Heath Town and other signalling diversions

[Andy Keane]

Paul, I would be interested to know what all the boards do, who supplied them and what you thought of them as a user - there is quite a lot going on there!

regards

Andy

[5BarVT]

Hi Andy,

 

The heart of the system is the board with the thin red outline - an RR-Cirkits Watchman board.  It’s an 8 way train detection board with another 8 way i/o. (There should be an 10 way ribbon cable from it to the top yellow outline board - oops!).  The train detection uses current transformers which I have chosen to combine on one board - the thick red outline at the bottom. I missed a trick when I converted from stripboard to PCB: with stripboard I was limited to 0.1” components so went for an 8 way latching header strip which I have to crimp individual terminals on each wire.  With the PCB I could have used an RJ45 connector and cut ethernet patch cords in two.

 

My reason for going to PCB design was that it saved the hassle of cutting strips and adding jumpers.   My reject rate on getting strip cuts in the wrong place was not good!
 

The left hand yellow board is a MERG DCT2 which has the current transformer on the board.  I wanted 9 sections on this board so I added these extra two and then used both of them ‘because they were there’.

 

The orange board is a home brew H bridge driver with opto isolator to keep clean and dirty supplies separated, again previously built on stripboard.  This one is only 50% populated as there is only one point on this board.

The purple outline is a voltage divider - separate as I couldn’t find space for it on the stripboard driver - it generates the mid voltage so that the op amps switch as the input go from high to low.

 

The yellow i/o board is just a multi purpose connector bank. The 10 way ribbon has Watchman +5V, 0V and 8 configurable i/o, the converter board can take 8 x 2way i/o + 0V, or 4 x 3 way double i/o or +5/0V and separate i/o depending on the application connected.  So the MERG DTC2 needs 0/5V and has two inputs coming back whereas the point board has a three way input for two drive signals. Detection comes back as a single input from the tortoise to the i/o.

 

The four way connector bottom left brings DCC (white) and dirty 12V (black) onto the board.  12V goes to the point drive voltage splitter and on to the point drive board. DCC goes to the current transformer board where it is split to the 8 separate sections - all the white sheathed 2c running all over the board.

 

The data connection (black 3c) runs on the opposite side of the board away from the dirty stuff (I’m probably being over pessimistic, it’s just a design standard I’ve adopted).  The RR-Cirkits boards run on a modified Loconet data network - an interface from loconet adds power so I don’t need a separate clean power bus round the layout.  Because i have a  Digitrax command station I can use the cheap interface, there is a slightly more expensive one for a stand alone system.  I have that on my test system.

 

One reason why I went for the RR-Cirkits system was the distributed nature of the boards - one watchman is just about right for most of my layout boards.  They also do a Motorman - 8 slo-mo point drives and 8 i/o, and a Signalman = 16 direct drive LED outputs which can have brightness adjustments, can flash etc.and a Towerman - 16 i /o.

 

I’ll be using the signalman on the scenic section for . . . signals! and also for the indications in my frame.  Tower man for the switch inputs on the frame and Motorman where I have concentrations of points.  Others use Arduinos to make loconet compatible I/o at a much cheaper rate per input than toweman, but obviously more work (you pays yer money . . .).  Watchman is probably the most cost effective  as 4 x DCT2 starts to add up!

 

Programming is JMRI - another plus for me and there doesn’t need to be a linkage between loconet ID and board i/o pin.

 

The importer is Coastal DCC - I have always had excellent service from Kevin.

https://www.coastaldcc.co.uk/products/rr-cirkits/

 

Paul.

[St Enodoc]

9 hours ago, 5BarVT said:

Another reference photo.  Final bit of wiring done this afternoon isn’t visible as I went for longitudinal rail bonds to the 1” section at the board joint.

 

Paul.

For bits of track that short I just solder the fishplates.

