Wednesday, December 21, 2022

Have a Merry Christmas and a Happy New Year

Christine and I would like to wish everyone a merry Christmas and a very happy New Year. May all that you hope for come about.

This would have to be the longest break between posts for the blog. I have almost sat down to type several times, but life has become a bit complicated in the last six months. I have managed to do a couple of things on the layout as well as make some signals for others but the main situation is around a now eight-month-old puppy. The last pup we had was in 2003 and it is amazing how you forget how much work a new pup is. We have just about got it toilet-trained and it seems to be starting to quiet down so things should improve soon. It certainly cuts the day into pieces of time where I can work, or alternatively, look after the pup when Chris is having a break from doing it. However, all considered, it is a lovely little thing.

I hope to get back to posting more often next year.

Sunday, June 12, 2022

Australian Made Servo Control Boards for DC and DCC

As we now know the past two years have disturbed the normal supply of many things. Among those have been my goto servo control boards from Tam Valley Depot. I have used Tam Valley Depot Singlet IIs, Tam Valley Depot Quad-Pics and a Tam Valley Depot Dual 3-Way board. Tam Valley Depot's owner Duncan McCrae stopped production due to ill health and at about the same time, COVID-19 hit the supply of electronic parts. Tam Valley Depot has since recommenced production but only a small part of its range. All of the above servo control boards are no longer being made.

As a result, I began a search for alternatives. I found suitable servo control boards in the UK as well as one in the US but given the cost of shipping, these become expensive.

I came across a local Sydney supplier, RailCon with an inexpensive range of boards in both DC and DCC versions.

At the time I was designing and printing linear servo signal bases for some US-made Tomar brand Upper Quadrant signals after an enquiry from a modeller here in Australia. Now the modeller didn't want three positions for the signal arms so I told him about the 2 position RailCon MD-003 boards so he purchased them for his signals and had them sent to me.

USA Tomar brand Upper Quadrant Signals with RailCon MD-003 Servo Slave Units

The RailCon MD-003 boards arrived and I found them quite easy to adjust the linear servos to get the correct throw of the signal arms but the speed of the throw was too fast. I called RailCon and spoke to the owner, Jim Crew about the issue. Jim said the boards were originally designed to throw point blades so the speed wasn't an issue then., however, he said he would have a look at slowing the speed. After a couple of days, Jim said he had done it so I sent the RailCon boards to him to be updated. They arrived a week back or so later and the speed was good. This speed has now become the speed for the boards in the RailCon range as it won't have any effect on the point blades since they only move a small amount.

The RailCon MD-003 Servo Motor Slave Unit will drive 2 servos and is priced at $24.60 which is $12.30 per servo output. The board requires a 12 Volt DC power supply.

For use on a DC layout, the MD-003 is used as is but for DCC the $24.85 SD-001 Central Control Unit is required. The SD-001 board connects to 4 SD-009 Servo Motor Slave Units. This gives $15.40 per DCC servo output.

While talking to Jim I asked about the possibility of a 3 position servo board for our NSWGR Upper Quadrant signals and he said he would look at it. About two weeks later Jim said he had a 3 position board and could I test it, the answer of course was yes.

Here is a video of a HO NSWGR Upper Quadrant signal arm being moved by the 3-position servo board. As I didn't have any Upper Quadrant bracket signals under construction I set up the Upper Quadrant arm on a Lower Quadrant timber post from my range.

As you can see the arm moves through the correct sequence going from CLEAR (vertical) to STOP (horizontal) in one movement then From STOP to CAUTION (45 Degrees) and then finally back to STOP. A three-position switch can be used to change the signal but the software has been written to ignore the centre switch position when moving from CLEAR to STOP. Of course, other switching methods can be used such as DCC, track detectors, relays, etc.

The MD-003a Servo Slave Unit is the 3-position Upper Quadrant version and is $24.00. The MD-003a is for DC and only controls one servo as the existing MD-003 printed circuit board has been used with three input connections instead of the four inputs (2 pairs) of the MD-003 unit. This was the simplest and least expensive way for RailCon to give us a 3-position option.

Although the website hasn't been updated as yet there is a DCC 3-position version, the SD-009a Servo Motor Slave Unit, The  SD-001 Central Control Unit will support 4 SD-009a Servo Motor Slave Units to control 4 servos in total.

