Buckingham West

[Richard Mawer]

Presumably you could arrange the circuit so that when the responding tapper is used you'll put a low resistance across the capacitor which would drain the capacitor pronto and extinguish the LED. That means that you could probably have a very high value capacitor to give you a long LED lit time without worrying that the LED will still be lit long after the bell signals have been exchanged for a following train movement.

 

And on that very note - excuse the pun - looking at things from the other direction so to speak, the LED would be lit when exchanges were the other way round. Could this be confusing?

 

Thanks Ray. Another good idea!

 

Being lit with the exchange the other way is not an issue. The problem lies when you are concentrating on something else and a bell goes, which one of the three is it? Any indicator being lit during an operation is fine.

[GWMark]

Presumably you could arrange the circuit so that when the responding tapper is used you'll put a low resistance across the capacitor which would drain the capacitor pronto and extinguish the LED. That means that you could probably have a very high value capacitor to give you a long LED lit time without worrying that the LED will still be lit long after the bell signals have been exchanged for a following train movement.

 

And on that very note - excuse the pun - looking at things from the other direction so to speak, the LED would be lit when exchanges were the other way round. Could this be confusing?  

 

 

I would be tempted to use a NE555 timer in monostable mode. That way the lit duration would be tuneable to some fairly decent times without the need of a big capacitor and you could also use the outgoing bell taper on the reset line of the 555 to turn the LED off immediately. The 555 as well as being less power consuming than charging the capacitor, would also be less expensive probably.

 

My other thought was to have an Arduino in there that counted the bell taps an understood the difference between the different codes - particularly train out of section. That has the potential of also being able to indicate track occupied in the interval between "train in section" and "train out of section" - although this would not be true occupancy detection, but it could add a little more to things. However this would of course be dependant on the operators getting it right - so no good for me 

[RJS1977]

Whilst it probably wouldn't be worth doing it just for the block bells, if you had sufficient CBUS units with spare inputs (CANACE8C) and outputs (CANACC8), you could operate the block system that way:

 

Box A tapper making contact - ON event Box B LED and bell. OFF event for box A LED

Box A tapper breaking contact - OFF event for Box B bell but LED remains lit

Box B tapper making contact ON event for Box A LED and bell. OFF event for box B LED

Box B tapper breaking contact - OFF event for Box A bell but LED remains lit.

[MarshLane]

My other thought was to have an Arduino in there that counted the bell taps an understood the difference between the different codes - particularly train out of section. That has the potential of also being able to indicate track occupied in the interval between "train in section" and "train out of section" - although this would not be true occupancy detection, but it could add a little more to things. However this would of course be dependant on the operators getting it right - so no good for me 

 

 

I have something similar to this working on a PIC processor, which interfaces with the DCC system, if that helps?  The PIC code uses a timer to work out the bell code (i.e. the gap between 2-1 on TOS being the splitter) and then sends a DCC decoder ON/OFF signal as required, which the computer pics up.  Far more complicated that you need Rich, but added to show that it is possible!

 

Rich

[Richard Mawer]

Wow!

 

Here's me thinking of an H&M point motor, a piece of wire and a piece of white card visible (or not) through a hole in a piece of wood. When all along, I could be using a capacitor, an NE555, an Arduino, CBUS CANACC8, or a PIC Processor. Silly me!

 

Whilst I do know basically what each of those are, I have no idea whatsoever, how one wires them up.

 

I am very endebted to all you guys for the interest and suggestions.

 

So, I can work out how the capacitor and LED idea works. One small problem is that the tappers which operate the bells at Evenley are all on negative side of the bells. To make it work, I think the tapper needs to be on the positive, but could reverse the polarity of the whole bells, tappers, buzzers system easily enough.

