3 Dec 2009
Electrical Pick-Ups, Occupancy Detection and LEDs
As part of my grand plan, I need to get good occupancy detection of trains on the layout. Achieving this is easy for locos - the Digitrax BDL168s referred to in earlier posts see the loco current draw and know it is in the relevant section - however the occupancy detection really needs to see the rear of the train as well as the loco at the front. Therefore I decided as a test to try and convert a brake van to trigger the occupancy detection.
While I am at it, I will also try and add some LEDs to the brake van to provide some lighting effects, in particular I want to try some of these fantastic looking oil lamps from DCCConcepts.
I started with a Bachmann 33-306A toad brake van.
Now the two aims - lighting and occupancy detection - both require electrical pick up from the track - this is not that easy at least for me it isn't. By default Bachmann insulate both wheels where they join with the axles - obviously if there was no insulation then the wheel / axle combinations would short the two rails out.
I intended to use the DCCConcepts pickup srings which fit over the axles but these need one of the wheels to be electrically connected to the axle. My initial attempt was using silver paint
but i just couldn't make this work - the paint is a suspension of silver and even after repeated attempts i couldn't get any electrical connection from the wheel to the axle. So on to plan b - more expensive - but more reliable perhaps. Replace the wheel sets for ones that have an metal bush on one wheel thereby making the axle 'live'.
One company that advertise live axles are Markits
I ordered two Markits wheel sets the 12mm for the brake van, and the 14 for a Bachmann Mark 1 coach - that I wanted the same treatment for.
2 X MRJ12TSNSL1 (12mm 10 spoke n/s live 26mm Axle)
2 X MRJ14BR1-1L (14mm 3' 7" contoured n/s coach disc live 26mm Axle)
These were not cheap, by the time postage was included these four axles cost 15 pounds. Worse the van really needed 13mm wheels and putting the smaller ones on it, meant that now the van was too low for the kadee height test...
Still I carried on anyway, thinking that if the project worked I could always go for some larger wheels later - the model came with 13mm wheels but Markits only do 12 or 14mm - and fitted the spring takeoffs to the axles ... you need some calipers to reset the wheel axle widths after you have done this.
The I soldered some insulated wire to the springs and covered the join with some heatshrink. So the underside looked like this -
To get the wires through the chassis I needed to drill a hole in the floor of the brake van, and to do this I had to prise the body off - the body is glued to the (black) chassis, and the body has a (grey) floor that is also glued in place and can also be (carefully) prised out. We also need to get into the body later to place various bits of circuitry and the interior LED. Here is the body separated from the floor prior to drilling...
Next came a very simple occupancy test - if you place a 10K resistor across two rails then that is sufficient to fire the BDL168 to show occupancy, but far too small to short the track or cause any significant current to flow. So next I tookthe two wires exiting from the top of the chassis and connect to a 10K resistor with some croc clips and voila my Heath-Robinson looking device is detected by the BDL168s.
Now having got power to above the chassis in the brake van - i need to work out what circuitry i need to both power the two very different LEDs and to make the occupancy work.
Now DCC is an odd sort of supply as it is squarewave AC - and so i asked Pearson(elder) for advice - you can see his response here (broken link).
So from that I believe that I need to adopt the following circuit. The choice of 10K for the locolamp is not from any theory but because the resistors supplied in the pack from DCC concepts seemed to range from 5kohm to 20kohm - so obviously they expect a very low forward voltage.
I lashed up a crude test of this with some components and sure enough it triggered the occupancy detector and lit LED 1 but not LED 2 ( the locolamp).
A bit more messing around led me to the following answer...
If like me you are a bit unused to identifying diodes then really anything that isn't a zener is likely to work - ideally look for a minimum breakdown voltage of > 50V other than that go for a nice small size that will fit in your model!
I built this circuit on a little bit of veroboard cut from a larger piece bought at maplin and using a yellow LED from Rapid that I had in stock - yellow to simulate an oil lamp lighting the inside of the vehicle.
Here is the circuitboard mounted on the van base - with the led being lit...
And even more satisfying here it is both stopping an oncoming loco and turning the associated signal to red - via the CML SIGM20 I mentioned earlier in the blog.