[Andy Keane]

8 hours ago, 5BarVT said:

Hi Andy,

 

The heart of the system is the board with the thin red outline - an RR-Cirkits Watchman board.  It’s an 8 way train detection board with another 8 way i/o. (There should be an 10 way ribbon cable from it to the top yellow outline board - oops!).  The train detection uses current transformers which I have chosen to combine on one board - the thick red outline at the bottom. I missed a trick when I converted from stripboard to PCB: with stripboard I was limited to 0.1” components so went for an 8 way latching header strip which I have to crimp individual terminals on each wire.  With the PCB I could have used an RJ45 connector and cut ethernet patch cords in two.

 

My reason for going to PCB design was that it saved the hassle of cutting strips and adding jumpers.   My reject rate on getting strip cuts in the wrong place was not good!
 

The left hand yellow board is a MERG DCT2 which has the current transformer on the board.  I wanted 9 sections on this board so I added these extra two and then used both of them ‘because they were there’.

 

The orange board is a home brew H bridge driver with opto isolator to keep clean and dirty supplies separated, again previously built on stripboard.  This one is only 50% populated as there is only one point on this board.

The purple outline is a voltage divider - separate as I couldn’t find space for it on the stripboard driver - it generates the mid voltage so that the op amps switch as the input go from high to low.

 

The yellow i/o board is just a multi purpose connector bank. The 10 way ribbon has Watchman +5V, 0V and 8 configurable i/o, the converter board can take 8 x 2way i/o + 0V, or 4 x 3 way double i/o or +5/0V and separate i/o depending on the application connected.  So the MERG DTC2 needs 0/5V and has two inputs coming back whereas the point board has a three way input for two drive signals. Detection comes back as a single input from the tortoise to the i/o.

 

The four way connector bottom left brings DCC (white) and dirty 12V (black) onto the board.  12V goes to the point drive voltage splitter and on to the point drive board. DCC goes to the current transformer board where it is split to the 8 separate sections - all the white sheathed 2c running all over the board.

 

The data connection (black 3c) runs on the opposite side of the board away from the dirty stuff (I’m probably being over pessimistic, it’s just a design standard I’ve adopted).  The RR-Cirkits boards run on a modified Loconet data network - an interface from loconet adds power so I don’t need a separate clean power bus round the layout.  Because i have a  Digitrax command station I can use the cheap interface, there is a slightly more expensive one for a stand alone system.  I have that on my test system.

 

One reason why I went for the RR-Cirkits system was the distributed nature of the boards - one watchman is just about right for most of my layout boards.  They also do a Motorman - 8 slo-mo point drives and 8 i/o, and a Signalman = 16 direct drive LED outputs which can have brightness adjustments, can flash etc.and a Towerman - 16 i /o.

 

I’ll be using the signalman on the scenic section for . . . signals! and also for the indications in my frame.  Tower man for the switch inputs on the frame and Motorman where I have concentrations of points.  Others use Arduinos to make loconet compatible I/o at a much cheaper rate per input than toweman, but obviously more work (you pays yer money . . .).  Watchman is probably the most cost effective  as 4 x DCT2 starts to add up!

 

Programming is JMRI - another plus for me and there doesn’t need to be a linkage between loconet ID and board i/o pin.

 

The importer is Coastal DCC - I have always had excellent service from Kevin.

https://www.coastaldcc.co.uk/products/rr-cirkits/

 

Paul.

Thanks, very interesting. Have you thought of writing this up as an article in the MERG journal. I am sure many would be interested. Minor question: why did you not use the MERG 8-way detector? Is the Watchman board a better solution for you? I have some very similar stuff but in the end use Arduinos for much of what I do. Do you plan any train automation?

regards

Andy

[5BarVT]

7 hours ago, St Enodoc said:

For bits of track that short I just solder the fishplates.

I thought about that, and about who would give that advice! :-). Only reason I didn’t is because the adjacent track is a fixed set track piece and I didn’t want solder in the fishplate area. In theory that means it could be reused on another layout.  Illogical, cos it won’t be!

Paul.