 

I'm sorry but I am not going down the route of CBUS just for bells. This is a DC layout and the only electronics are MERG SERVO4 boards for the servos. I like simple/old school. Similarly Arduino and PIC Processors that count beats etc are very clever and would be great as a virtual second operator (automatic Crispin huh Tony!), but part of the fun here is for the operator to deal with the bells. The timetable already says what to expect when.

 

The real issue seems to be possible confusion as to which bell rang call attention. Once in a bell conversation, it is fine. So an indicator as to which of the 3 bells just rang call attention is the issue.

 

That said, having an indicator of TOL would be very handy. In fact indicators of Line Clear and then TOL would prevent lapses of memory. Hang on, we are reinventing the block system and instrument! I can't see why there should be such indications automatically from the bells when I easily make some block instruments. But that is all too cumbersome for now. Maybe in the future.

 

So, my choice would have to go with Mark's NE555 approach.

 

The big issue is, I have no idea where to start, so is that an offer to make 3 Mark? The bells are on 12v DC not 5v.

[MarshLane]

Here's me thinking of an H&M point motor, a piece of wire and a piece of white card visible (or not) through a hole in a piece of wood. When all along, I could be using a capacitor, an NE555, an Arduino, CBUS CANACC8, or a PIC Processor. Silly me!

 

 

And you thought it could be complicated!!! Lol!  As a totally random thought - would the principal that they used to use (and maybe still do) at Shrewsbury Severn Bridge Junction SB be any good?  They had a piece of string with a weight (I think a button) attached to the bell somehow, so that it 'jiggled' for a few seconds after the bell rang - if the signalman was busy or didn't pick up which one, he just looked along the shelf for which piece of string was swinging!  

 

Nice and simple

 

Rich

[MarshLane]

Similarly Arduino and PIC Processors that count beats etc are very clever and would be great as a virtual second operator (automatic Crispin huh Tony!), but part of the fun here is for the operator to deal with the bells. The timetable already says what to expect when.

 

Me again Sorry I forgot the PS .. which was to say, wasn't suggesting you went down the PIC route - as you know with me, its mainly me operating, and the boxes either side won't be modelled so its a virtual signalman (don't tell Network Rail of that idea!) - but was just showing it could be done if needed!  I usually find the KISS (Keep It Simple Stupid!) principal is always the best and most effective!

Rich

[t-b-g]

And you thought it could be complicated!!! Lol!  As a totally random thought - would the principal that they used to use (and maybe still do) at Shrewsbury Severn Bridge Junction SB be any good?  They had a piece of string with a weight (I think a button) attached to the bell somehow, so that it 'jiggled' for a few seconds after the bell rang - if the signalman was busy or didn't pick up which one, he just looked along the shelf for which piece of string was swinging!  

 

Nice and simple

 

Rich

 

I like that!

 

A friend of mine had a layout with many stations and bells and he had a paperclip attached loosely to bells when there were several together. I always thought it was purely a modellers trick but I am delighted to see that it had a prototype.

[Richard Mawer]

I like that!

 

A friend of mine had a layout with many stations and bells and he had a paperclip attached loosely to bells when there were several together. I always thought it was purely a modellers trick but I am delighted to see that it had a prototype.

I love the ideas, but my bells are out of sight, below the baseboard at Evenley.

[RJS1977]

 

 

I'm sorry but I am not going down the route of CBUS just for bells. This is a DC layout and the only electronics are MERG SERVO4 boards for the servos. I like simple/old school. Similarly Arduino and PIC Processors that count beats etc are very clever and would be great as a virtual second operator (automatic Crispin huh Tony!), but part of the fun here is for the operator to deal with the bells. The timetable already says what to expect when.

 

 

 

And I wouldn't have expected you to. I couldn't remember whether you were using CBUS already (it's not just for DCC) and I know from various ideas I've sketched out that the number of outputs required to be worked from MERG kits rarely matches the number of channels available, so most CBUS applications wind up with some channels left over.

 

But I'm working on exactly those lines for having a CBUS-operated 'Digital Crispin' on my own layout once it's been resurrected.

