Upgrading A Peco Insulfrog Crossing For DCC Operation On An Existing Layout

One of my customers has been having some issues with his layout since switching to DCC operation; they find that trains short and stop when crossing Peco Insulfrog crossings. In this post I’ll explain why this happens and show you how I fix it.

The Peco Insulfrog crossing comes in short and long versions and can be used in many ways. The term Insulfrog refers to the areas where rails cross each other, which is called a ‘frog’. An Insulfrog has a plastic separator to isolate the rails electrically. With an Insulfrog crossing, the two tracks are electrically separate from each other. So differently controlled trains can run on each line, as long as they don’t hit each other.

But the problem comes right at the centre of the frog where trains with wide wheels cross the plastic isolation. In the image below I’ve circled the area called the ‘frog’. The downside is here the larger the plastic frog, the bigger the dead spot is, or area without power, for the locomotives. So Peco has made it as short as possible to limit the dead spot. However, if a locomotive or regular item of rolling stock has wide metal wheels, that is, wider than the rail, the wheel will make contact with both incoming rails at the tips.

When using these on a traditional DC or Analogue layout there tends to be no issue with shorting at the frog because power is normally only given to the tracks a train is running on either by a power routing point or switch, so it’s more than likely the other track is switched off. However, with DCC all the track is powered and shorts at the frogs are very likely. This is why Electrofrogs, which means the frog is made of all metal, are ideal for DCC. But the polarity of the frog, positive or negative needs to be switched depending on the direction the train needs to travel.

My customer has several of these crossings, the one below is at a major junction between four double slips. To solve the issue of trains shorting on this crossing I’m going to upgrade it to work like an Electrofrog crossing, just without the metal frogs.

Fortunately my customer had already installed insulating fishplates at the ends of the frogs. This is always needed with Electrofrog points, crossings and slips but not for DC/Analogue. If these were not already there I would need to cut the track using a slitting disc in a Dremel-style tool to create the gap.

To completely isolate the frog section, I need to cut the rails on the inside of the frog as well. The cuts want to be roughy in the middle of the rails section; this is because Peco has placed a wire under the plastic frog to bridge it. You can see the bridge wires below. The cuts need to be between these wires.

Using a slitting disc in a Dremel tool I’ve cut the rail in four places, above the gap in the sleepers to help keep the integrity of the rail support.

It’s important to make sure you cut deep enough to go all the way through the metal rail.

The frog area should now be completely isolated from everything, and currently, the two frog rails are separate from each other. To get power to them, I drill a hole just outside the frog and feed a wire through, I always use green, because most frogs are green! I strip more than normal off the end of the wire and twist the strands.

Then I loop the end, not a full 180­° but to the same angle as the frog.

The stripped wire can then be pushed into the web of the rail which is below the railhead and out of the way of any passing trains. The wire needs to be in contact with both rails.

The wire can now be soldered to both rails which creates a basic Electrofrog. The excess wire is pulled back through the base board.

This is repeated on the other side of the crossing.

Although the frogs are now connected to a wire they will need to switch to the right power feed depending on which way a train is going. If it’s going from top left to bottom right the left frog will need to be connected to the bottom rail, but if travelling from the bottom left to top right it will be the top rail. This could be done with a manual switch but to make things easier I like to use Autofrogs. These are electronic relays or automatic frog polarity switches which detect if the frog is at the wrong supply and switch it before the DCC command station realizes there’s anything wrong.

On the reverse of each one are three solder pads. The one marked F, on the left, is for the frog wire. The other two marked R, are for rail connections. It doesn’t matter which way round the DCC is connected to these pads as the Autofrog will always pick the right one itself.

I first add some solder (often called ‘tinning’) to all the pads.

Then I connect the power cables This particular layout uses four core cables to connect power and track power to its point motors so I used the same wires.

The frog wires are then connected to the Autofrogs under the layout and the four core cable is connected back to the main DCC bus.

At this stage, I run trains through the crossing to check it all works as anticipated. As a train crosses in a different direction to last time you should hear a faint click from the Autofrog as it changes the polarity, the train should carry on without even noticing. Which it did.

The last thing to do is secure the Autofrogs, and they have a single hole which can be screwed to the base board.

Doing this eliminates the shorting issues and trains now run without shorts and stalls.

The Gaugemaster Autofrogs work well but they’re not the only product that works like this. Tam Valley makes a product called a Frog Juicer which is all electronic, unlike the Autofrog which has a mechanical relay.

I’ve used both many times and both are very good. Although the Frog Juicer is slightly more expensive it does come with screw connections so soldering is not required.

When it comes to upgrading to DCC there are other ways to try and overcome the problem of shorting, but as a permanent solution with Insulfrogs this fix can be done without lifting the track or replacing it, and enables trains to run smoothly so you can get on with enjoying your layout.

Building A Helix

As well as 3D printing model railway parts and kits I also build a lot of layouts for customers. Sometimes it’s a whole layout or sometimes, as in this post, I get asked to just do one bit. This week I wanted to share with you a recent job I did for a customer installing a pair of helixes in their new layout.

The customer is building a new layout in OO gauge (1:76.2) with staging yards on a lower level, 12.5″ (317.5mm) below the main baseboard, and they wanted to join both ends of the staging to the layout with a helix in each corner. Using 3rd Planit, which is 3D model railroad design software (https://www.trackplanning.com), I was able to draw out their baseboards and the track the customer had already laid in the staging yard. This allowed me to create a plan for the point work required, but most importantly shows the customer just how big the helixes were going to be.

As you can see the new helixes overhang the baseboard. The outer track on each helix has a radius of 22.5″ (572mm), the inner track is 20″ (505mm). The customer was concerned about the size as they’d expected them to be a lot smaller and fit onto their baseboards, but there are several good reasons for choosing these sizes.

Firstly is the availability of helix kits. I could design my own, cutting all the material and making a truly unique helix for this build, but that is very time-consuming and therefore costly, and as there are already several great kits on the market it makes sense to use one. The kit I used for this layout came from Model Railway Solutions, they provide two sizes of helix kits for OO and this leads to the second reason.

The first OO helix kit MRS produces is for 2nd & 3rd radius curves and the second kit is for 3rd & 4th radius, but what does that mean? The terms 2nd, 3rd, and 4th radius refer to the radius of Set Track, the curves often supplied with train sets, starter packs and sold separately by companies such as Hornby and Peco. Set Track is described as a range of rigid curves, straights, crossings and points (turnouts), made to the standard British geometry. With the curves, as the number gets bigger, so does the radius, and below you can see Peco’s Set Track curves from 1st to 4th radius. The key advantage is you can easily keep multiple tracks parallel around curves.

MRS’s helix kits have been specifically designed to match Set Track and for this build, I used the 3rd & 4th radius kit. But why choose that over 2nd & 3rd radius kit which would be smaller? The answer depends on what type of layout you’re building. Some larger locomotives have a minimum radius they can navigate and a few specify above the 2nd radius curve which is 17″ (438mm). So if you have large locos, which my customer does, they may struggle with the tight curves. But the main reason is the gradient.

These helix kits climb 3″ (76mm) with every revolution. As the radius increases the distance traveled increases and consequently, the gradient reduces. In the table below you can see how this works out.

RadiusLength of Full CircleGradient
2nd Radius17″ (438mm)108″ (2752mm)2.762 %
3rd Radius20″ (505mm)125″ (3174mm)2.395 %
2nd Radius22.5″ (572mm)141.5″ (3595mm)2.115 %

Although the difference in gradient may not seem a lot, it can make a huge difference to a locomotive pulling a train uphill around a curve. My customer wants to run steam locomotives pulling at least six coaches so in order to give them the best chance the largest radius is recommended.

You may have noticed from the track plan at the start that the helixes are not mirrored but both climb in a clockwise direction. This is because trains in the UK run on the left so making them climb a helix in a clockwise direction means they are going up on the larger radius which is an easier gradient and coming down on the tighter radius which is slightly steeper. Model trains tend to find it harder pulling uphill than braking downhill.

The helix kits from MRS have three main parts. An entry/exit ramp, a first-level base kit, and a riser kit. The entry and exit ramp is a tapered section that allows the helix to start on top of a flat baseboard without modification.

All the small bits come boxed together, the amount depends on your build. All these parts are for two helixes, each with four rotations.