[5BarVT]

40 minutes ago, Andy Keane said:

Minor question: why did you not use the MERG 8-way detector? Is the Watchman board a better solution for you? I have some very similar stuff but in the end use Arduinos for much of what I do. Do you plan any train automation?

Opposite way round for the reasoning!

I have TrainController for automation and needed a control system on its list of supported interfaces.  MERG CANBUS is not on that list.  That led me to a Loconet based system because it has a wider support base (mostly from the US rather than Europe) and I settled on RR-Cirkits.  The Watchman board has more sensitivity adjustment than the MERG kits, although I’ve not needed any yet as everything I detect is loco motors, although that is intended to change in the fullness of time.  I think (at the time) that an 8 way MERG plus the 16 way I/o was more than the Watchman board - and would require construction time!

Paul.

[5BarVT]

On 07/09/2023 at 22:31, 5BarVT said:

(There should be an 10 way ribbon cable from it to the top yellow outline board - oops!).

One from the recoveries pile only a smidgen too long is now in place, routed slightly differently to take account of the extra length.

Paul.

[Clive Mortimore]

On 07/09/2023 at 22:31, 5BarVT said:

Hi Andy,

 

The heart of the system is the board with the thin red outline - an RR-Cirkits Watchman board.  It’s an 8 way train detection board with another 8 way i/o. (There should be an 10 way ribbon cable from it to the top yellow outline board - oops!).  The train detection uses current transformers which I have chosen to combine on one board - the thick red outline at the bottom. I missed a trick when I converted from stripboard to PCB: with stripboard I was limited to 0.1” components so went for an 8 way latching header strip which I have to crimp individual terminals on each wire.  With the PCB I could have used an RJ45 connector and cut ethernet patch cords in two.

 

My reason for going to PCB design was that it saved the hassle of cutting strips and adding jumpers.   My reject rate on getting strip cuts in the wrong place was not good!
 

The left hand yellow board is a MERG DCT2 which has the current transformer on the board.  I wanted 9 sections on this board so I added these extra two and then used both of them ‘because they were there’.

 

The orange board is a home brew H bridge driver with opto isolator to keep clean and dirty supplies separated, again previously built on stripboard.  This one is only 50% populated as there is only one point on this board.

The purple outline is a voltage divider - separate as I couldn’t find space for it on the stripboard driver - it generates the mid voltage so that the op amps switch as the input go from high to low.

 

The yellow i/o board is just a multi purpose connector bank. The 10 way ribbon has Watchman +5V, 0V and 8 configurable i/o, the converter board can take 8 x 2way i/o + 0V, or 4 x 3 way double i/o or +5/0V and separate i/o depending on the application connected.  So the MERG DTC2 needs 0/5V and has two inputs coming back whereas the point board has a three way input for two drive signals. Detection comes back as a single input from the tortoise to the i/o.

 

The four way connector bottom left brings DCC (white) and dirty 12V (black) onto the board.  12V goes to the point drive voltage splitter and on to the point drive board. DCC goes to the current transformer board where it is split to the 8 separate sections - all the white sheathed 2c running all over the board.

 

The data connection (black 3c) runs on the opposite side of the board away from the dirty stuff (I’m probably being over pessimistic, it’s just a design standard I’ve adopted).  The RR-Cirkits boards run on a modified Loconet data network - an interface from loconet adds power so I don’t need a separate clean power bus round the layout.  Because i have a  Digitrax command station I can use the cheap interface, there is a slightly more expensive one for a stand alone system.  I have that on my test system.

 

One reason why I went for the RR-Cirkits system was the distributed nature of the boards - one watchman is just about right for most of my layout boards.  They also do a Motorman - 8 slo-mo point drives and 8 i/o, and a Signalman = 16 direct drive LED outputs which can have brightness adjustments, can flash etc.and a Towerman - 16 i /o.

