Please get vaccinated.


Plastic cement applicator

I have been using a bottle of Model Master plastic solvent cement for building plastic kits for about 25 years. The bottle has a very nice metal tube allowing very precise application of the 'glue'.

I put some red and yellow paint on the black cap so that I could find it easily on my workbench.

Sadly this week it came to an end. I did however have a very similar bottle of Testors solvent glue but alas, it does not have such a fine application tube, but rather a plastic tube. I used it a few times and found that it kept getting clogged, something that never happened with the "Model Master" bottle. 

I looked about online and could not find the "Model Master" glue available, only Faller and Revell seem to have the metal applicators. I ordered a bottle of Faller cement but delivery is not expected for 3 to 6 weeks!

I came up with a plan to transfer the content of the Testors bottle to the old Model Master bottle.  I initially tried pushing the metal tube into the plastic nozzle of the Testors bottle and letting the solvent run through, but of course there is no way for air to escape from the lower bottle, so I decided that was not working.

Sixteen years ago a train friend Eric Joerg gave me some small syringes and 'stubby' needles. These are not sharp, but attach to the syringes just like a hypodermic needle. They have been very useful for adding smoke fluid to steamers, etc. and I decided to use them for this task too.

The stubby "needles" appear to be the same size as the metal applicator, so I pulled the tube out of the old bottle.

I then drilled a small hole in the new bottle.

I then drew solvent out of the new bottle and pushed it into the old bottle.

When done, I inserted the applicator tube into the bottle.

Well, overall, it worked but here are some important things to know:

  • I am merely reporting what I did. I am not suggesting anyone does this as there are risks involved. If you do this or anything similar you do so at your own risk.
  • The cement does not flow easily, it is very hard to draw the solvent out, it takes time for it to come up into the syringe. I removed the cap so that air could get in. It is also slow to squirt the solvent into the destination bottle, there is no way for air to escape. I tried squeezing some air out of the bottle before inserting the needle but solvent always seemed to ooze out. I also tried pulling air out using the syringe but it is hard to judge where the air is.
  • The operation took me an hour to transfer the glue.
  • Solvent does spill out. This task should be done outside, I ended up sniffing glue for an hour.
  • I tried replacing the cap on the new bottle so I could squeeze the side to speed up the transfer of solvent to the syringe, but I think I may have ruptured a seam because the next thing I knew the bottle was covered in glue. I got solvent all over my hands. Gloves would have been a good idea. I used some cloth rags to soak up spilled glue, and disposed of them outside (in rain) until they could be placed in the trash.
  • To speed things up, I used a second syringe so that one could be drawing solvent out while I used the other to put solvent into the other bottle. It was tricky holding each bottle, so it would have been better to clamp the bottles so that they did not have to be handled.
  • Bits of solvent did spray about, either through syringe operations, or caused by air escaping from the destination bottle. My reading glasses acted somewhat as safety glasses.


1960s model lighthouse

Though this is not actually part of my train layout, it is very much a model and close to HO scale.

When my paternal grandfather passed away in 1975, I said I would like to have ‘The Lighthouse’.

This was a model lighthouse that he had made, many years before, possibly in the 1950s or 1960s, and it used to sit on their mantelpiece in their living room. Standing about 29cm high it includes some rocky terrain and a small lighthouse keeper's house. The magic thing was that one could pull a small shaft on the side and the shed and tower would light up, the lighthouse lantern room light would go on and off periodically. I recall that it simulated the occluding type, meaning it was off for a longer period than it was on.


3D printed LED exterior light

While assembling a plastic building, I decided it would look much better if the exterior lights actually worked, so I set about designing some exterior lights.

I used OpenSCAD to design a half hemispherical shade (5mm diameter) with a small wall mount. A 1mm diameter hole allows for the wiring of a surface mount 0402 LED. After three 3D printing iterations I ended up with a usable design.

I painted the inside with silver acrylic paint, and threaded an LED in to test it out.


As expected, there was a lot of back scatter of the light, so I decided to paint the top black.

At this stage they look rather like toilets.

I threaded in the LEDs. In order to keep the LED facing the right way I found I had to tape the wires to the work surface.

I then added a drop of super glue to each LED and let them dry.

To add them to the building, I drill a small hole with a pin vice.

After threading the wires through the wall I could see that a second coat of black paint was in order to ensure complete coverage.

A drop of super glue holds the light in place.

I will use these on the row of Vollmer buildings I am constructing. I will probably run two 5V power circuits to the buildings, one, always powered, for internal lights and the controller for the butcher shop and the other, switched, for night time when exterior lights are appropriate.