One of my NSWGR Steel Bracket Signals with
Upper Quadrant and Lower Quadrant signal arms

I hope you have found this information useful and I think that RailCon should be supported as it is a very good local product.

Saturday, March 26, 2022

Wollar Roundhouse Completed Finally

Way back in June 2011 and July 2011, I did several posts on my build of an Anton's Roundhouse kit (click on the links above or see the old posts list to the right and down).

Well, I finally decided that I needed to finish it, only 10 years late.

The problem that stopped me originally was the large rear windows that were wrong, or at least based on a plan and some photos I have of the type.

I measured the existing window frames of the model and worked out that I needed to modify them to get the windows anywhere near correct. The problem is that the roundhouse is designed to go with the Anton 75 foot turntable and the turntable has incorrect spacing(read angle) of the radiating tracks. This angle difference means that the rear walls of each stall are wider than they should be and this makes the proportions of the windows in the rear wall look wrong. It also became apparent that some of the rear wall framing was incorrect. I worked out what I needed to do to make things look proportionally correct even if not dimensionally so.

Cowra Roundhouse rear windows showing proportions

I turned my computer on and drew up some replacement window frames and the two different size windows in Sketchup for 3D printing.

This view shows that the model framing for the windows is quite wrong

The smoke vents that I made from various Evergreen strips and sheet can be seen in the above photo. These days I would of course draw them up and 3D print them. It is amazing how much 3D printers have added to my enjoyment of the hobby.

After printing a frame and some windows I checked how they might look. Apart from being wrongly placed the large louvres above the centre of the windows then looked too tall.

Incorrect framing cutaway and printed frame and windows in place

The large louvre supplied with the kit was then replaced with a shorter 3D printed louvre.

All 3D printed frames, windows and new smaller louvres are in place.

The windows and window frames were printed on a filament printer with PLA filament and a 0.4mm nozzle and a 0.1mm layer height. Here are the STL files for those who wish to use them to improve the Anton's roundhouse. The link points to a folder with the STL files. Please note that there are two small window STL files as the 0.4mm nozzle made slight variations in the window frame, a function of the printer trying to get the frame dimensions correct but needing to accommodate the nozzle size I guess. The difference is only tiny but it can stop the wider small window from fitting into the frame. It is best to print one each of the wide and narrow small windows first then try them in the window frame to work out how many of each size you will need. 

I spray painted the roundhouse 'timber' frame with a very dark grey and brush painted the windows and lower roundhouse wall with off-white.

After painting, I was able to clad the roundhouse with Evergreen corrugated sheet. It used up quite a lot of the sheet that I had been collecting over the intervening years. I found that each piece of sheeting had to be measured carefully to fit as no wall or roof section of the roundhouse seemed to be exactly the same size as another notionally equivalent section.

A light grey was applied to the walls and roof cladding. The edges of the various corrugated sheets were drawn with a fine point pencil to show each corrugated iron sheet. The roof corrugated iron sheets were delineated with black pastel powder and a brush to dirty up the roof. Further weathering will be applied with rust pastel powder until I am satisfied with the look.

After all the work on the rear windows, the roundhouse is placed such that the rear can't be seen, but I know they are looking a lot better and especially through the front of the roundhouse.

Tuesday, January 18, 2022

Casula Hobbies Z19 1954 - Modification of the Marker Lighting

My Casula Hobbies Z19 arrived a couple of days ago, a lovely little model and it runs really well. A big well done to Joe and his technical team.

Casula Hobbies Z19 1954 fresh out of the box

I had ordered 1957 but received 1954 for some reason, perhaps too many had ordered 1957 and I missed out. I don't mind though as 1954 is the same Baldwin tendered version with cut-out cab and cowcatcher. Given the COVID-19 situation, I wasn't up to taking it back or going to the Post Office to send it. Christine and I had our booster shots today and we won't be getting out and about until two weeks after.

I am not a fan of Loksound although I do have four Loksound 4 decoders in some steam locos. I decided that I would take the Z19 with the Loksound 5 sound option as I investigated the Loksound 5 Micro Next 18 decoder and determined that there were two extra power outputs that I could use to improve the marker lighting.