 

I’ll be using the signalman on the scenic section for . . . signals! and also for the indications in my frame.  Tower man for the switch inputs on the frame and Motorman where I have concentrations of points.  Others use Arduinos to make loconet compatible I/o at a much cheaper rate per input than toweman, but obviously more work (you pays yer money . . .).  Watchman is probably the most cost effective  as 4 x DCT2 starts to add up!

 

Programming is JMRI - another plus for me and there doesn’t need to be a linkage between loconet ID and board i/o pin.

 

The importer is Coastal DCC - I have always had excellent service from Kevin.

https://www.coastaldcc.co.uk/products/rr-cirkits/

 

Paul.

Whooosh, that was close but luckily went over my head.

 

2 wires for each track at both ends, so 16 going into a 25 way D plug at the left hand when looking at the photo and 14 track wires and 3 wires to the point motor into another 25 way D connector, but this time a socket at the right hand end. And DC is lots of wires.

 

As for a watchman, that is me with my Mk1 eyeballs.

[5BarVT]

Hi Clive,

 

Main take away for you is that DCC with train detection sections is more wires than simple DC.  Only 2 wires? Pah!

 

Paul.

[5BarVT]

On 11/09/2023 at 23:04, Clive Mortimore said:

Whooosh, that was close but luckily went over my head.

That’s twice I’ve nearly got you so far.  You need to start worrying, one might hit!

:-)

Paul.

[Clive Mortimore]

10 minutes ago, 5BarVT said:

Hi Clive,

 

Main take away for you is that DCC with train detection sections is more wires than simple DC.  Only 2 wires? Pah!

 

Paul.

Hi Paul

 

I bet you that even with your train detection you will have a "Oh bu**er, I didn't want that one to move" moment.

[5BarVT]

2 hours ago, Clive Mortimore said:

Hi Paul

 

I bet you that even with your train detection you will have a "Oh bu**er, I didn't want that one to move" moment.

Not as many as you might think (might hope?! :-) ).  And indirectly because of the train detection.

The train detection is there to allow automation and I tend to operate using that rather than a ‘controller’ (hand held, or on screen) so not only do I know what I’m trying to move (by number) but also where it (allegedly) is.  That means two thinkgs have to be wrong before I make a mistake and the ‘diversity’ does help reduce errors.
Ooh Butter moments have usually been the opposite - I’ve run something on the controller and forgotten to release the loco then part way through the next automatic move the controller sends out a Speed 0 command and the train stops dead.  That’s had me head scratching a couple of times now.

Paul.

[5BarVT]

HELP REQUEST

 

I have moved the track on my station throat to improve its distance from the baseboard edge (clash with spring clip for holding the boards together).  Unfortunately that means that a point motor now clashes with the board bracing . . .

 

Green is normal position - no chance!  Option 1 (red) is the offset operation position - still a bit tight.

So option 1a is to put a greater offset step in the tortoise operating wire - has anyone tried that?

By chance, I came across the MTB MP10 tortoise equivalent and purchased 2 to trial on a little used crossover (in case they weren’t as successful as I’d hoped).  The same bracing clash occurs as with the tortoise as they are deliberately the same footprint.  However, they come with a fairly long operating rod (with vertical adjustment on the machine).  I’m wondering if I could put 2 right angle sets in the operating rod to allow an offset drive with the MP10.

 

Anyone tried that?

 

TIA, Paul.

 

P.S. Option 3 is to cut a hole in the bracing to allow the MP10 to fit under - possible due to the low height so not an available option with the tortoise.

Edited September 14, 2023 by 5BarVT
P.S.

[Nick C]

What about a remote mount - something like this: https://www.digitrains.co.uk/800-6100.html

[5BarVT]

41 minutes ago, Nick C said:

What about a remote mount - something like this: https://www.digitrains.co.uk/800-6100.html

Thought about that after I posted.  I’ll need to see what / how much space it needs at the point end.

Paul.

[Andy Keane]

Paul, could you try a woodworking solution were you re-route the bracing perhaps?