A 5mm light in HO scale represents a 43.5cm light - a reasonable size. I could probably make them even smaller.


Butcher shop - Metzgerei

I was given a set of Vollmer buildings called "Bahnhofstraße" (Station Street) item 3675. It is a box of five buildings that were/are also available separately. I started with the butcher shop which is an end building with two large shop windows. The building is also sold as item 3674. This post describes how I built it and added super detailed interior.


Extension tracks landscaped

Now that I have installed all the turnouts and three sensors per track, I have added the cable ducts, indusi magnets, and some weeds to the storage area known as Oberbad.

There are cabinets for storing additional lengths of signal cabling too..

First trains have arrived.

I still have to design and manufacture the dwarf shunting signals, and the little connection boxes for the magnets and signals. The software to control the signals is all done and ready.

I ran an S-Bahn train from Oberbad to the underground station (Wilsnack Tief) and it was nice to see 'Oberbad' appear as the origin on the station platform sign!

Previous post on Oberbad:


Märklin turnout motor 7549 reliability

For many years I have used a tip I read online about disabling the electrical switch mechanism in the Märklin K-track turnout motors. As I made the modifications to my last few motors I decided to show what I do and took some pictures of the steps.

Note that over the years the shape of the Märklin 7549 turnout motors have changed. This post applies to the older version which can still often be found at train fairs and the second hand market.

The turnout motor comprises two solenoids that pull the mechanism in either direction. The design includes an electrical cut-out to prevent the coils from remaining energized after fully actuating. This prevents the coils from overheating and melting the plastic thus allowing the motor to survive being switched with continuous current. Unfortunately the physical switch adds a tiny amount of load to the mechanism and this sometimes causes the motor to not fully switch. Since I switch all my turnouts with a pulse (and my software is stable enough to not forget to switch the current off after each pulse) I remove the physical switch to reduce the possibility of needing to pull the motor out of the layout due to failure.

In addition to removing the switch, I also lubricate the mechanism with graphite powder.

The plastic cap of the switch motor simply pops off revealing the two coils and the mechanism that slides up and down.

The yellow wire provides power and the two blue wires complete the circuit when grounded. Two copper leaves isolate the 'yellow' power from the ' blue' ground depending on the motor position. This means that when the mechanism has been pulled to the right, the right hand coil is switched off and vice versa.

Removing the two switches


It is sad that one needs to explicitly say this, but modifying a switch motor may result in its destruction. The modification described here is not reversible. It is a permanent modification. I am not responsible for anything that may result from anyone attempting to follow what I describe here. That includes damage to any switch motors, tools, trains, power systems, bodily injuries and psychological effects, loss of warranty, or any other nonsensical claims against me.

The first step is to pull up one of the copper leaves. I push the mechanism to the left and then pull out the right hand leaf. I use some tweezers to grab it and pull it out from underneath the plastic slider.

Pulled up

Once it is pulled up past the plastic slider, I fold it back and forth to cause it to break off at the circuit board due to metal fatigue. 

Three or four motions left and right are normally enough to break it off.

I repeat all that on the other side too.

Now, since I have removed the switches, current from the yellow wire can never reach either coil. So I have to bridge the connection on the left and right.

Fortunately there are solder pads I can use! I bridge the two pads marked A and B where the red line is in this photograph:

Right hand side done:

I do the same on the left, making sure that the yellow wire remains connected.

If I find that I end up with a big blob of solder bridging the pads, I check that it does not foul the end of the moving mechanism. If it does, I simply snip off the end of the plastic as can be seen above.

At this stage it is prudent to test that the motor actuates in both directions. I hold the bare end of the yellow wire onto a yellow wire from a 16V AC transformer and then alternately ground (brown) each of the blue wires for just a moment. As each blue wire is grounded it should snap the mechanism left or right.

Lubrication of the mechanism

These switch motors are not expected to be lubricated, but I lubricate mine using graphite powder. 

I dip a small screwdriver into the powder and tap it where the plastic mechanism slides against other parts.

I also try to get some into the area where the tongue actuator moves in and out of the motor..

I then move the mechanism up and down a bunch of times to get the powder in place. One can feel how easy the movement becomes.


Track power, AC, DC, and digital

There is a tremendous amount of confusion in the hobby about the nature of the electricity found in model train tracks. This is aimed and explaining the characteristics of the different types..

Before we get into actual trains let's get the terminology straight first.

Alternating current (AC)

From Wikipedia

Alternating current (AC) is an electric current which periodically reverses direction and changes its magnitude continuously with time in contrast to direct current (DC)

The usual waveform of alternating current in most electric power circuits is a sine wave, whose positive half-period corresponds with positive direction of the current and vice versa. In certain applications different waveforms are used, such as triangular waves or square waves. These currents typically alternate at higher frequencies than those used in power transmission.