ESU LokSound 5 Micro DCC Next18 output diagram

It is obvious why the marker lights were set up the way they came as the soldering of the extra wires is perhaps not a great idea for a production model as it restricts the decoder from being a simple plug-in for the modeller who wants to add sound later.

Now, the exploded schematic of the model that is supplied in the box shows that there are four screws but they are mostly hidden beneath the ends of the front and rear bogies.

Warning, do not do what I did. I unscrewed the bogies at their pivots and then tried to move the bogies to clear the body mounting screws. The bogies have pickup wires through the tender floor and can't be moved far. I did manage to just clear the rear bogie to unscrew the rear body mounting screws but in trying this at the front a disconnected one of the bogie chains. Now, the chain is glued over a short small peg on the end of the bogie side frame and it was a nightmare glueing it back when I had finished.

The best way to access the tender body mounting screws is to carefully spread the bogie side frames and pop the axle out that is over each screw position. Don't bend the side frames out too far as the axle ends run in a metal inner frame for the electrical pickup.

Anyway, back to the reason for accessing the two extra outputs. The marker lights on the model are able to be switched on and off and are independent of the headlight(s). The model marker lights are designed to have white at the front and red at the rear when moving forward and the reverse of this when moving backwards. This is not prototypical when pulling a train as the red lights should be off.

Here is a photo of the underside of the Loksound 5 Micro Next18 decoder showing the additional output solder tabs which are AUX5 and AUX6. There are three other outputs but they are logic level and will not power a LED without some additional circuitry.

The underside of LokSound 5 Micro SCC Next18 decoder
showing the AUX5 and AUX6 outputs

The orientation of the decoder in the above photo is not the same as the ESU diagram as I decided to take the photo of the decoder in the same orientation as it is in the tender just flipped over.

In the following photos, I decided to use the number 3 output on the circuit board at the front and the number 2 output at the rear of the tender as they were easier to access with the soldering iron. These equated to the front and rear white marker light LEDs.

The white wire attached to Output 3 must be unsoldered from the board as it will be soldered to a 1KOhm 1/4 Watt resistor which has another piece of white wire soldered to it. This white wire is then soldered to the AUX5 tab on the decoder.

Output 3 white wire disconnection spot on circuit board

The yellow wire attached to the rear yellow surface mount LED will be unsoldered from the board at the R5/C8 connection in the lower-left corner next to the speaker. The photo below shows the yellow wire already unsoldered from the R5/C8 connection spot on the circuit board. I forgot to take a photo before I unsoldered the yellow wire.

Output 2 LED and  yellow wire disconnection spot on circuit board

The yellow wire is then soldered to a 1KOhm 1/4 Watt resistor which has another piece of yellow wire soldered to it. This yellow wire is then soldered to the AUX6 tab on the decoder. 

Please note that all white wire solder joints, yellow wire solder joints and resistor leads have 0.6mm heat shrink insulation tubing applied.

Extra marker light wiring and resistors in place

After re-wiring, I now have each pair of white and red marker lights on a separate function button. So the marker lights are a bit more prototypical.

  • White marker lights at the front when and red markers on the tender when running light (locomotive by itself). The reverse when running tender first of course.
  • White marker lights at the front when on a train and no red markers on the tender.
  • Red marker lights at both ends when shunting.

When testing the marker lights I found that if the red markers are on and you turn on the white markers then the white dominates the red and you can't even see a tinge of the red.

Once the re-wire was complete I weathered 1954 and added a crew.

I used two different greys, a darker and a lighter one and followed this with a light application of black pastel powder to the tender bogies as I felt that they were too light and clean. I then used some cream coloured pastel powder to add a spilt sand effect to the sandboxes area just in front of the cab on both sides. Finally, I added a very light application of rust coloured pastel powder to the lower half of the smokebox door.

Here is a link to a JMRI DecoderPro file I made for the Z19 default CVs. Before I did any of the above I read all the CVs into DecoderPro and also did my best to work out what sounds are on what function numbers (see the Function Labels tab in DecoderPro). I could be wrong for a couple of them so would be happy to be corrected.

EDIT: Here is a link to a just-released PDF document about the Casula Hobbies Z19 that includes the default function list.

One final note, I was disappointed to find that there is no short whistle for use when reversing, three short whistles. There is however a function button that will play four different long whistles, very strange!

Some whistle files can be programmed to allow a short or long whistle to be played depending on how long the whistle button is pressed but this isn't one.

I hope that the LokProgrammer sound file for the Z19 will be made available as that missing short whistle is very annoying.

If the file is made available then I will have to borrow a LokProgrammer from a friend and add a short whistle.

Overall, once again I can only say congratulations Joe Calipari.


I received a comment question about the coal load and if I have any intention of doing something about it, here is my reply:

"The short answer is yes. The longer one is that I hope to do something about a keep-alive but there isn't much room in the tender. The model coal load is quite flat and it is removable, well I hope it isn't glued in too strongly.

My idea is to remove the plastic load and investigate how much room can be found with a 'taller' load. The keep-alive will have to be a homemade one as the supercapacitors may need to be spread around inside the tender and coal load.

Then there is the final issue.... I have to find my supply of coal. I have obviously put it somewhere other than where it was usually kept."

Thursday, December 23, 2021

Have a Merry Christmas and a Happy New Year


I hope everyone has a safe and enjoyable Christmas and New Year.

Sunday, October 10, 2021

A Home for the BYLONG Layout Circuit Breakers

Several years ago the layout gained a mysterious short circuit even though I hadn't been doing any wiring or trackwork recently. After a lot of searching under the layout and checking the track for closed rail gaps, I called in the big gun, Marcus Amman.

Marcus spent an afternoon with me checking all the same things and we still came up with nothing. At this point, it was decided that I needed to replace the NCE EB3 circuit breaker board as the EB3 boards were known to be a bit flaky. So, three NCE EB1 circuit breakers were bought and temporarily installed in the wiring. Now what I mean by temporary is that they just hung down below the layout suspended by the wiring to them. This situation stayed that way until about a month ago when I decided it was time to fix the mess.

I won't go into a too long-winded explanation of what I did so I will present the photo below and it will become self-evident. The EB1s are mounted on the perspex panel with 3D printed stand-offs. The stand-offs were 3D printed because we were in COVID-19 lockdown and I didn't want to do Click and Collect at Jaycar where there is a whole range of different sizes. Longer stand-offs were also printed to mount the panel on the layout fascia. The lit red LEDs on each EB1 board indicate that all is well.

Now, I should mention that somewhere along this journey when the EB3 board was replaced by the EB1s the short circuit disappeared and hasn't shown up again.

I should also state that the mainline was divided into two districts and the Cassilis branch was the third district. The mainline was divided at Coxs Gap loop but for years the two mainline districts would shut down when there was a short in either, I obviously had a sneak path in the track wiring. I searched for this for a long time and eventually accepted that this was just the way it is. Of course, when doing something totally unrelated under Coxs Gap loop I found some wiring that went from one end of the mainline outside the loop to the other end, it was tucked up behind some layout framework. A quick snip and all was well with the BYLONG world.

In the photo above a toggle switch and a 5 Amp ammeter can be seen. This measures the total amperage drawn by the layout when operating. The toggle switch is to cut the ammeter out of the power to the track when not reading as it uses some voltage. It is turned on when needed to check the current draw.

I am very pleased with the installation and it has taken away my concerns about those EB1s hanging in the wiring, also they were not very professional looking.

Sunday, October 3, 2021

A Track Cleaning Train

 As we all do, the search for a good and easy way to clean track can cause us to try all sorts of things. Of course, one way is to use the Peco track rubber or other brand, another is to use what used to be known as an ink rubber (eraser), a slight abrasive was bonded in the rubber. I have noticed that these are very hard to get anymore which is a pity as they were better than the Peco rubber. I am not going to get into the argument about the abrasive rubbers scratching the rail head as my track is long past protecting from that potential issue.

One thing I have always done is to use a 38mm paintbrush and drag it along the track at about 45 degrees to remove dust which I believe is a major part of dirty track. Combine dust with some conductivity fluids and you make the grey stuff that builds up on the wheels. Anyway, that's my theory. The dust by itself can also cause erratic power pick up and running. I will use the brush if I haven't run the layout for a few weeks.

In regards to conductivity, I started decades ago by using CRC 2-26, an electrical conductivity spray, this certainly works well but must be applied sparingly and away from grades. I applied it to about 50cm of each rail on level track in several places around the layout as well as on the point blade and stock rail contact areas of points.  I did find many years later that it can build up to a slightly sticky form at the point blade contact areas and start to restrict blade movement. This was rectified with an application of enamel paint thinner on a cotton pipe cleaner. This is how I enhanced conductivity until I started looking for other ways several years ago.

The next trials involved the use of graphite pencil sticks which seemed to work well but also had to be kept away from the layout grades. These can be found in art supply shops.

My latest efforts involve a geared motor rotating a cleaning pad against the rail tops. This motor is built into an NSWGR 48 Class HO model that runs as a dummy with another diesel locomotive, usually a Trainorama 44 Class. Behind the diesel locomotives are several wagons that have other track wiping and cleaning fittings.

Here are some photographs that show what I mean.

A white metal Protype BCW with an abrasive rubber (from Walthers I think)
The middle wagon has a piece of 1mm cork mounted on the Dust Monkey brush
Noch also makes the axle mounted Dust Monkey brushes

The cork disc and the cork on the Dust Monkey have a small amount of CRC 2-26 electrical conductivity fluid applied to them before operating. The CRC 2-26 can also be applied to the rails in a number of places around the layout to rejuvenate the cork. Any buildup of the grey dirt from the track can be cleaned from the cork carefully with some enamel thinner. Be careful not to saturate the cork with the thinner as it may cause the glue holding the cork on the plastic to soften. I use a cotton pipe cleaner to clean the cork.

The 48 Class has a piece of grey sponge mounted behind the leading cowcatcher to wipe the dust away before it can get on the locomotive's wheels and cause problems. I am still looking for a small brush arrangement to replace the foam.

The motor that spins the cork disc is a 1000 rpm geared 12volt DC motor that can be found here (at the time of this post of course):

On ebay


On Aliexpress

The body of the 48 Class is an old Trax body and the chassis of the 48 Class is made of several 3D prints. The non-powered 48 Class bogies are from the early Trax 48 Class and I have used two of them. They clip into the 3D printed chassis the same as they did on one end of the Trax 48 Class. It was lucky that I kept the Trax 48 Class chassis' when I made new mechanisms some time ago. See the blog post here.

The prints were done on a 3D filament printer except for the fuel tanks which were done on a resin printer. The fuel tanks might work on a filament printer with appropriate orientation and supports.

If anyone is interested and has an old Trax 48 Class body and a couple of the non-powered Trax 48 Class bogies then here are the STL files of the chassis parts and the fuel tank side (two required).


Chassis Motor Bottom Support

Track Cleaner Disc

48 Class Fuel Tank Side

The chassis parts were printed flat on the bed of the filament 3D printer.

Of course the adventurous could always design and print the bogies and use the white metal bogie side frames if they are still available.

The motor sits relatively loosely in the centre square hole and the square part below the chassis is glued into the chassis to retain the motor. The fuel tanks screw onto the chassis and retain the body. The disc slides onto the motor shaft which has a flat keyway. The disc can move up and down under its own weight but a small piece of foam is used to apply slight downward pressure. Too much pressure can stop the motor from rotating. The longer two box shapes can be used for some weight and/or a place to put the DCC decoder. Although my 48 Class isn't fitted with a Keep Alive as yet it would be a good addition.

Various trial materials were used on the disc - 2 types of material, cork,
2000 grit Wet and Dry and the last one has glue on it from a covering I removed
Under chassis view showing a piece of foam that exerts
a slight downward pressure on the disc.
View of the chassis, body and two fuel tanks that retain the body.
View showing motor sitting in the square box mounting.

1000rpm 12 volt motor

The decoder has Back EMF which can assist in keeping it spinning. The decoder is not in a DCC consist with the 44 Class as it needs to run at full speed. Any decoder with Back EMF would be suitable. The 44 Class is run slowly to give more cleaning revolutions of the disc on the rails.

Finally, this method of making a track cleaner could be applied to any other dummy model diesel by designing and printing an equivalent style chassis, etc.

Here is another motor and gearbox style that could be used. The link was good at the time of this post.

If your diesels are having intermittent power pick up issues then have a look at this earlier post where I put forward a theory about it and solve it with some easily fitted power pickups.