Disappointingly though the connection via the rails up to the circuit board is too poor to work well for occupancy detection. The led flickers all of the time as the van circulates on the track - i could fix this using a capacitor circuit like flickerfree (also from DCCConcepts) but this won't help with the occupancy issues. If the computer and/or signals see the occupancy going on and off they will get confused. A bit of messing about shows the problem is not the spring pickups on the axles but the fact that the two live wheels aren't always in contact with the rail.
Accordingly I am going to have to try and get eletrical pick up from the other two wheels on the van and probably also weight the van down more. This will mean fiddling about with nickel silver springy wire - not something i am looking forward to.
So in the meantime I have been messing with the locolamp. These are really incredibly small.
They come with enamel coating over the wires for insulation. This is not easy to remove - I finally did it with a combination of a match to burn the enamel off and a piece of emery cloth.
Here is the hole drilled on the van for the red oil liamp light...
More when I try and connect a pickup to the other two wheels.
17 Sep 2009
Block Detection Visible on Computer
Here is the first screen showing a live block occupancy detection on demo version of RR&Co software.
As the loco enters the section the block turns from yellow to red and as it leaves it turns back to yellow again.
Now I just need to try putting three separate detection sections into a single logical block.
13 Sep 2009
Block Occupancy Detection
A key part of the infrastructure on the planned railway is 'block occupancy detection' - a set of circuitry that allows you to know if part of the track is occupied.
For this I am using the BDL 168 from Digitrax. It is suitable for the other loconet devices that I am running For example it tells the DTM30 that a section of track is occupied via loconet avoiding lots of long wires for LED displays, it also interacts with SIGM20 signal controller to automatically control signals, and finally it provides vital information to the computer to allow it to automagically run trains, and hopefully avoid crashes.
It took me sometime to understand the wiring diagrams, they are highly ambiguous - but for my simple test setup you just connect Zone Common wire to the main Rail A feed. then for the section to be detected remove the Rail A wire from wherever it was previously connected and join to any of the 16 detection section pins on the DBL connector. And that is it.
Learning points - I will need to plan lots of rail gaps and feeds into the layout before construction to allow sectioning.
Here is the board and connector mounted on a scrap piece of hardboard :-
And here is it working for the first time, the top LED indicates that the class 25 at the back of the shot is 'in section'.
And finally here is the full array of electronics I am currently using -
From the left we have a Team Digital SMD 82 static controller that is operating the one point motor, then the Digitrax controller and PSU and then comes the box containing the CML DTM 30 and finally the DBDL168. I still have to add a SIGM20 and replace the SMD82 with a CML DAC 10.
8 Sep 2009
Signalling Panel
I have been extravagent and bought three new boards from CML electronics. The first to be implemented is a DTM30 board,
which allows traditional style control of points and (reporting of signal and occupancy states) but using very modern techniques.
In a "traditional" panel you push a switch and it activates a point motor, this is OK, but it has several drawbacks -
Firstly, it means you have lots of wires flowing from the panel to the various motors.
Secondly, it means if, like me, you want to be able to control routes both by computer and from the panel then that is not possible without conflict.
Using this CML board, then there are no direct wires to the point motors instead the board outputs and inputs packets to the Digitrax loconet bus. These instruct the point motors to change state this is then "seen" by the computer and so it is aware, it is also seen by the
static decoder on the point which actually activates the point. Similarly if the computer or "throttle" gives a command to change state of a point that will be picked up on the bus by the DTM30 which can track the change and change the indicator light.
This all means that you only need three connections from the board to the outside world - two loconet connections and a power supply.
I used a maplin £4.99 ABS box, combined with some Hexaganol plastic spacers (12mm) to mount the board on off the floor of the box and a power supply left over from a BT broken home hub that gives up to 2A at 9V DC. I then connected that to a Chassis mount 2.1mm DC socket (FT96E) so that the whole box could be unplugged.
On the top of the box, I have used some 5mm green HB LEDs (CK39) mounted on 5mm LED Clips (YY40) and a sub miniature push switch (JM01B).
Like everything else this is just for test at the moment so I have not made the box too pretty.
I purely want to test out the concepts for the main layout. As so much wiring is involved I want to be sure what I am doing before I start.
30 Jul 2009
(Link for this specific entry...)
6 Mar 2009
New Test Layout
I have now moved into the new room - hoorah, and I have taken my single board layout and am working on turning it into a "tail chaser". I intend t use this layout to do a dummy run before I build a more seroious layout.