Direct Current (DC)

From Wikipedia

Direct current (DC) is one-directional flow of electric charge. The electric current flows in a constant direction, distinguishing it from alternating current (AC)


Pulse wave

  From Wikipedia

A pulse wave or pulse train is a kind of non-sinusoidal waveform that includes square waves (duty cycle of 50%) and similarly periodic but asymmetrical waves (duty cycles other than 50%).

Digital Cab Control (DCC )

This is the predominant standard digital protocol, and is maintained by the NMRA. The electrical characteristics are defined by NMRA: S-9.1 which defines the signal on the rails as:

The NMRA baseline digital command control signal consists of a stream of transitions between two equal voltage levels that have opposite polarity.

From Wikipedia

The command station/booster quickly alternates the polarity on the rails, resulting in a modulated pulse wave.

DCC signal wave form

Now, let's look at some model train history. There are two ways of supplying current via the tracks to electrical model trains, three-rail and two rail.

  • Three-rail systems have an electrical contact between the two outer rails and power is picked up by a slider that runs along the third rail and returns power through the wheels to the two running rails which were electrically connected.
  • Two rail systems use both running rails to provide the two electrical poles.
Historically some model train manufacturers, such as Märklin, opted for 3-rail and others, 2-rail.

In addition to the number of electrical rails used, manufacturers needed to choose the nature of the current. Before digital came along there were two analog options AC or DC. Märklin standardized on AC and others went with DC. Both systems controlled the speed of locomotives by varying the voltage of the current. Märklin controllers varied the running speed by varying the AC voltage from zero to 16 VAC and the wave form (at full speed) can be represented as follows:
Representation of AC current over time

The wave form of the analog train track voltage was sinusoidal (that wave shape you see above), simply because it was created by transforming mains voltage (sinusoidal 220VAC or 110VAC) to 16VAC.

Changing direction was initially achieved by flipping a switch on the locomotive, and then later by sending a high voltage AC pulse. The pulse actuated a solenoid which switched the wires through the field coils.

DC systems controlled the speed by varying the DC voltage and direction changes were achieved by reversing the polarity so that the motor turned in the opposite direction. Here is a representation of a DC current at a constant speed setting:
Representation of DC current over time

The combination of those two decisions resulted in a world where most 3-rail systems used AC current and most 2-rail systems used DC current. This in turn, resulted in many people referring to the two dominant systems as either  AC/DC or 3-rail/2rail through association. It is quite possible to have a 2-rail AC system and a 3-rail DC system.

The upshot of this nomenclature is that many people now think that any 3-rail system is AC and that 2-rail systems are DC. That was not much of a problem until digital control came along.

As can be seen by the definitions of AC and the NMRA DCC specification, there is no other logical conclusion other than the fact that DCC is a type of alternating current. The current alternates polarity rapidly, which is what defines AC current. The DCC signal is not sinusoidal, but as is clear from the Wikipedia definition, AC current is not limited to being sinusoidal.

Here are three different AC waveforms:

Representation of AC current over time

Representation of pulse wave current over time

DCC signal wave form

Now that all describes the nature of the current in the rails. When a digital locomotive gets a digital signal from the track, it feeds that into a digital decoder. The decoder decodes the information in the voltage polarity oscillations into data that it uses to decide if it needs to do anything (such as go faster/ stop/ turn on lights etc.) and it also converts (rectifies) that alternating current into a DC current. In addition to powering the decoder itself, the DC current is also fed to lights, and the motor with the polarity matching the desired direction of travel. The speed of the loco is governed by pulsing the DC current on and off at different ratios. This is called Pulse Width Modulation (PWM).

Modern decoders prefer using DC to drive the motor for two main reasons:

  1. It is backward compatible with DC motors found in analog locomotives as well as universal AC motors.
  2. It allows load control using back EMF - essentially it is able to get feedback from a DC motor to determine how fast the motor is turning.
This means that if one is converting an analog AC locomotive to use a modern decoder, it is best to convert the universal AC motor to a pure DC motor. This is achieved by replacing the field coils with a permanent magnet.

So, if you have been fully following along, a modern digital locomotive is picking up an AC digital pulse wave from the tracks and sending PWM DC power to the motor.

This is why referring to a locomotive as either AC or DC can be confusing. AC could (sloppily) refer to 3-rail, or the form of the track current or the nature of the motor in the loco.