[5BarVT]

43 minutes ago, Andy Keane said:

Paul, could you try a woodworking solution were you re-route the bracing perhaps?

Not easily, though not impossible.  I suspect that one of the other options will be successful before I get to complete brace options.

Paul.

[5BarVT]

SOMETHING IS STIRRING

Plan in place for marking out . . .

 

and some track has appeared. Four turnouts reused from Heath Town v4.  A change of wiring policy (now Back to Black, previously the opposite) means I had to decide whether to remove and refit the droppers . . .  there will be a warning label underneath saying that the four to the left are wired opposite polarity.  Normal labelling is black on white, the warning will be black on yellow to stand out.  The two points on the right being new will be wired to the new convention.

It’s taken a long time, but it’s good to be making progress on the layout.

Paul.

Edited September 15, 2023 by 5BarVT
Spelling

[St Enodoc]

9 hours ago, Nick C said:

What about a remote mount - something like this: https://www.digitrains.co.uk/800-6100.html

 

8 hours ago, 5BarVT said:

Thought about that after I posted.  I’ll need to see what / how much space it needs at the point end.

Paul.

If that doesn't work, you could try a Cobalt, which is physically smaller, with/without a right-angle drive adaptor.

 

Edited September 14, 2023 by St Enodoc
added photo

[St Enodoc]

2 hours ago, 5BarVT said:

 A change of wiring policy (now Back to Black, previously the opposite) means I had to decide whether to removed and refit the droppers . . .  there will be a warning label underneath saying that the four to the left are wired opposite polarity.

I think would change the droppers and in your heart of hearts I suspect you think the same...

 

As you are well aware, an engineering solution is always preferred over a procedural one.

[5BarVT]

1 hour ago, St Enodoc said:

I think would change the droppers and in your heart of hearts I suspect you think the same...

 

As you are well aware, an engineering solution is always preferred over a procedural one.

You’re right, I know what I should do.  I’m also lazy!.

 

 

1 hour ago, St Enodoc said:

If that doesn't work, you could try a Cobalt, which is physically smaller, with/without a right-angle drive adaptor.

I wasn’t aware of the Cobalt RA drive adaptor.  Not sure if it would fit - I think it’s only about half the clearance in your photo, but pushed right up to the bracing it might work.  Thank you for the idea.

Paul.

[St Enodoc]

5 minutes ago, 5BarVT said:

You’re right, I know what I should do.  I’m also lazy!.

 

 

I wasn’t aware of the Cobalt RA drive adaptor.  Not sure if it would fit - I think it’s only about half the clearance in your photo, but pushed right up to the bracing it might work.  Thank you for the idea.

Paul.

https://www.dccconcepts.com/product/cobalt-right-angle-adapters-with-mounting-hardware-3-pack/

 

Usual disclaimer.

[5BarVT]

18 hours ago, Nick C said:

What about a remote mount - something like this: https://www.digitrains.co.uk/800-6100.html

A further option has been discovered.  A home made remote mount using MDF and the WFRM laser cutter.

Watch this space and thank you for the inspiration.

Paul.

[5BarVT]

DRIVE BAR v1

75x16.5mm, Drive from tortoise goes in LH slot drive to point comes through RH hole both with 90 deg bend.  Green is a clamping piece with engraved slot to hold the point wire in place.  Blue slots allow the whole thing to slide around No.4 pan head screws into the underside of the base board.

If it’s too sticky or too sloppy or just binds, something more complicated will be designed.

 

Paul.

[5BarVT]

On 15/09/2023 at 00:22, St Enodoc said:

I think would change the droppers and in your heart of hearts I suspect you think the same...

 

As you are well aware, an engineering solution is always preferred over a procedural one.

On 15/09/2023 at 01:38, 5BarVT said:

You’re right, I know what I should do.  I’m also lazy!.

What I really need is an engineering solution that does not require unsoldering and resoldering droppers.

 

IKEA PYSSLA to the rescue.

 

Paul.