In particular I want to check out track laying (using glue rather than pins) and ballasting techniques, and most importantly I want to work on the electrical fit out to allow computer and panel control as well as throttle control.
Here is the current progress, with the cork layed.
2 Mar 2009
Team Digital SMD 82
- Can operate a pair of points from a single output
- Operates on a switch transition rather than level
- LED driven onl from tortoise?
- poor to operate tortoise?
21 Feb 2008
Heljan Western Disassembly
There are no retaining screws for the main body moulding.
But it is hard to grip anywhere - DO NOT grip the chassis by the fuel tanks as these are only held on by four skinny plastic clips and it is very easy to break them ( I speak from experience). 
The best disassmbly mechanism is to insert two thin sheets of soft bendable plastic one on each side below the drivers door and lift one end of the body up using these pieces of plastic as levers.
DCC fitting - the pin 1 diagram is not that obvious on the instruction sheets - here is a picture that should hope make it clearer. I used a LENZ silver, the silver direct would NOT fit under the body
PIC}:-insidewestern.jpg;
Kadees - I used number 19s directly on the NEM sockets. They seemed to work fine.
20 Feb 2008
Bachmann Mark 1 Disassembly
As part of a longer term plan to add lighting to the coaching stock, I needed to dismantle a Bachmann Mk 1 coach.
As always this was more fiddly, and stressful than one would like. As models have got more detailed, they have also got more delicate. So here for my failing memory, as well as anyone else out there - here is my guide to dismantling Bachmann Mk 1s.
1. Remove Water Pipes
Remove the water pipes on each end they can just be pulled out with pliers or tweezers, if you just remove the lower fixing they can be gently bent clipped around the side of the coach.
2. Remove Bogies
Turn coach upside down and remove each bogies by unscrewing the centre screw.
3. Remove Coupling Assemblies
There are clever, but delicate little coupling assemblies below each bogie, these can be carefully unclipped.
4. Unscrew the base
Three screws can now be seen and removed. and the seating slid from the chassis.
When you come to reassmble, please note that the middle screw is not in the centre, so the base and the floor need to be lined up - if you put one in the wrong way round, the middle screw can not be done up.
PIC}:-bogie one way.jpg;
Also note the bogies fit on with the 'gap' pointing towards the coupling socket.
26 Dec 2007
Bachmann Standard 5MT disassembly / DCC fitting
Whilst I like this model enormously and it is equipped with a DCC socket - it is very tricky to carry out the fitting without damaging the plethora of delicate features on the model.
This is made unnecessarily worse by the lack of any good disassembly instructions....I hope this blog post will help others as some instructions turn a tricky prospect into a relatively simple one...it will also help me recall what to do when I have to it again.
To separate the body from the chassis requires three screws to be removed
Firstly the piece of the puzzle that I wish I had known was to remove the screw connecting the con rod to the rear driving wheel. This allows the connector from the con rod to the body - presumably this was the speedo connection on the protoype?...if you don't remove this then you end up like this....
The second screw to release is the countersunk screw at the front of the loco, indicated by the arrow in the picture below...
The last screw to remove is the hardest to find / get at , as it is hidden below what presumably are the brake actuating mechanism...but again my knowledge of the prototype is so limited that I am only guessing...In the picture below then I have labelled four components -
(1) brake rod running down the centre of the loco, is just clipped in place, can either remove completely or just unclip from the component marked '3' below and bent gently to one side
(2) can't be unclipped just have to bend out of the way.
(3) a 'nodding donkey' shape of component which is easily lost - clips each side and then has a locating lug holding it in place in the chassis.
(4) the screw we are trying to get at. Can't be seen on the next photo, look at the one below that.
OK now you can separate the two halves of the loco, let's turn to the DCC decoder, I have used a LENZ silver 8 pin with plug. The only thing to note is that to fit in the nicely designed slot for it to go in, you should trim any
heatshrink back tight to the edge of the little circuit board as here:-
And here is a picture of the controller fitted into the loco - as you can see the controller is stood vertically in a slot designed for it mounted in the smokebox area.
When reassembling the loco there is a plate that fits under the cab that will have come out when you separated the body from chassis. Here it is below...
The plate fits back with the semicircular bit pointing to the front of the loco and the hole in that fitting over the lug highlighted in the next photo...