People who have digital systems sometimes try to debug a problem by measuring the current in the tracks by whipping out their multimeter. Some multimeters demand to be set to either AC or DC depending on the type of current to be measured, newer ones may also have a combined mode in which it figures out if the power is AC or DC. The problem is that most meters assume that any AC current to be measured is sinusoidal in nature, because that is the usual form of AC current that most people encounter - we all have it supplying power in our homes and old analog train controllers.

Now, sinusoidal wave forms of AC have a very very short period where the voltage is at the maximum and minimum. For this reason AC sinusoidal voltages are expressed as the Root Mean Squared (RMS) voltage - the voltage that gives the same effective amount of resistive power had it been DC instead of AC. i.e. the RMS voltage of a sinusoidal AC current is determined dividing the average peaks by the square root of 2, which is 1.41.

Look again at the representation of the wave form from a 16V sinusoidal AC power source such as a Märklin analog controller:

Representation of AC current over time
Note that the peaks and troughs are in fact at 22.4 Volts not 16 Volts. This current would supply the same effective heating power as a 16V DC power supply.

When measuring this with you fancy RMS meter, you want to get a reading of 16VAC not 22.4VAC. You can think of it as the meter automatically dividing by 1.41 for you. This works so long as the wave form is sinusoidal.

What we have learned however, is that the AC current in a DCC digital system is not sinusoidal. This means that using your RMS meter to measure the voltage of the track current will not give you a usable result. What you get is undefined and thus meaningless. Meaningless. It does not mean that the power in the tracks is not AC and must therefore be DC, nor does it mean it is, 3-phase, or some alien mixture of power unknown to Wikipedia. The only meaning is that you do not have a quick way of determining the amplitude of the wave forms.

If you want to know what the maximum amplitude (V) is (perhaps to compare different systems?), you need to rectify the power and measure the DC voltage after rectification. You can use an off the shelf rectifier or 4 diodes (such as 1N4001) arranged like this:

rectifying and measuring digital track voltage

Brake modules

Some manufacturers offer what they call brake modules that will stop a digital locomotive in a non-digital manner. It works by rectifying the track current (so that it is DC), and locomotive decoders which support braking modules are designed to slow to a stop when they detect DC current. That tells you that the digital current is not DC, leaving only AC as the alternative (given that there is no 3-phase system).

Other digital systems

In the text above I have described the AC wave form of DCC, but there are also other digital systems, such as the original Märklin digital system. Detailed information on the electrical characteristics are not as readily available as for DCC, but the Märklin (item 0303) book: Model Railroading Digitally Controlled by Georg Fuhs and published in 1988 gives clear evidence that the track signal is AC.

On page 11 was decided during the design phase of the Märklin Digital system to limit it to only two "conditions", namely "positive voltage" (approx. +20 volts) and "negative voltage" (approx. - 20 volts)

When describing how the original c80 decoder works on pg. 71 it states:

The c 80 locomotive decoder receives information sent by the Central Unit or Central Control. This information is first checked for frequency. The decoder is able to differentiate conventional operation (50 Hz), information for solenoid accessories (approx. 10 kHz) and information for locomotives (approx. 5 kHz). 

DC current does not have a frequency to detect. It goes on though:

The "digital current" is rectified at the entry point to the decoder so that a continuous current for controlling the locomotive motor through the track is available independent of the flow of data. The power supply for the auxiliary function, by comparison, comes only from the negative part of the third rail potential.

So, they rectify (convert from AC to DC) the current, and furthermore they declare that the aux function is only half wave rectified, a clear indication that both positive and negative voltages are present. That means it is AC.

On pg. 73 they discuss what happens if there is DC current on the track:

If DC current is present in the track, the locomotive will function only when there is a positive potential in the third rail. If the potential is negative, this corresponds to the condition "turn on" in the Digital control and results in a quasi "stand by": the locomotive remains at a halt and the auxiliary function is shut off. The information for speed and auxiliary function remain stored, however. As soon as the third rail potential becomes positive again, these stored commands are carried out.

So this conclusively confirms that the Märklin Motorola based digital track current was also AC.

Personal note: In 1989 I visited Georg Fuhs at his residence and demonstrated the program I had written which he tested on an N scale Arnold layout he was building in his basement, and it worked.

Note that unlike DCC, the Märklin Motorola (MM) system encodes 1s using a positive voltage and 0s with a negative voltage. DCC instead makes the duration of both a positive and negative cycle twice as long for a 0 than that of a 1. This means that if you were to keep sampling a voltage directly on a MM system you might see a higher voltage when more 1s are being transmitted than 0s. On a DCC system the ratio of 1s to 0s will make no difference when averaged.