In the box, you get 3D printed pillars ranging in size for the first level base kit, regular pillars, pillar caps, clips, superglue, superglue setter & cable ties. The box of screws in the photo are not part of the helix kit, they’re for fixing down the baseboard tops.

The actual helix deck is laser cut from MDF, each one has a stepped end to allow easy and accurate joining.

The one thing not included in the kit is the track and it’s amazing how much you need for a helix. I chose to use Peco Set Track and to build these two helixes I needed 64 3rd radius double curves and 128 4th radius curves (they don’t come in double).

As I said at the beginning, the helixes are larger than the baseboards and although my customer had cut holes in their benchwork, extra support was going to be needed.

The support needs to be under the pillars of the helix and as long as it’s flat it doesn’t really matter what it’s made from. I was able to use the offcut material to make a flat surface, instead of buying new timber. The important thing to note is the hole in the middle, which is crucial as once the helix starts going in you won’t be able to reach the backtracks without it.

I added a triangular support at the front and some outcrops so each pillar has somewhere to sit.

In MRS’s first level base kit the 3D printed pillars are labeled A to H in ascending order. these sit under the helix and screw through the deck into a regular pillar. These first pillars are effectively the feet. They do have a thread in their underside should you wish to fix them down but the overall weight of the helix once complete will stop them moving.

But before we start fitting these the entry/exit ramp needs to be fitted and for this, the superglue is used along with the two bulldog clips.

Because the height between the helix decks is at a premium any fixing that protrudes above or below the deck is not useful so the joints are created by gluing the steps together and holding in place with the clips. This superglue holds fast in about 10 minutes. The spray is a superglue actuator that causes it to set instantly. I used it on this first joint but as for the later connections, I found I didn’t need to. Simpy putting superglue on the step, fitting the next board, and clamping for 10 mins with the clips worked perfectly.

The first 3D printed pillars can now be fitted along with the standard pillars. The standard ones on top act as the nuts, fixing the lower pillars. When screwing these in don’t overtighten them. You don’t need any tools, finger-tight is sufficient.

If the pillars are sat on a flat surface, the helix will climb at a constant gradient.

I test fitted the second deck section to make sure everything was in the right place, the end of the second deck fitted over the entry/exit ramp. Don’t fix the second section in place yet as you need to start laying the track onto the helix first.

With the helix in the right place and the second section removed, I put a few screws in the entry/exit ramp to hold it all in place, then started laying track. When using Set Track you instantly realize the benefit; it’s a perfect fit and holds a constant radius with ease.

You may have noticed I haven’t used cork or underlay under the track, this is to maximize the height between the decks. I also chose to pin the track down. Because it’s Set Track and doesn’t flex you only need a few pins, which is good because there’s some bounce if you try and pin between the pillars, but close to them it’s okay. One thing to consider is the length of the pin. The deck is only 7mm thick so if a track pin is put all the way in it’ll stick out below and I guarantee you’ll scratch the back of your hand when cleaning the lower tracks at a later date! The solution is to put the pin in only enough to hold, then bend it over as shown below. As long as the pinhead is below the top of the railhead it won’t create problems with train clearance.

Something else to consider before you take the build too far is track power. Even on the inner line, a full loop is 125″ (3174mm) long. Times that by four rises and that’s a long way if you only have a power feed at the top or bottom of the helix; your loco may start to slow down as it gets further away from the power feed. My solution for this is to put a power feed to every level of the helix and solder the fishplates/rail joiners together for the other joints on that loop. In the photo below the power feeds are at the bottom of the photo. All the fishplates/rail joiners have been soldered with the exception of the ones at and above the entry/exit ramp.

After several rotations, track laying and soldering as I go, the helix takes shape. Building from the inside is much easier, you can now see why that access hole was crucial.

As can see above the top rotation is not a full circle, which would put the track facing the wrong way to enter the main layout, but that also means some of the elevation gain is lost. You may recall that the staging was 12.5″ (317.5mm) below the layout. The helix climbs 3″ (76mm) per revolution, so with just over three and a half it has only come up by about 11.4″ (290mm). To overcome this I build a ramp at the same gradient of the helix onto the new layout baseboards.

The track on this section will be laid with flexi track and have larger sweeping curves as it’ll be a visible section of the finished layout.

The cable ties supplied with the kit are used to hold the track feed wires together so they don’t snag on passing trains.

The helix on the other side is basically the same, although the top section is near the wall so the trains are again climbing on the outside of the helix. You can see the pillar caps holding down the top deck sections. These screw onto the pillars and act like a nut as well as covering the last of the exposed treads.

Again I constructed a ramp to make up the height difference that runs at the back of the new layout benchwork.

With all the track down and power connected, a quick test had to be done. The loco is a Heljan Class 28 Co-Bo Diesel.

These kits are a great way to add a helix to your layout and can be built in a variety of height combinations to suit your needs. They are also available from MRS for N Gauge Set Track and Kato Unitrack.

I build a lot of layouts in all shapes and sizes and I look forward to sharing some more with you in later posts.

Updating Your Digitrax Products

As with most equipment, manufacturers often make improvements to their products after you’ve purchased them.  The nice thing with electrical equipment is these improvements are often related to the way it works rather than a physical change, so this can be updated.  In this post I’ll show you how you can check to see if your Digitrax equipment is up-to-date and if not, how to change it.

Digitrax equipment is very reliable and robust, but most of their equipment has had a small improvement made at some point.  Just like a computer or mobile phone, electrical equipment has a program that runs on the circuit boards. This program is called firmware, unlike software that runs on a computer, firmware is the program that tells the components what to do.

Below is an example of some common Digitrax equipment.  From left to right we have a DT602D radio throttle, a DT500 throttle, a DCS 240 command station, a DCS 51 command station and a PR3 USB to PC interface.  All of this equipment has firmware and all of it has an update available.

Some of the equipment, such as the newer DT602, tells you what version the firmware is when it starts up. Below you can see this throttle has SW Version 0000.1 and is dated July 15, 2021.

Other equipment such as the command stations and older equipment don’t have a such a clear display. But there are ways to find out what the firmware version is.  To do this Digitrax have a piece of software called DigiIPL and this can be downloaded from their website www.digitrax.com.  This software can be installed on a PC, laptop or Windows-based tablet, but don’t start it up yet.

To connect the Digitrax equipment to the DigiIPL program you will need a USB cable, which should have been supplied with the command station.  If not, the same type of cable often used with printers will work. The cable needs a USB Type A fitting on one end and a USB Type B fitting on the other as shown below

The newer command stations, such as the DCS 240 have a USB port on the front of the unit but the older models such as the DCS 51 do not. I’ll show you how to connect those later.

With the USB cable plugged into your computer and the DCS (DCS stands for Digital Command Station) connected to its own power supply, the PC will assign a COM port to it. This is the address of the connection, depending on what else you have plugged in this will be different for each user, we will find out what it is in a minute.

With the DCS connected you can now start the DigiIPL program and it will look like this.

You’ll notice at the top left a Loconet Port has not been selected, this is where the COM port is entered. Luckily you can only select from Com ports that are active, mine had been assigned to Com 4. If there’s more than one, try the first and if that doesn’t work try the next. It’s also important to note the ‘Bit Rate’ should be set to 16457 although by default it always is.

At this stage I should point out that if you have lots of Digitrax equipment, particularly lots of the same item, it is a good idea not to have it all plugged in to the command station at the same time when doing this. Theoretically it should work but it’s recommended to do this individually.

Once you’re ready you can push the ‘Find Devices’ button and it will open another window listing the equipment, and below you can see the DCS 240 listed and is at SW version 0.3.

If I plug in the DT602D from earlier and push the ‘Find Devices’ button again you can see it listed showing the SW Version 0.1 as we saw before.

So now we know the current SW (Firmware) version but how do we know if it’s the latest? Heading back to the Digitrax website on their downloads page they have all the available updates listed. Below you can see the DCS 240 is at version 0.4 and the DT602 is at version 0.1 so only the DCS 240 needs to be updated. (I did the DT602 the other week).

To update the DCS 240, or any Digitrax equipment, download the relevant firmware to your computer. Then using the ‘Select File’ button, select the downloaded firmware file.

Pressing the ‘Start’ button will now update the equipment.

Once done you can check to see if the update was successful by pressing the ‘Find Devices’ button again to check the version.

When it comes to the older equipment, such as the DCS 51 that don’t have a USB port, an interface is needed. This is where the PR3 shown in the first photo comes in. Below you can see the USB cable connected to the PR3 and the PR3 connected to the DCS 51 via a Loconet cable.

With the new equipment connected the DigiIPL has to be restarted. This time it has connected on Com 5. Using the ‘Find Devices’ I can see the PR3 and DCS 51, but only the DCS 51 needs an update.

And thats it, all up to date. It’s always a good idea to update any firmware, Digitrax or otherwise. The chances are you will not see a difference in how your device works but there’s a good chance it will fix a problem you didn’t even know was there.

Solving Common Bad Running Issues With Kato Diesels

Kato makes some fantastic N scale locomotives and they run very well, that is until they don’t.  In this post, I’ll share with you the three most common issues I find as to why these reliable locomotives stop working.

There are of course typical things such as dirty wheels and track, but the three issues I’m sharing with you take a little bit more to fix.  They’re also common on other brands of locomotives that have similar parts.

The first issue has become apparent on this Kato SD40-2, as shown below. This model was custom painted by Paul Begg.  This is a 6-axle locomotive that picks up the power on all 12 wheels.  The power is transferred to the chassis halves via copper strips and then into the circuit board.

With the shell removed you can see the circuit board. This particular locomotive is DCC fitted with a Digitrax DCC drop-in decoder, but I’ve had the same issue with DC versions.

The circuit board or decoder drops into the chassis and pushes forward which clamps it in place and at the same time makes electrical contact with the chassis halves and motor contacts.  Below, you can see the decoder has slid back exposing the electrical contacts at the front.

Close up you can see the electrical contacts, and the slots in the chassis halves they slide into.  Ideally, this should be a tight fit requiring a little force to push the decoder forward.  The friction then prevents the decoder from moving and ensures good electrical contact.  But sometimes the decoder is a loose fit; this may because the DCC decoder is ever-so-slightly thinner than the original circuit board or maybe the chassis has been affected by the pressure of clamping the decoder.  The consequence is the decoder slides in and out easily, and the electrical contact is very poor.

The electrical contacts or pads on the decoder are copper, which is ideal to solder too, so to fix this issue I flash over the pads with a soldering iron to put a thin layer of solder on top.

The layer of solder can’t be too thick, otherwise, it’ll prevent the decoder from fitting at all. To get a thin layer I first use the iron to put some solder on, then I clean the iron tip on a wet sponge and quickly run it over the pad again. Excess solder will be removed on the iron, leaving a thin flat surface.  This may take one or two passes with a clean tip each time.

The decoder can then be pressed in.  It will take a good press as the pads are now thicker, but if it feels like you need to press too hard, rather than risk breaking the decoder, pass a clean iron over the pads again to remove a little more solder.

Once fitted the decoder should not move and you’ll have a solid connection, because the solder is softer than the copper and the chassis will dig into it.

The second issue I see a lot is to do with the actual pickups in the trucks.  The AC4400CWs below, again custom painted by Paul Begg, suffered from this.  The first fix above has already been done. This is one of the older chassis where the chassis screws need to be loosened slightly to release the truck.

With the truck bottom, and side frames unclipped the pickups will fall out.  Each wheelset has a plastic axle and metal wheels with a spike on the outside of the wheel.  The spike fits into the cone of the brass pickup.

The issue is dirt and crud that builds up inside the cup.  Below you can see two pickups from the same truck. I’ve cleaned the cups on the lower one.

To clean these cups I use tiny cotton swabs dipped in Isopropyl alcohol. These fit perfectly into the cups to clean them out.  Another alternative is a small Philips/positive drive screwdriver of a similar size.  One note about using Isopropyl alcohol to clean contacts: if they’re from a 3D printed truck, clean the Isopropyl alcohol off fully before refitting the pickup because 3D printed materials and Isopropyl alcohol don’t mix very well.

As well as cleaning the cups the wheel spike should also be cleaned and for this, I use a regular cotton bud dipped in Isopropyl alcohol.

The last issue is also to do with trucks and is something I’ve come across several times.  Below is another SD40-2.  The decoder is not loose in the chassis and the pickup cups are clean, and as you can see below it’s receiving power as the light is on.

However, if the front truck is lifted off the track the power stops.

Lifting only the rear truck shows the power returns so one or both sides of the rear truck is not working correctly.

This is one of the newer Kato chassis and the trucks simply pull out and clip back in.  When removing the trucks the drive shaft will also fall out, so be extra careful not to lose the bearing on the end of the worm gear as they will also fall off.  Should it fall off note that there is a tiny plastic washer between the bearing and the worm gear.

With the truck removed the problem becomes apparent; you can just about see it in the image above but it’s clearer below. The righthand pickup, with the cups in, has a post that stands up to make contact with the chassis copper strips and it’s bent.

Compared to the one on the other side you can see it’s pointing towards the truck rather than up.  This will prevent making contact with the copper strip and reducing the locomotives pickup by half on that side.  The solution is to bend it back into the right place and this can be done with a pair of tweezers or small needle nose pliers.

Now with the truck reinserted in the chassis, the locomotive should work with either truck on the track.  And with both trucks picking up power, clean pickup cups and a good connection to the decoder, the locomotive should run as well as new.

Clean wheels are also important, and a little lubricant on the gears is also a good idea while you have the trucks out.  If you’re wondering what oils or lubricants to use on the gears I’ve written a post about that which can be found here.

Cutting Out Etched Brass

Etched brass features in several of my model kits and has been used as a way to model tiny detailed parts for many years.  But how do you cut the parts out?

A sheet of etched parts, or fret, normally consists of a sheet of metal with all the parts attached by a half-etched tag.  This means the area around the part has been fully etched away except for a small tag which is only half as thick as the rest of the sheet.  Below is the etched brass ‘Additions’ sheet for my N scale Alco C-855B locomotive.  As well as the larger handrail section there are also several other parts, such as ladders and grab irons, which are much smaller and more delicate.

The half-etched tag serves two purposes; it marks where the part finishes and needs to be cut off, and it makes the actual cutting of it easier, as the material is only half as thick.

With thinner sheets and softer metal such as brass, the tags can be cut with a sharp knife.  I use what is commonly called a Stanley Knife as the blade is strong but sharp.

However, there’s a risk of damaging or bending the parts as the pressure needed to cut the metal is more than the force needed to bend the parts, particularly with tiny parts. Stainless steel etches are harder to cut than brass etches, because the material is harder. This is more noticeable especially on a metal or hard surface, and you may struggle to cut the etch at all. If you cut either material etch on a cutting mat the blade will drag the part down as you exert the force needed to cut through, resulting in bending the part.  The stainless etch below has some very small parts, the squares on the cutting mat are 0.5″ (12.5mm).

This stainless steel etch is from a kit by Keystone Details and zooming in you can see some of the tiny details.  Under the ladder is an electrical box that needs to be cut out and folded to make the box.

If I attempted to cut any of these parts out with a knife they would certainly get damaged.  So I use a special pair of scissors designed for doing this job.  Mine are made by Tamiya specifically for photo-etched parts.

The tips are small and curved which allows them to easily fit in the gap between the part and the fret so you can cut the tab releasing the part.

It’s not always possible to cut the tab off exactly where the part starts, so this little burr will need to be filed off.  Be sure to hold the part firmly between flat surfaces otherwise the filing action could also bend it.

Sometimes the fret is made from fairly thick metal.  The HO DT6-6-2000 etched brass Additions are made from 0.5mm thick brass compared to the N scale ones at only 0.25mm.  This was done to give the desired size of the handrail on the HO model, but it does mean the tabs are much harder to cut.

I use a much larger pair of scissors here with a strengthened set of blades for thicker metal.  These are ideal for removing sections of the fret to enable better access to the parts.

So I use a mixture of the Stanley Knife and scissors, depending on the part I’m removing.  It’s best to test cut a section of the fret that you don’t need, to gauge which tool is best.

This isn’t the only method for cutting parts from etched frets, there are also other tools for cutting photo-etch, however, I haven’t tried them yet because what I have works well for me. As modelers, we’re inventive in our use of tools and materials, but it’s helpful when we find a tool designed to get the job done, ie cut out parts that aren’t bent or burred.

Checking for Shorts when DCC Fitting A Wrenn Locomotive

The Wrenn Locomotives, despite being much older than most things you can get today, are still great locos and normally great performers.  They are not easy to convert to DCC but it can be done and I’ve previously written about the 3D printed sleeves I produce to allow you to do this.  But I sometimes get questions from customers who’ve done the conversions themselves, with my sleeves, and the loco runs very poorly even though it ran well on DC.  In this week’s post, I’ll show you what the most common reason for this is.

Coincidently this week I’ve had two Wrenn locomotives in for DCC fitting so I can use them to point out the issue.  The two locos, as you can see below, are a former LMS Duchess 4-6-2 and GWR Castle 4-6-0.

As well as being very different models visually they are different mechanically as well; the Duchess chassis at the back has a vertical motor and I covered the DCC installation procedure for this here.  The Castle has the horizontal motor and that was covered here.

Before I go any further I should point out one other issue which can cause problems with DCC fitting these locomotives and that is the current draw.  Sometimes older motors, and worn-out motors, can draw lots more current than intended and the DCC decoder can’t handle it.  To find out what the amperage draw is for a locomotive a stall test should be done.  You can read how to do this here.  Both of these locomotives had a stall current of less than 1 amp, so they are ideal for DCC fitting.

So what is the main cause of problems with these?  Starting with the Duchess below you can see I’ve cut off the wires as per the DCC install instructions.  Both motor brushers are still fitted and you can see them touching the collector on the armature.  The left brush holder, which is at the front of the locomotive, is isolated from the chassis, and the right, or rear one, is not.  The problem is often that the left/front brush isn’t totally isolated.  The brush still fits inside a brass sleeve which is wrapped in a rubbery paper-type material to create the isolation.  Over time, remember I said these were old, this material breaks down.  It’s possible heat from excessive running has affected it as well.  The material hasn’t totally disappeared and it’s not a dead short, otherwise the loco wouldn’t run at all, but a very tiny intermittent electrical short happens between the brass sleeve and the chassis.  Running the locomotive on DC doesn’t really affect it too much.  Although it’s not good for the controller, the tiny short will affect the running, but the controller is able to push more amps through the motor to compensate. But under DCC, the decoders are much more sensitive to shorts and are not capable of delivering as many amps.  The result is the locomotive runs very slowly or has no pulling power.

To check to see if this is going to be a problem remove the brush cap, spring, and brush from the left/front sleeve and inspect the insulation.

If it appears to be okay, basically not falling out, an electrical test with a multimeter can be done.  A continuity test, setting the multimeter to the symbol shown below, will check to see if there’s an electrical connection between the meter probes.

With the brush removed and the brush cap replaced, check to see if there’s anything between the two brushes.  This doesn’t work with both of the brushes fitted, as there’s a connection through the motor.  If, when performing the test, the multimeter gives the slightest suggestion that there’s a tiny connection there, it will cause a problem.

The solution for this is to remove the brass sleeve and isolating material and fit a 3D printed sleeve to the left/front as well as to the right/rear.  As the new sleeves are plastic you are guaranteed to have no short.  The customer’s Duchess above is actually in very good condition and is perfect with no sign of a short, so I won’t change the front sleeve, but once the decoder is fitted, if there’s an issue it will be changed.

The Castle with the horizontal motor can suffer from the same thing although it’s not so common. Both motor brush holders are at the back on either side of the motor.  Again I’ve cut the existing wires off but left the heavy gauge wire on the right, which runs from the connecting point to the brass sleeve on the right.  This is because it’s a better connection than relying on the spring to deliver the power.  The right-hand sleeve has the isolating material.

To remove the brush simply pull back the spring and it will slide off and the brush should fall out.

You can then do the same test as shown below.

I originally supplied my Wrenn DCC conversion sleeves in pairs to provide a spare incase something went wrong and one broke, but in hindsight I see it was a good idea as you may need to change both.  The sets I sell are:

Two Wrenn horizontal motor isolating sleeves.

Four Wrenn horizontal motor isolating sleeves.

Two Wrenn Vertical motor isolating sleeves.

Four Wrenn Vertical motor isolating sleeves.

Two Wrenn Vertical & two horizontal motor isolating sleeves.

This vertical motor design was also used in the Hornby Dublo locomotives, 2 and 3 rail, so should you wish to convert any of the locomotives to DCC or repair a DC locomotive which is shorting, the 3D printed sleeves will work.