inch box. So Anyrail was used to work up a scale plan - this defined the overall case size to be a little larger at 40x24 inches, or 1016x610mm. Each half of the case would need to be about 2in high to provide plenty of clearance for rail + underlay + stock. Here's the track plan:
     

     
    The case parts were then designed:
     

     
    Wood was duly acquired. The sides are a bit non-standard; they are moulded 14mm softwood skirting trimmed to 50mm wide by removing the moulded long edge. The bases are 6mm ply. However, the DIY shop managed to cut the ply out of square on their saw-bench, so the two pieces had to be suitably trimmed square. Doing this shaved a few mm off the short sides - my eventual base sizes were therefore 609x495mm.
     
    So all we need to do next is assemble the bits!
×
×
  • Create New...