Additional notes

  • Amazingly it is possible for both MM and DCC to coexist in the same track. This is achieved by controllers such as the Intellibox, allowing both MM and DCC decoders to run on the same track.
  • Many decoders can handle multiple protocols and even analog, switching to whatever style of current they detect.
  • If you are still not convinced that digital track signal is a form of AC, place an LED with a 1K Ohm resistor across the two track poles. LEDs can only tolerate power in one direction. They will burn out if AC current is applied or DC is applied in the incorrect polarity. If the LED lights up  and stays on, then you have DC or pulsed DC current. This can be confirmed by reversing the LED, switching the polarity. If it then burns out, it confirms that you have DC or pulsed DC power. If an LED burns out no matter what the polarity, then you have AC power.


Android Cab control

At times I seem to get fixated on making a cab view control and to date I have made five such cab controls, each time with different technology, and each time it gets a bit better.

My first effort was built into the train control software.

Then I got an HP PDA device that had a wifi connection so I wrote a program that ran on that.

Then I made one that would display in a browser. In this version, I managed to superimpose signals showing the current signal aspect into the image.

Then I made one that would run on a phone using Blynk technology. It did not allow the flexibility to position the signal in the image so I displayed signals to the side, but at least it was now on my phone.

And now, I have made one that runs on a phone (and Windows) using AppGameKit

This version has a better speed display, in addition to the track speed limit and current desired speed of the train, speeds above the speed limit of the (lead) locomotive are shown in red. The speedometer scale also changes dynamically according to the maximum speed of the current locomotive.

As before, all available digital functions can be invoked by buttons, but now the buttons also make a realistic click sound. It also superimposes signals onto the track image. When one touches any control in the cab, the image outside the window goes out of focus for half a second to simulate the depth of field of the driver's vision.

It now also features a cab window frame with windscreen wiper.

The Locomotive name and current odometer is also displayed. The odometer changes as the train travels about and is an actual distance covered by that model, accumulated over the years.

When the signal changes it shows immediately.

A new feature is the ambient light of the track image can also be simulated from bright and sunny to pitch dark. The locomotive headlights light up the image too! Here it is getting somewhat dark and the loco lights are on.

In total darkness and no headlamps!  The ambient light data comes from the layout control software which will relay the current ambient light of the layout room. So, as a sunset occurs, we see sunset colors in the cab control on the phone!

(I plan on perhaps not having the reflective stripes show up without the headlamps.)

When inside an underground station that has its own lighting, the image is of course not darkened.

Oh, and it can also announce destinations etc. in German. When the train being viewed is dispatched there is a voice announcement stating that the driver has permission to start and what the destination is. If a trip is cancelled, that is also announced. 

When the train has been dispatched, the desired speed is indicated with a yellow marker on the speedometer, and also presented (below the digital speed indicator) in yellow. 84Km/h in this example:

As before, it can also:

  • Turn layout track power on and off
  • Disable unexpected train alerts
  • Pick any train and dispatch it to any destination.

  • Trigger any event in the train control software
  • Set any accessory address to red/green
  • Monitor any sensor address

The track images are cached on the phone. If the software is directed to use an image it does not have, it fetches the image from the Bw software. It can also request all the images for all tracks and caches those for instant loading as needed.

It is based on the RemoteSign command set, communicating with my Bw train software over the network.

I might add the Buchfahrplan information display so the person driving the train sees the prototypical journey information as they progress along their trip.


I have added a throttle so one can control the acceleration and braking of the train.

I have also added Indusi, Sifa and AFB features.

The Indusi system warns the locomotive driver about upcoming signals and they have to acknowledge these warning by pressing an acknowledgement within 30 seconds, and bring the train speed below a certain threshold within 20 seconds. If these conditions are not met the emergency brakes are applied. The allowable speed for the train after passing 1000Hz Indusi transponder is also indicated with the prototypical flashing. The system automatically sets the Indusi category (which in turn governs the speed limits allowed) based on the train definition in the main layout control software.

Here an ICE has passed a 1000Hz Indusi magnet and has 20 seconds to get below 85km/h (and is only doing 70)

Sifa is an additional driver attentiveness system that requires a button to be pressed periodically. If the button is not pressed, then a verbal warning is made in the cab and if the Sifa button is still not actuated, the emergency brakes are applied.

The AFB (Automatische Fahr- und Bremssteuerung) mode can be switched on and off. When it is on, the system controls the speed of the train itself. When it is off, you have to set the throttle/brake control manually to regulate the speed of the train.

Here an ICE is coasting along at 107km/h with AFB off.

A train with the lowest speed category in a hidden area awaiting a green signal

An S-Bahn train in the middle category: