Eastwood Town - A tribute to Gordon's modelling.

[gordon s]

18th of Jan and today saw golf cancelled for me anyway as it was carry only.  As the years slip by I am conscious of minor changes in my body and level of fitness.  Seems like only yesterday I was carrying a bag and had no problem with 36 holes in a day.  Sadly those days have gone and now I have to use a trolley and whilst 36 holes in a day are still manageable, those days are reserved for when the sun is shining and the shorts are on.

 

It was a funny old week and I seem to have spent a disproportionate time on RMweb.  There must be something about track threads that bring out the worst in people and tempers were getting frayed. Unfortunately there were some statements made without substance, opinions became entrenched and in some cases, they were unable to see the others point of view.  Eventually the thread was locked and the parties involved retreated to argue another day.  It's one of those subjects where there appears to be a very simple answer but the more detailed discussion becomes, the more complex it becomes.  It also means that words have to be chosen very carefully as often one word in the wrong context or position can be sufficient to light the blue touch paper.

 

Anyway I digress.....

 

Some work has progressed and I'm now able to see what areas will look like.  The section across the stairs will have four different levels.  One fixed double track loop plus two descending single line tracks which are the feed and return from the reverse loop/storage roads.  The final pair of tracks are the climb up to ET terminus.  Looking at a plan is one thing but it's not until you have cut wood can you really judge the vertical and horizontal spacing of these tracks.  Thankfully they appear to be OK and just standing some Slaters sheet alongside the ET feed shows some promise.

 

 

 

So that's all positive, but sadly now trackbeds are in place, it has raised another issue.  One of my fundamental goals in any layout plan is that all pointwork must be easily accessible.  No matter how careful you are any mechanical item with moving parts is going to fail at some time or other and I was horrified to see just how restricting the loop track beds were going to be once I had moved from a computer generated plan to real component parts.  I really didn't spot this was going to be such an issue and had a tie bar or point blade fail, it would have meant considerable dismantling to get access.  My original thought was to have the throat of the storage sidings all on one board that could be slid out in one piece.  Unfortunately that cannot be done as there is only 70-80mm clearance from the track bed above and Tortoise motors are at least 82mm deep and as such there is not room to remove the board in one piece.

 

 

Back to the drawing board....

 

Just spent a couple of hours this morning moving all the pointwork that feeds the reverse loop around the corner and away from it's current position.  First views suggest it is promising, but I have had to drop to 32" radius on a few feeders from the C10 turnouts, which goes against my design principles, but at this stage, I cannot see another option.  The second problem is how and where to add riser supports to the upper level tracks where there are low level storage loops beneath.  I wanted all of these to coincide with baseboard edges, but may have to accept one or two trackbeds will overlap board joins if that is the only way I can provide risers that do not clash with lower levels. 

 

The second alternative will be to move the pointwork inboard and away from the upper level loops.  The down side of that is that it will mean shorter storage roads.  How much shorter will remain to be seen.  That will take the pointwork under ET terminus, but that's not such an issue as clearance under there is around a foot.

 

Watch this space.

 

Next time anyone suggests building a multi level layout is easy, I will scream....:-)

 

Edited January 18, 2016 by gordon s

[47137]

It was a funny old week and I seem to have spent a disproportionate time on RMweb.  There must be something about track threads that bring out the worst in people and tempers were getting frayed. Unfortunately there were some statements made without substance, opinions became entrenched and in some cases, they were unable to see the others point of view.  Eventually the thread was locked and the parties involved retreated to argue another day.  It's one of those subjects where there appears to be a very simple answer but the more detailed discussion becomes, the more complex it becomes.  It also means that words have to be chosen very carefully as often one word in the wrong context or position can be sufficient to light the blue touch paper.

 

Anyway I digress.....

 

. . . 

 

So that's all positive, but sadly now trackbeds are in place, it has raised another issue.  One of my fundamental goals in any layout plan is that all pointwork must be easily accessible.  No matter how careful you are any mechanical item with moving parts is going to fail at some time or other and I was horrified to see just how restricting the loop track beds were going to be once I had moved from a computer generated plan to real component parts.  I really didn't spot this was going to be such an issue and had a tie bar or point blade fail, it would have meant considerable dismantling to get access.  My original thought was to have the throat of the storage sidings all on one board that could be slid out in one piece.  Unfortunately that cannot be done as there is only 70-80mm clearance from the track bed above and Tortoise motors are at least 82mm deep and as such there is not room to remove the board in one piece.

Track topics  - me too (and some exasperation).

 

How about miniature servos (as used for model aircraft) for point controls?

You can buy kits which include mounting brackets for model railway use, and also driver boards. You can set up end stops and operating speed. There are write-ups on line and on the RMWeb.

 

- Richard.

Edited January 18, 2016 by 47137

[Siberian Snooper]

Track topics  - me too (and some exasperation).

 

How about miniature servos (as used for model aircraft) for point controls?

You can buy kits which include mounting brackets for model railway use, and also driver boards. You can set up end stops and operating speed. There are write-ups on line and on the RMWeb.

 

- Richard.

MERG supply the operating hardware for servo's in the shape of Servo4, which will operate 4 servo's each they are about £7.00 each or 5 for £30. You do need to join MERG to get them though. The servos themselves can be bought in bulk for about £1.30 each based on buying 10 at a time for the SG90.

 

There are 2 methods for setting them up, a servo set box also from MERG or download the software to a laptop.

 

The servo's can also be used for semaphore signals.

[tender]

MERG supply the operating hardware for servo's in the shape of Servo4, which will operate 4 servo's each they are about £7.00 each or 5 for £30. You do need to join MERG to get them though. The servos themselves can be bought in bulk for about £1.30 each based on buying 10 at a time for the SG90.

 

There are 2 methods for setting them up, a servo set box also from MERG or download the software to a laptop.

 

The servo's can also be used for semaphore signals.

 

But you have to build the MERG units yourself which is not everybody's cup of tea. If you want to go down the servo route, Signalist (SC2) or MegaPoints offer a ready made solution (among others).

[martin_wynne]

How about miniature servos (as used for model aircraft) for point controls? You can buy kits which include mounting brackets for model railway use, and also driver boards. You can set up end stops and operating speed. There are write-ups on line and on the RMWeb.

 

How about an ordinary industrial solenoid? Magnetic-latching pulse type:

 

http://uk.rs-online....=PSF_421416|cav

 

12V pulse to operate (no continuous power needed, held in magnetically). Reverse polarity pulse to release.

 

No gears, no motor, no electronics, no flimsy plasticy bits, only one moving part and almost nothing to go wrong. Just add a return spring in the linkage. I think you could safely bury one of these behind the scenes and forget about it. Do the crossing (frog) polarity switch at the control panel, where it is much more reliable and accessible.

 

regards,

 

Martin.

[gordon s]

Moved on a bit since my last post.  Thanks for the info re servo's and solenoids, but I will stay with Tortoise motors as I have 72 of them that I bought a few years back from the USA.  I hadn't seen those solenoids before from RS, Martin. They are interesting and I will hold the info for future reference.  Whilst using my Tortoise motors does mean some re planning, sheets of A4 and the odd bit of ply are still relatively cheap.

 

Tried a few more alternatives whilst the hours slipped by.  Moving the pointwork inboard of the loops doesn't work as the length of each road is compromised.  After a while I started thing about my 36" minimum radius and whilst this definitely applies to point work I'm sure I won't have a problem in the storage roads as they can be used for DMu's and local stock headed by tank engines. Running all the roads through in one length is a win win situation.  The pointwork is all moved to areas where they are easily accessible and the minimum radius on the inside curve is still 32" so I'm sure that won't be an issue on plain track.  Not to mention a considerable increase in the length of each road.

 

I'm just taping up the template sheets now, so I should be able to see any difficulties once laid out on the boards.  My only area of concern is the riser supports, but I have some ideas how to deal with those.

 

 

 

[martin_wynne]

Just add a return spring in the linkage.

 

p.s. For a permanent layout, a falling weight on a cord over a pulley is more effective than a spring -- it applies a constant force (which is easily adjustable by changing the weight).

[Rowsley17D]

On those 32" radius tracks, don't forget to increase the distance between parallel tracks. You don't want long stock hitting each other.

[gordon s]

Thanks Jonathan.  I've have checked that out on Templot and Mk1's will just clear each other.  It's unlikely the two inner roads will hold Mk1's though as they will be reserved for suburban stock and freight.  No harm in widening them a fraction though as sod's law says I'll forget.......

[G.M.R.]

Hi Gordon. Long time since we last chatted. Happy New Year to you.

 

Great to see that ET is firmly back on your agenda. Your current layout looks excellent and I can imagine it taking shape within your train room.

 

Maybe I can offer a couple of thoughts from my experience in admittedly a much smaller room.

 

If at all possible, try to avoid hidden inclined lines especially at the back of the room. Access can be very difficult, and I got frustrated at not seeing my trains for long periods whilst they made their way from storage to scenic levels. I now have my inclines very much in view, and in their own scenes. It is just a matter of how you look at the trains passing.

 

How do you walk from one side of the storage lines to the other? I have removed all duck unders and now have lifting or dropping flaps so that I might still be able to get around the train room as each year passes! Do you get to either side from the stairs?

 

Now that I've found your post again, I will look forward to watching your progress on this latest version of ET.

Edited January 19, 2016 by G.M.R.

[gordon s]

Hi Geoff, great to hear from you again.  Once the weather warms up a bit, we ought to get out for another round of golf with Dave.  

 

It's hard to see from the plans, but I have learned many lessons over the years and a visit to your own layout was part of that education.  It's probably not obvious from the plan, but all the inclines to and from the reverse loop and those up to ET terminus are visible and away from the outer edges of the room and the overhanging roof line.  The completely hidden part of the loop is plain track all on the same level as I certainly made that mistake before.  Right now I'm giving a lot of thought to the storage loops and whilst most of them are fine being at least 300mm below ET terminus, I am considering minor changes to the pointwork either end to ensure that all turnouts are in accessible places.

 

I've also realised that one end of the storage sidings will be out of view once scenic sections are added, even with the pointwork easily accessible.  

 

Whilst trains will be easily seen beneath ET terminus, there is a question how far a train is driven down a storage road before clashing with pointwork or the adjacent road.  The back end of the train is no problem, it's just the front end that is an issue.  The easy way of dealing with that is a carefully positioned IRDOT-1 sensor that will provide a visual indication when a train has reached it's forward limit.  As I said, once in position the back end can be easily seen, so double stacking in a long storage road is not an issue.  It brings with it additional cost of around £100 for nine sensors, so moving the turnouts inboard saves that cost but of course shortens the storage roads.

 

The other issue that's going round my head is the width of the track bed for double track.  With lines on 50mm centres and SMP track with a nominal sleeper width of 32mm, you have 82mm plus the width of the ballast shoulder.  I had hoped to secure the trackbed to the risers from the top surface so that trackbeds can be worked on away from the layout and screwed into position once the track had been laid and all the wiring completed to each trackbed.  That would make construction so much simpler, with just each trackbed to be connected to the main bus afterwards.  Of course that means the screws have to be outside the ballast shoulders or they will be hidden from view under the cork underlay and that would prevent my preferred construction method.

 

With regard to the riser that also has to be wider than the screw positions or you end up with screws splitting the riser if they are too close to each end.

 

None of these challenges are insurmountable, but I do find that construction has slowed a little as I consider these issues.  It's so important to get these fundamentals right before going to far or you end up with a lot of scrap material.  Thankfully A4 paper is very cheap and far easier to get rid of than 12mm ply....

 

With regard to duck unders, there is just one and that has 1m clearance.  It's not ideal I know, but I will live with it.  There are only two turnouts the other side of the duck under with all operation etc on the accessible side, so it's hoped the number of times you will need to go under will be few.  A lifting section is virtually impossible with the storage roads and ET terminus above.  Although not ideal, I tried in vain to get the key ingredients I wanted in ET without a central peninsula, but  couldn't find a solution.

 

I would hope to resolve these issues before the weekend and now I know the gradients are fine in terms of length and alignment, I should be able to start tracklaying in a two or three weeks time.

[Rowsley17D]

Hi Gordon, I use IRDOT-1s for my hidden storage roads. The longer ones have two. I wouldn't be without them as you cannot have too much storage capacity. So get the detectors and have the benefit of the extra space. After all, how much is a new club?

[gordon s]

I'm pleased you said that Jonathan as the additional storage space is important to me.  Reading about IRDOT's appear they are easy to install and will work in low light conditions, so it's great to know that your experience has been very positive.

 

I've also has a Eureka moment in the bath on the riser challenges....

 

I had discounted it before, but now having thought it through the answer to my trackbed issues is to have the risers attached to the trackbed module.  I didn't want to go wider than 140mm on the overall trackbed as that gives additional problems with retaining walls and the parapet being too far from the track edges on the upper level.  I can machine risers that sit inside the 6mm MDF side cheeks that I use to ensure my trackbed stays flat between the risers and then fix them to the module with screws underneath the cork underlay.  That will stop the wood from splitting and then the complete module can still be taken in and out by undoing the screws holding the module/riser assembly to each frame.  I will need to check if the screws alone will be sufficient for consistent alignment or whether something additional like a simple alignment dowel will be needed. 

 

I will do some experiments this morning to see if that stacks up, but hopefully it will resolve something that has been bugging me for days now.

 

The longer storage roads are a problem where they cross under upper levels and block the addition of risers in that position.  I'm wondering if threaded studding could be used to solve that issue.  I may need to widen out the radii to accommodate threaded rod and stop it clashing with Mk1's on the curve so it's back to Templot to test out some theory's.  I seem to recall there are some spacing ring issues that Martin had provided in his dummy vehicle option, so that may be useful.

[eldavo]

£100 for 9 irdot sensors! There are much cheaper solutions than off the shelf if the cost is a factor. Waton has infrared sensors to trigger signal changes and I managed to rig them up for around a fiver!

 

Do the screws securing the trackbed to the risers have to be vertical? Couldn't you have a block below the trackbed and screw horizontally in through the riser at the board joins or some such?

 

Cheers

Dave

[martin_wynne]

With regard to the riser that also has to be wider than the screw positions or you end up with screws splitting the riser if they are too close to each end.

 

Hi Gordon,

 

If attaching near the end of a piece of wood, to avoid spitting use a machine screw instead of a wood screw, and glue in (Araldite?) something like this:

 

 http://uk.rs-online.com/web/p/inserts/7031281/

 

Hundreds of similar designs on the web sites, plastic or metal.

 

Martin.

[BoD]

Go on, computerise and automate the whole lot.

You know you want to.

[gordon s]

£100 for 9 irdot sensors! There are much cheaper solutions than off the shelf if the cost is a factor. Waton has infrared sensors to trigger signal changes and I managed to rig them up for around a fiver!

 

Do the screws securing the trackbed to the risers have to be vertical? Couldn't you have a block below the trackbed and screw horizontally in through the riser at the board joins or some such?

 

Cheers

Dave

 

 

Ah, but you're an electronics expert...

 

When and if the time comes, I'll speak to you about them, but start talking technical spec beyond solder A to B and I may struggle.....

[gordon s]

Go on, computerise and automate the whole lot.

You know you want to.

 

Forget to use multi quote.  I did consider that once upon a time, but the placement of detectors became very complicated so I sold the few bits I had to try out.  Let me get something up and running first and then see how that pans out.....

[Junctionmad]

Gordon, I'd ban you from templot and tell you to lay track

 

paralysis by analysis and all that 

 

The goal is to make the little plastic trains go round and round :D

 

dave

[gordon s]

If only it were that easy.....and I am building something.  Just stopped for some lunch.

[Junctionmad]

hi gordon

 

In software engineering,  the accepted approach  until the last 10 years ,was " top down " design.  essentially you built a whole specification before you cut any code.  

 

This was found to be a poor way to design systems as the spec constantly needed changing and was constantly revisited and often the actually coding was pushed further and further away and allocated less and less time, leading to massive overruns and software that didn't meet the customers needs.

 

Today , in software design , we have the combination of "agile" and test driven development , I often call it bottom left hand corner design!.

 

Here you only define small pieces and build them as simply as possible.  The key is to cut code as quickly as possible and get something running , so that the specifiers can see the real thing as quickly into the development as possible ( I'm simplefying this a bit obviously ) 

 

The code  is the minimal required to meet the " test" , at no stage do you add features that are not at this point needed,

 

You then constantly return and refactor the design as it builds, replacing code with better implemented solutions , that still meet the same test 

 

Today large projects couldn't be completed without agile development methodologies , especially whare large teams are used 

 

 

The key takeaway, is that large complex overarching detailed theoretical planning, is actually a hinderance to project completion 

 

you plan a little, build a bit, test, refactor/rework and continue 

 

 

just a useless tuppence worth

 

dave 

[Ian_H]

Hi Gordon,

Just going back to your idea for train detection using the IRDOT-1 sensors, you mention that you have nine roads so would require nine sensors but since you plan on double stacking then presumably you will require more sensors and as you say they are quite expensive. Home made sensors would be cheaper, unfortunately a couple of weeks ago you threw a printer away that had the very sensors that could have been used ... too late now   but there are circuits to build that would do the very thing for you at a much better price and they are relatively simple in construction.

A short while ago I came across some YouTube videos made by Richard W of Everard Junction fame. One of Richards videos deals with Train Detection and Automatic Signalling using a Picaxe micochip. In his video tutorial Richard explains how to program the Picaxe using a software where the user simply designs the operating software using flow charts (if this happens do that else wait this length of time and the other) as opposed to programming languages such as 'C'. Richard explanation is very laid back and I think very very easy to follow, its almost Lego!

I know you're not ready for the system now but if there's nothing on the TV tonight, is there ever, and you have 50mins to spare have a look:

 

 

 

Ian

[gordon s]

Thanks for that Ian.  Just stopped for a cup of tea, so I'll look at that tonight.  Just to make it clear though.  All the trains come in at one end only and it's only the last three feet that are truly hidden.  The remainder of the storage roads sit beneath ET terminus and there the clearance is over a foot, so visibility is not a problem.  I need to run trains in to stop at a pre determined place so that the loco is not fouling either the turnout or the adjacent track as the clearance reduces to go to the same turnout.

 

Once the train has stopped in the right place, the back of the train can easily be seen, so it's only the front stopping point that needs monitoring.

 

Ah, and the printer is still in the hall waiting to go to the tip.  Is it easy to spot the bits that could be useful?

[G.M.R.]

I use a lot of irdots very successfully, but here is an alternative which I am about to try:-

 

http://www.teamdigital1.com/prod_catalogue/dbd22_product/dbd22.html

 

Just have an isolated length of track and feed one side of the DCC to it with the wire passing through the current transformer on this board. No connections needed, except to the LED you want on your control panel. ( and a power supply). £16 for two detectors. coastaldcc.co.uk stock them and will be at the Alton show first weekend in February. I intend going along to pick some up.

 

You can also make them yourself with parts sold by Coastal or with bits from rs-online as shown here:-

 

http://home.cogeco.ca/~rpaisley4/DccBODvt5.html

 

That site also shows you a lot about infrared detection.

 

They have the advantage that they will show occupancy over an entire section, not just at one point. Could be useful for stacking and, of course, are ideal for full automation.

Edited January 20, 2016 by G.M.R.

[G.M.R.]

A short while ago I came across some YouTube videos made by Richard W of Everard Junction fame. One of Richards videos deals with Train Detection and Automatic Signalling using a Picaxe micochip. In his video tutorial Richard explains how to program the Picaxe using a software where the user simply designs the operating software using flow charts (if this happens do that else wait this length of time and the other) as opposed to programming languages such as 'C'. Richard explanation is very laid back and I think very very easy to follow, its almost Lego!

I know you're not ready for the system now but if there's nothing on the TV tonight, is there ever, and you have 50mins to spare have a look:

 

Ian

Interesting video Ian. Automating signalling is a great interest of mine. I've been using hard logic, but am now using an Arduino. One thing I have learnt from the various real signalmen on this forum, is that a green signal must stay green whilst the entire train passes it, so that a guard sees the clear signal, not just the driver. I am now busy changing my software to accommodate this. Means waiting for the detector to see the end of the train, not just its presence. You may be aware that Heathcote electronics make a board which does all the work of Richard's project and includes the ir detector. It will also drive other signals further back down the line. Edited January 20, 2016 by G.M.R.