[MarshLane]

I love the ideas, but my bells are out of sight, below the baseboard at Evenley.

 

 

Ah!  Well as Cleggy used to say on Last of the Summer Wine .. "Its just one damm snag after another!"

 

This may be a daft thought, but going back to one of the previous suggestions, using the electric feed to the bell solenoid, could you not tap off that through the appropriate resistor and capacity to an LED on your control panel?  If the LED stayed lit for 4 to 5 seconds, then extinguished as the capacitor emptied, is that not the simplest way around the problem?  Avoids the need for timers etc...

 

Rich

 

Rich

[Richard Mawer]

Ah!  Well as Cleggy used to say on Last of the Summer Wine .. "Its just one damm snag after another!"

 

This may be a daft thought, but going back to one of the previous suggestions, using the electric feed to the bell solenoid, could you not tap off that through the appropriate resistor and capacity to an LED on your control panel?  If the LED stayed lit for 4 to 5 seconds, then extinguished as the capacitor emptied, is that not the simplest way around the problem?  Avoids the need for timers etc...

 

Rich

 

Rich

Yes Rich, it is the simplest and it does whats necessary.

 

So is there a nice kind soul out there who is willing to send me a quick diagram of the circuit and to suggest a value for the capacitor and the resistor to go in front of the LED? Does the capacitor charge from just the initail "call attention" ping? That's a very short connection!

 

Or have I missed the point?

 

That's why I thought the 555 timer was a good idea.

 

Rich

[MarshLane]

Yes Rich, it is the simplest and it does whats necessary.

 

So is there a nice kind soul out there who is willing to send me a quick diagram of the circuit and to suggest a value for the capacitor and the resistor to go in front of the LED? Does the capacitor charge from just the initail "call attention" ping? That's a very short connection!

 

Or have I missed the point?

 

That's why I thought the 555 timer was a good idea.

 

Rich

 

Im not particularly knowledgable on capacitors, but I would have thought that a short burst of electric, could be enough to part charge a capacitor that would lite an LED for a few seconds.  LEDs draw so little current.  I could quite easily be wrong tho!

 

Rich

[5BarVT]

Paul,

Thanks for your input. My layout is indeed GWR, 1930s. You learn something every day! I didn't know the TOS was not repeated. That makes life a little easier anyway.

Thanks also for the advice on the cap and LED. I am sure a fading one would be fine.

Cheers

Rich

Being 'brought up' on WR signalling, it was only when I visited Banbury North after decommissioning that I learned that other regions did reply to TOS (40 years between the two events)!

 

I will be able to draw/design a circuit for you but I'm a bit pushed for time over the next few days. Just so that I know what I'm working with, I assume that each bell circuit is a separate wire (on the real thing both ways worked over the same wire to save on pole route copper) with the +12 connected to the bell and the tapper making the connection to the -12. (There's a bit of WR signalling circuit pedigree in using a positive common rather than common negative too!). I'll need to try a typical capacitor value to see how big is needed for 5 sec delay. Others are right it could be quite big and a 555 would allow a smaller capacitor, but the bell coil will push nasty spikes into the 555 which it might not like.

Paul.

[Richard Mawer]

Being 'brought up' on WR signalling, it was only when I visited Banbury North after decommissioning that I learned that other regions did reply to TOS (40 years between the two events)!

I will be able to draw/design a circuit for you but I'm a bit pushed for time over the next few days. Just so that I know what I'm working with, I assume that each bell circuit is a separate wire (on the real thing both ways worked over the same wire to save on pole route copper) with the +12 connected to the bell and the tapper making the connection to the -12. (There's a bit of WR signalling circuit pedigree in using a positive common rather than common negative too!). I'll need to try a typical capacitor value to see how big is needed for 5 sec delay. Others are right it could be quite big and a 555 would allow a smaller capacitor, but the bell coil will push nasty spikes into the 555 which it might not like.

Paul.

Paul, I too went round Banbury North box in the summer, but its about 12 miles from me. You are in Glasgow?

 

I don't have the ability to draw on here and am only using an ipad. But there is a 12v +ve feed to Evenley which is connected to all three bells (and all 3 tappers. We can igore those tappers because the operate bells elsewhere of course). A single wire from the other terminal of each bell goes to the tapper at each of the three other stations. Those tappers are connected to the -ve power feed (0v presumably). So there are 3 circuits in parallel. Each with its own discrete wire between bell and tapper. Each circuit is + bell tapper -. I can easily swap over the power to have 0v connected to the bell and +12v connected to the tapper because the 12v dc is run off an old H&M Duette. I just need to turn the knob the other way!

[Richard Mawer]

THE MATRIX - PART 3

 

The 4 wiring looms are added. Blue to Down dial. Orange to up. The 2 black and reds to the point motors. Red for normal and black for reverse.

 

 

 

The dials on the control panel at Banbury select the road. You then push the red square button to fire the CDU.

 

 

 

I have chosen route selection so I do not have to think in terms of tracks, points and storage loops. I just want the train to go into the tunnel and "onto the rest of the system". Route selection just following a letter on the timetable is close as I can get..

 

I will fix the matrix to a baseboard leg and then I need to work out which coil on each motor makes that point normal or reverse. The motors are fitted where space permitted. They are not all facing the same way.

 

[Stubby47]

A bit late to this issue with which bell has rung, but would different tone bells help ?

[Richard Mawer]

A bit late to this issue with which bell has rung, but would different tone bells help ?

 

Ha ha! Well after reading all the posts on LEDs and timers etc. you can be easily forgiven for thinking that they must all sound the same. The fact is that they aren't. They are different tones, but for a relative newcomer to the layout, with quite a lot going on, and one's concentration on something else (such as shunting etc), and with the bells out of sight, the different tones can be hard to differentiate between. Especially when one just does a single ping out of nowhere.

 

I tried to get them as different as I could find and get to work. One is a fairly small (higher pitched) church type bell, one is a lower pitched cow bell and one is a mid pitched bar (like a glockenspiel).

 

I did try to get a really small and high pitched bike bell to work, but I could only get a clunk. The beater has to bounce off it far faster than I could manage.

 

I can tell the difference normally, but then again I am normally operating Buckingham West when I have guests. When I have trained someone else up on Buckingham, I will be able to see how my ears do when operating with others.

 

Thanks

[Richard Mawer]

Stu (Stubby47),

 

Thank you for being my 40th follower on here. I never expected my project to have more than about 2! So I am grateful for the interest and for all the ideas and advice from everyone.

 

I hope you enjoy the ride Stu!

 

BTW, I may have to employ some your building expertise soon. I have some drawings of a Brunel train shed to go over the platforms at Buckingham West. They were kindly donated to me through RM Web, but I cannot see me round to building it for a long time yet.

[The Stationmaster]

A bit late to this issue with which bell has rung, but would different tone bells help ?

 

The old answer was chewing gum - nowadays it is probably blutac, a nice little 'slug' of it inside the bell to alter the tone.  The other trick is to have a fair idea of which one it was that rang and feel it as there will still be a trace of vibration - works a treat with Western bells but probably not so easy with those abysmal LMR plastic Triang like block instruments with built-in bells.

[GWMark]

Sorry Rich, I've been distracted and not followed the conversation.

 

The big issue is, I have no idea where to start, so is that an offer to make 3 Mark? The bells are on 12v DC not 5v.

 

 

Yes, more than happy to do that. I'll have a play and see if I can get something to you.

 

Mark

[Richard Mawer]

Merry Christmas everyone.

 

Friday evening and yesterday I was given special dispensation and have been running the layout singled handed with the clock. I can easily manage running at a clock speed ratio of 4:3 or 1.33x. In other words 24 model hours every 18 hours real time. I'm aiming for 1.5x.

 

With group running we have been running at 2:1 or 2x quite easily. I had hoped for 3x.

 

I do find it very satisfying, but as someone warned me, I am at a dangerous point - starting to enjoy operation rather than continuing building. I still haven't fitted the matrix.

[Ray H]

Belated Christmas wishes to you Richard and all the best for 2017.

 

I think you'll find that 3x will be possible once things becomes second nature to people. We have a schedule for the club's layout that normally takes about an hour to run through but we've got down to forty minutes by the time the same team have run it for three times consecutively.

 

Something else you may find is that you'll start to lose the keenness to operate solo after a fairly short while and will seek out other tasks to do whenever you can't muster a team for a session - or that may just be me!

[Richard Mawer]

There's one heck of a lot of movements in this timetable! Not sure we'll ever get to 3x, but it can be an aim.

 

Today I finally remembered to run-in the re geared Dean Goods in reverse! I have been meaning to for weeks.

 

I also added more lead to the pony truck of one 43xx (they were both playing up the other day) and got it behaving much better. However, I spent an hour on the other one with far less success. I have been here before with these locos (see earlier entries), but then I was trying to increase the traction. Now I want them to stop derailing. I ended up with so much weight on the second truck it was lifting the front drivers off the rails - with a distinct lack of traction.

 

If I take the pony off, it runs fine. If I put it on, it tends to derail on the points and crossings at Charlton Junction.

 

Rightly or wrongly I think it isn't so much about jumping, as the pony being pushed/being allowed, to go the wrong way at the points - sideways, and then derailing.

 

There seems to be a lot of side play in the drivers. The body can twist with all the drivers still on the rails. Maybe if the body is twisting one way, it pushes the pony over to the wrong route at the frog. Could it? So I am thinkng of gluing some plastic strip onto the sides of the chassis, betwwen the front drivers and the chassis, effectively making the chassis wider. This will hopefully reduce the amount of play on the axle. Then doing the same at the rear. The central axle will be left free to slide more from side to side to get round the curves in the dumbbell. Hopefully that will mean the loco is more consistent in pointing directly down the track, rather than being able to waggle. Does that sound logical?

 

My second thought is to use some wire to create some sort of centre springing of the pony. If I can keep it in line with the centre line of the loco body (but with some play aganst the spring to turn a bit for curves), then that might help.

 

I should add that neither pony truck has any springing from the loco chassis or body. Some seem to. Mine don't.

 

Apart from new chassis, any other ideas?

Edited December 26, 2016 by Richard Mawer

[Richard Mawer]

43xx FIXED (I know I have said that before!)

 

We have been away for a couple of days and I was thinking about these 43s. I became more convinced the possible waggle due to side-play between the driving axles and chassis could well be the cause of the pony truck going astray. If the loco body could twist say, clockwise, left over the front axle and right over the rear) then the fixing of the pony would move left forcing the pony to face slightly right and possibly take a right hand divergence. And vice versa.

 

I decided to pack out the gap between the driving wheels and the chassis block to prevent possible twists/waggle. There was quite a gap each side. Not having suitable plasticard I searched for a suitable thickness of plastic and came up with a plant pot label. The sort you write on and stick in a pot of seeds. I cut it in half length-wise and slid a piece down each side behind the drivers. There was just a tad of play left on the axles.

 

 

 

I tried it and hey presto, no derailing! I then tried it round the dumbbell and it ran round that too.

 

I shortened each piece so it slotted in. They are not glued. They fit above the axles and in indents in the chassis. There is no way of sliding out. Body back on and its like a different loco!

 

I might just shave it thinner behind the centre wheels to allow a bit more play for curves, but its not a big issue. It would be on tighter curves.

 

I need to paint them black, but overall I'm really pleased with such a simple solution.