Charlton Bridge - 4mm BR(S) - Building Bridges

[hayfield]

Ian

 

What I finds works best is

 

First cut the isolation gaps first (before any rails are laid) (check every timber electrically (potentially saves time and trouble later)

Fill gaps with filler and once fully set (24 hours) sand flat. (this is very simple and easy to do) and invisible

Solder rails

 

This method whilst a little more involved at the start potentially saves so much time and the turnout can be tested with a loco whilst building, and if care is taken totally invisible 

 

 

 

 

[ikcdab]

2 hours ago, hayfield said:

Ian

 

What I finds works best is

 

First cut the isolation gaps first (before any rails are laid) (check every timber electrically (potentially saves time and trouble later)

Fill gaps with filler and once fully set (24 hours) sand flat. (this is very simple and easy to do) and invisible

Solder rails

 

This method whilst a little more involved at the start potentially saves so much time and the turnout can be tested with a loco whilst building, and if care is taken totally invisible 

 

 

 

 

Hi John, yes I do pretty much the same as you.  I cut the timbers, gap them, test then fix them to the template.

I use a very shallow half round file to cut the gaps. Two or three strokes and you are through. As it is a shallow half round, the edges of the gaps are tapered and disappear once painted.

Ian 

[Michael Edge]

I cut the gaps close inside one rail, they aren’t so visible there.

[St Enodoc]

5 hours ago, Michael Edge said:

I cut the gaps close inside one rail, they aren’t so visible there.

Unless you pick the wrong rail...

[ikcdab]

Sometimes you really have those "slap your head, duh" moments. Having built my trackwork, i thought it would be an easy job to wire it in.

My turnouts are operated  by servos driven by the MERG servo 4 unit.

In the case on my harbour station, i previously had three routes: loop, main and bay, connected by two turnouts selected by a rotary switch. The routing logic was easy:

route 1: turnout 1: on

route 2: turnout 1: off

route 3: turnout 1 off, turnout 2 on.

I am now adding a fourth route. This requires turnout 1 on and the new turnout (crossover) 3 to be on as well. route 1 then becomes turnout 1 on, turnout 3 off.

Easy, i thought. All i need to do is to wire a diode across the rotary switch terminals such that when route 4 is selected, curent also flows to route 1,  but not the other way.

i spent a whole day fiddling about and just could not get it to work. I tried everything and eveb started to believe that the diodes were faulty.

Then after two days, i realised that servo 4 operates by shortcircuiting the terminals - there is no current involved. All you have do is to connect the servo terminal to the common and the servo throws. So the dioide would never work as there is no current to block (thats as i understand it)...

 

So what do i do now. The ony answer i can think of is to completely rewire this part of the layout such that the rotary switch operates dpdt relays via a diode matrix. One side of the dpdt will complete the circuit to operate the servos, and the other side will set the crossing polarity. I know this works as i have used it elsewhere on my layout. But it will be a major job to remove the switch panel, rebuild it and reinstall. but it will simplify the wiring a bit. 

Unless anyone else knows how you can selectively shortcircuit servo4 boards to provide routing, this is now what I am going to do!

I'll post pictures later.

Ian

Edited December 19, 2023 by ikcdab

[5BarVT]

4 hours ago, ikcdab said:

Then after two days, i realised that servo 4 operates by shortcircuiting the terminals - there is no current involved. All you have do is to connect the servo terminal to the common and the servo throws. So the dioide would never work as there is no current to block (thats as i understand it)...

There is current, it’s just that it flows out of the servo4 terminal not in.  That should just mean putting the diode the other way round.

This is what I think your wiring is (black) and what you need to add (red).
 

If that doesn’t work it will be because the voltage across the diode prevents the input from registering, but at ~0.7V I wouldn’t expect that to be the case.
 

Paul.

Edited June 25, 2023 by 5BarVT
Added diagram.

[ikcdab]

2 hours ago, 5BarVT said:

There is current, it’s just that it flows out of the servo4 terminal not in.  That should just mean putting the diode the other way round.

