Kato Unitrak Double Crossover Electrical Improvement N Scale

[u18b]

I have found is not a rare thing on internet discussion boards that someone complains about their Unitrak double crossover in N scale.




I have owned one for years and have also encountered problems.   Kato Unitrak is generally often dubbed “bullet-proof”, but I’m not sure that applies to this item (especially depending on how you use it- including the modeler not using properly).

Here are the problems I’ve encountered with both DCC and analog.

--occasional stalls, usually around the middle.
--weird electrical surges.  My loco would enter the crossover at a relatively slow/average speed and then speed up on the other side.  Clearing indicating a voltage drop or increase.
--poor electrical conductivity.  Sometimes there would be intermittent conductivity.  I even saw this with a crossover on my workbench.  Using a multimeter to check for continuity, there was an intermittent loss.  I knew this would need to be solved.
--or complete lack of continuity.    Unitrak in general is like puzzle pieces.  You can change layouts as desired (when it is not glued down).   Sometimes the new design would lack power and I had to add jumpers.  This is not so much a flaw as much as a frustration of something I was not understanding.

Recently, I acquired two more crossovers and I decided to improve them all and make them more reliable.   So this tutorial will show how to take them apart and significantly improve the electrical reliability.   My focus is on electricity, not mechanics.  So sharpening the switch points and adjusting the screw that holds the points for fewer derailment and point-picking is not covered here.   My concern is almost exclusively with the internal guts.

First, let’s look at how this thing actually works electrically, because THAT is the source of a few problems I and others have experienced.  In short: I discovered I was not using it properly.  Plus I discovered other problems.

Bottom line (without modifications):  You need SIX power attachments (jumpers) if you want these double crossovers to run properly.  Yes six.

Put an isolated crossover on your work bench and use a multimeter and you will find that if electricity is applied to TWO rails (one pair of the four pair), then 3/4 out the turnout will be electrically dead.
This honestly surprised me.

Now a work bench is often different from a layout since a layout is often a loop.  Applying electricity to two rails in a loop will send the juice around the loop back to the other side.

I assumed that Kato designed a crossover that was entirely powered.  I was wrong.

I have a plywood plains layout (just track on a flat surface).  I have a three track mainline so I can run three trains.  The track is DCC or analog:  changed by plugging into the required source.   

This this is a temporary layout, so I only have ONE power source with no power bus and no feeders running everywhere (and all turnouts are manual- no wires everywhere).  This meets my needs to a) run trains for fun;  b) let me focus on building/repairing stuff (and testing locomotives) and c) writing articles.   I hope to one day build a permanent decorated layout.  But for now, this is it.

Now, a long time ago, I discovered  I needed to add jumpers from one track to the other.   But after this latest construction and the problems described above, I set out to know exactly how these crossovers work.

[u18b]

So, with a multimeter, here is what I discovered.
See the following drawing testing on a work bench (not a layout):




In this example, with points set to straight, when I put a probe on the bottom pink rail…. That rail is the ONLY rail in continuity in this whole assembly.

Likewise, with the matching rail (light blue), you can see that only a couple of rails are powered.
What I discovered is that the top half and the bottom half are completely and totally isolated.
Further, the right and left half is also isolated.

Flip the points to crossing and you get this:





Now, in this situation, the blue is only powering two rails. That’s it.   

Now, in defense of Kato, this crossover would work well in situations where the blocks required the tracks to be isolated.    But I have no such need.

So this all explains a couple of things. 

1.  In order to run most reliably, this piece of track requires electricity to SIX of those eight end rails.  The very top and very bottom rails are solid and only require one jumper each.  But ALL four of the inner rails need power for this to work properly.