This is what I think your wiring is (black) and what you need to add (red).
 

If that doesn’t work it will be because the voltage across the diode prevents the input from registering, but at ~0.7V I wouldn’t expect that to be the case.
 

Paul.

Hi Paul, thats really useful, thanks. In my two days of headscratching i am sure that i tried the diodes in every conceivable position and orientation! But i will try again. It may be over-complex wiring elsewhere that is allowing undesirable results - in one instance i just had servo one powering back and forwards continuously until i turned off the power (i know, sounds like an intermittent connection).

When i measure the resistance across the IN4001 diodes, i get around 400ohms in the forward (normal) direction.  WSould that be enough to stop the servo4 inout registering?

Ian

[5BarVT]

4 hours ago, ikcdab said:

When i measure the resistance across the IN4001 diodes, i get around 400ohms in the forward (normal) direction.  WSould that be enough to stop the servo4 inout registering?

It’s only 400 ohms at whatever voltage your ohmmeter puts out!  But that does show it’s not dead.

Looking at the spec for the 16F630 used in the servo4, it seems to be a standard TTL input spec.  That says anything lower than 0.8V counts as Low.

The voltage across a forward biased silicon diode is nominally 0.7V so should be sufficiently low, but looking at the spec for 1N4001 it seems to be only just under 0.8V so quite close to tolerance.  I’m not sufficient of an electronics expert to say whether a diode input to TTL works or not.

So it could be the diode that is causing the problem, but might not be.

 

If your servo4 is accessible, there is a fairly easy way to test.  Short from i/p 4 to 0V to check that the input works the servo.
Then do the same with a diode.  Nothing will happen if it’s the wrong way round, so just try it both ways.  If nothing happens either way round then it looks like the diode is not pulling diwn the input far enough to be recognised as On.

 

Paul.

[ikcdab]

12 hours ago, 5BarVT said:

It’s only 400 ohms at whatever voltage your ohmmeter puts out!  But that does show it’s not dead.

Looking at the spec for the 16F630 used in the servo4, it seems to be a standard TTL input spec.  That says anything lower than 0.8V counts as Low.

The voltage across a forward biased silicon diode is nominally 0.7V so should be sufficiently low, but looking at the spec for 1N4001 it seems to be only just under 0.8V so quite close to tolerance.  I’m not sufficient of an electronics expert to say whether a diode input to TTL works or not.

So it could be the diode that is causing the problem, but might not be.

 

If your servo4 is accessible, there is a fairly easy way to test.  Short from i/p 4 to 0V to check that the input works the servo.
Then do the same with a diode.  Nothing will happen if it’s the wrong way round, so just try it both ways.  If nothing happens either way round then it looks like the diode is not pulling diwn the input far enough to be recognised as On.

 

Paul.

Thanks.  I have tried that and I had intermittent results.  First time round I got no reaction from the servo at all (with diode tried in both orientations).  Second time I did get reaction from servo, but then it stopped.

I guess the diode is giving a marginal input that is right on the threshold, so is not reliable.

Ian

[5BarVT]

5 hours ago, ikcdab said:

I guess the diode is giving a marginal input that is right on the threshold, so is not reliable.

I fear that might be the case.

It could be done with a single pole relay but it would need protection to avoid back emf getting into the pic and frying the input.

I assume this is Foxhole Quay not the main station?  If so, then I think the easiest way is to change the switch for a 2 pole version and wire it thus:

That keeps the two inputs separated without involving any diodes or relays. Uses a 2 pole 6 way in place of single pole 12 way.

 

Paul.

[ikcdab]

Hi Paul, that's a good solution and one that I hadn't thought of. I can't use it though because I am already using a two-pole rotary switch. One side operates the servos, and the other side operates the relays for crossing polarity. So what I will do is to rebuild the switch panel with dpdt relays.  I can energise these through a diode matrix operated from a single pole switch on the panel.  Then one side of the dpdt relay can run the servo and the other side can switch the crossing polarity. It's what I have done elsewhere and I don't know why I did this one differently!

I have started making this. Here's the detailed drawing and the embryonic panel.

R3 and R4 (and R8 and R9) are  crossover pairs so the two turnouts are linked at the servo4 board, hence only 1 relay needs the dpdt function for the servo.

R6/R7 is a turnout and catch point so only one lot of polarity changing needed!

 

 

Tomorrow I'll finish the wiring on this panel.

 

Ian

 

 

[ikcdab]

Realised I haven't posted on here for ages.

I'll update on the new siding soon. In the meanwhile I have been building a new pair of block bells to add to the existing ones.

 

First I made the cases.  Just ordinary B&q stripwood stained and varnished.

 

Then next was the base and tapper.  Tapper is from 6mm square brass strip, drilled and tapped:

 

Then the striker assembly.  Again, 1" steel angle.

 

 

Bell and bell pillar:

 

 

Winding then coils using the lathe on the slowest setting. 

 

 

The finished innards:

 

 

And one of the bells in position by the mainline control panel:

 

 

More to follow soon!

 

Ian

 

Edited December 24, 2023 by ikcdab

[John Besley]

Very nice looks very much like parts of the SR mainline to Exeter... is it possible to reload the missing photos?

 

Noted your comments on heating, I have used two tubular oil filled radiant heaters in my garage / workshop for the last 20 years running them 24/7 executive in the hight of summer, they keep paperwork dry and tools rust free

[ikcdab]

12 hours ago, John Besley said:

Very nice looks very much like parts of the SR mainline to Exeter... is it possible to reload the missing photos?

 

Noted your comments on heating, I have used two tubular oil filled radiant heaters in my garage / workshop for the last 20 years running them 24/7 executive in the hight of summer, they keep paperwork dry and tools rust free

Hi John, I'll see if I can find the missing pictures.  I need to do an update anyway.

Whatever I am doing with the shed is working as it's remained bone dry for the last three years despite all the rain we have had!

[ikcdab]

Its been a long time since I did any updates, but we have not been idle. Since January we have been concentrating on scenics. I had been reluctant to do this, I am not sure why, but my daughter firced me to start and we have not looked back. Now i have sat down ti write this, I realise I have not taken many suitable pictures, but here is one looking round the sweep of the lines to Charlton Bridge Junction.

 

More soon

Edited March 24 by ikcdab

[ikcdab]

At the moment my focus is on the viaduct. The span is about 1.5m with a depth of about 200mm. Originally I had planned for this to be a standard brick arched viaduct of which there are numerous real life examples. I erected some plywood piers with the intention of covering with brick paper. This idea morphed into a girder bridge on brick piers and I spent a lot of time designing and then 3d printing the girders. This again gave me issues as the 3d printing warped and I found it difficult to get nice, straight girders. That was a long story and changing the type of resin in the printer helped hugely - but I was not 100% convinced I had the right look for the bridge. One of my issues was that I had laid a 6mm plywood track base and traditional sleepered track, whereas what I really wanted was the unballasted "baulk road" effect you see on girder bridges. I had also made the wooden piers too narrow and i needed to add extension pieces to widen it.  All in all, I was not happy.

Then a friend visited and threw another massive spanner into the works saying that I had too many piers...

So yet another rethink. 

 

This is the (rather embarrasing) current situation with the abortive 3d printed girders (all now scrap!), the ply track bed and the ply piers. 

 

 

My new (and final answer) to this is to rip the whole viaduct section out and replace it with a structure based on Little Petherick Creek bridge on the North Cornwall LIne. Many pictures here.

 

I had amazing help from fellow RMWeb members who sent me drawings and pictures of the bridge. A lot of hours later I have drawn the structure in Fusion360 and i am now ready to start printing. Although based on the original bridge, that was single track and my bridge will be double line and my piers are much taller. But it will resemble the original. 

here are the piers drawn up and ready to go. I will have a trial print tomorrow and see how it goes. Luckily I only need two sets - each one is going to take 13 hours to print!

 

 

 

 

I am still looking for pictures and details of the bridge if anyone else can help!

Ian

[ikcdab]

And here are all of the main span components laid out.

For printing I had to break it up into sections, so i will print the end girders as a unit, then the central girder and join together. 

 

I do have an outstanding question though. At the top of the cast iron piers are square bearing plates which the main girders are sat on - you can see in the picture below. What is unclear from any pictures or drawings is whether these are two separate pieces (as per my drawing) or whether it is infact a single piece spanning the piers. Both answers are problematic. If its two separate pieces, then how are the ends of the deck supported...if its a single piece then how is that supporting itself? As i wont be able to print anything now til after easter, i might take the opportunity to drive down and have a look - its not that far from me. 

In the meantime, if anyone has any insights, I would be pleased to hear them!

Ian

[ikcdab]

Having just posted that, I realise that the diagional braces on the main girder should lean outwards! As it is, I prefer the look of what I have done and I'm not keen to change now. I think that what I have done will still be structurally sound. I might change it, we will see, it depends how much it nags me.

[ISW]

1 hour ago, ikcdab said:

In the meantime, if anyone has any insights, I would be pleased to hear them!

Ian

Ian,

 

I was always under the impression that there is a cross-member between the 2 outside trusses located at the base of each vertical member of the truss (and at each end). There is also generally some horizontal diagonal members between these cross-members to increase the stiffness.

 

For example, see:

http://www.mvlbridges.co.uk/pony-truss-girder-bridge.html

and

https://www.instructables.com/How-to-Design-a-Basic-Truss-Bridge-With-Fusion-360/

 

Ian

 

 

 

[ikcdab]

9 minutes ago, ISW said:

Ian,

 

I was always under the impression that there is a cross-member between the 2 outside trusses located at the base of each vertical member of the truss (and at each end). There is also generally some horizontal diagonal members between these cross-members to increase the stiffness.

 

For example, see:

http://www.mvlbridges.co.uk/pony-truss-girder-bridge.html

and

https://www.instructables.com/How-to-Design-a-Basic-Truss-Bridge-With-Fusion-360/

 

Ian

 

 

 

Hi Ian, yes, there has to be a cross member of some sort. I have just looked at this picture. You can clearly see the cross members and if you look closely, i think you can see a further one right up by the concrete pads. 

Ill have a play on fusion 360 and see if I can add it in. The concrete pads at 8ft square, and the girders rest equally on them, so it means that any cross girder would still be 4ft from the ends of the deck.

A visit is called for!

[ISW]

9 hours ago, ikcdab said:

The concrete pads at 8ft square, and the girders rest equally on them, so it means that any cross girder would still be 4ft from the ends of the deck.

Ian,

 

There has to be a 'back' or 'end' span between the outermost piers and the adjacent embankments. Length would vary depending on the topography but not very long. It would be supported on the embankment by a small retaining wall.

 

Ian

[ikcdab]

So success with printing the piers. Both sets now made.  One I've cleaned up, the other needs to wait it's turn. Really pleased with these.

 

 

[stevel]

Wow, just discovered this thread and have enjoyed reading through it. I love the railway in the landscape feeling you have captured. Look forward to seeing the progress.

 

Stephen 

[Fishplate]

9 hours ago, stevel said:

Wow, just discovered this thread and have enjoyed reading through it. I love the railway in the landscape feeling you have captured. Look forward to seeing the progress.

 

Stephen 

 

I've just finished reading through as well👍. Very nice layout.

[ikcdab]

11 hours ago, stevel said:

Wow, just discovered this thread and have enjoyed reading through it. I love the railway in the landscape feeling you have captured. Look forward to seeing the progress.

 

Stephen 

 

2 hours ago, Fishplate said:

 

I've just finished reading through as well👍. Very nice layout.

Thanks to both.  The bridge is coming on well will post details soon