2.  Now I know why my locos were surging and slowing (depending on direction).
Imagine this in a layout.





Power from the power pack enters at point A.   But with all the points set to straight, electricity does NOT get to the B side in the turnout.  But since I have a loop  and not a point to point layout, electricity travels from A counterclockwise all the way around the 9 foot layout (thus traveling about 18 feet) and comes to point B.  Thus I have a voltage drop between A and B.    A loco traveling from right to left enters the crossover at B going slow, getting it’s juice from the back of the layout all the way from point A.  As soon as the front wheels hit the blue rail in the center of the crossover, the locomotive speeds up (surges) because the voltage drop is gone.   

3.  This also partially explains some stalls.  If this track is not fully powered, then a small middle rail might not get the juice it needs.  This may also have to do with the power routing slider inside and/or the track power connections.


Now that I understand this crossover, I’ve concluded that for my needs, there is no reason for this lack of power.  One of the good things about DCC on a small layout like mine is that there are no blocks.  All track has power all of the time.   I see no reason why this can’t be the case for this double crossover.

Looking at this photo below…




So here is what I’m going to do:   I’m going to solder  INTERNAL jumpers connecting:
1 and 3.
2 and 4.

After I do this, almost half of the turnout will now be powered with only the two small inner rails power-routed internally.



I will also do the same on the other side.   5 will jumper to 7,  and 6 will jumper to 8.
When this is complete, all external rails will be powered internally, unified,  and no more voltage drops will occur.

[u18b]

Let’s get to work.

Turn the crossover up side down and remove three screws from one side that retain the cover plate.  I’ll only describe the procedure for one side, since the same procedure will be done on the other side.



With screws removed, getting the back off is a little tricky.  There are four locking tab projections on the ends, so the middle must be lifted without bending the back.   I use a knife to gently lift the cover plate without damaging anything inside.





Here are the locking tabs.



Once lifted up a bit, I can flex the back and get one side of tabs clear.  Now the cover plate is lifted off easily.



Be careful, because there are loose moving parts in there.  As you can see in the next photo, you do not want to damage the coils.





There are a total of four circuit boards in this assembly (one per turnout).

Here is a close up of one board.  I immediately noticed that the pickups to the rails are NOT soldered to the board.  Tabs are folded over and press fit.     My plan is to solder the tabs for a more secure connection.



How does this assembly get power to the rails?    The screw with the pivoting portion sends power to the frog.   The fingers send power to the rails.

In the photo below, you can see that there are a total of 8 power connectors to the rails.  These fingers are welded to the bottom of the rail.  However, I found that these can come loose.  More on that later.



One more orienting note.   In the photo above, you will see the black sliding turnout actuator.    That actuator has a wire running under it.   On end goes to the turnout points, the other is anchored in the middle.   These wires will probably come loose, and you’ll have to put them back.

Below is the wire controlling the turnout points.


And here is where they anchor.   There is a square cover plate that comes off very easily.  Notice the orientation.




And here is the plate removed showing the wires anchored properly.



This is what you’ll have to correct.

[u18b]

OK… Let’s get to work.
We do not want to get flux or solder on the circuit board pads where the power routing happens.
So I use painter’s tape (easily removed) to cover the areas.   I only need access to those folded tabs.




I used a very fine tipped soldering iron to get in tight spaces.




 I added flux on the far edges of the tabs and then added solder.




O I’ve soldered all four tabs per circuit board.
Is this wise?    Am I ever really going to remove this circuit board in TOTAL disassembly?   I VERY highly doubt it.  So- No.

Now, be careful, you don’t want to be blocking the pivoting plate that power routes the frogs and internal rails.  On the middle tabs, I’m careful to solder in the far corners.   I also don’t want to touch those fine wires powering the switch machines.  (By the way, since I only manually use this crossover, I remove the red/black power wire.).

With the tabs all soldered, I then measure, cut and solder jumpers wires as shown.
Here is the first jumper installed (gray), from one circuit board to the next.  Top to top.




And here is the second jumper (orange).  Bottom to bottom.  I used different color wires in order to keep my thinking straight and avoid a short.




The other tabs go to the power routed middle rails- so no jumpers for them.    I then carefully peeled off the tape.

[u18b]

Wait, while this is all open, there is one last thing to do.

Remember, I said that I sometimes had intermittent powering?  I discovered that the power tabs to the rails are not soldered. They seem to be tack-welded.

Sometimes those joints come loose.  Here, I’m using a blade to gently lift on a finger to check the joint.   This one is tight.




The ones to the inner rails were all tight. 

But the tabs to the long outside rails were not tacked and were only gently touching- NOT reliable. 
Here is a top-side view of one that was loose.




I decided to carefully solder any that were loose.   
Admittedly, with all the jumper wires I added, this may be overkill (but, hey,  it made me feel better knowing I had a better connection).

But the broader point is this:  if I had a tab come loose on a solid outer rail, if it was on one of the little inner rails, there is a great opportunity for a dead spot.

Room is VERY tight.   I have a soldering iron with a very fine pointed tip to get at places like this.

In this view, you can see that the power fingers to the outer rails are located under the black plastic actuators. 




Since I had a loose connection at one of these spots, I removed the actuator.



Space is very tight to get in there. 





So I tried soldering the joint from the top-side.
This one came out great.




Another one did not come out great.  The solder flowed up onto the outside of the rail and would have to be scraped off.



So I decided that maybe it was better to solder the joint from the underside.

It melted the plastic a bit, but it is under the rail and you cannot tell from the top.




Having completed the repairs and mods, I reassembled that half, did some checking with my meter to make sure I had no shorts, and then repeated it all on the other side.

Now, I have full power to all eight main rails of this crossover and it can be provided by only two power leads from my power supply.

And since I have two of these crossovers for three main lines, my two power leads are still sufficient to get electricity to all of the lines.  Without these modes, I’d have to have external jumpers to power the inner rails.

After a few weeks of living with these mods, I’m having no more power issues like I was having before.

[Carolina Northern]

Ron,

This is a good write up, but I have always used the six jumpers that you referenced. You don't even have to open the turnouts. Solder the jumpers on the ends to the underside of the rails.

I used them like this for many years.

My recent loss of function in my hands has made me leave N Scale, so I no longer have examples to show, but I'll happily answer questions if this isn't clear.

Don

[Maletrain]

Those are really excellent pictures that I am bookmarking in case I ever need to open one of these.

But, I just provide power feeders to all 4 tracks connecting to this turnout, and never have any problems.  In effect, they are doing the same thing as your internal jumpers.  I would think you could do the same with external jumpers, with less risk of messing up the innards.  But, if you really were having problems with the press fit of the tabs used to route power as the turnout is thrown, then this procedure does address that nicely.

One comment about your 2-wire-only feed to power your whole layout:  The slow downs and speed ups that you were experiencing are a good indication that you need more feeders to your loops.  Yes, these jumpers will smooth out the differences by removing the electrical gap that separated 2 sections of rail in a path, and even increase the voltage a bit by providing 2 parallel paths to every point on a loop, but I expect that there will be substantial voltage drops at the farthest track distance from your feeder wires, because the track and rail joiners are not going to conduct as well as a bus wire.  At least that was the effect on my 2-loop 9' x 3-1/2' test layout made with Unitrack.

Even if the resulting slow-downs are not noticed or not annoying, the voltage drops may (or may not) interfere with your overcurrent protection.  I suggest that you at least do the "quarter test" and short your rails at various locations around all of your loops to make sure that your power source trips its connection to the rails.  If it does not, then you should add some more feeders until it always does.  Otherwise, an unnoticed short on the track could cause a hot spot that can damage track and/or equipment.  The hot spot does not even need to be at the location of the short - it might be (mostly) on the loosest rail  joiner between your feed point and the short.

[u18b]

Ron,

This is a good write up, but I have always used the six jumpers that you referenced. You don't even have to open the turnouts. Solder the jumpers on the ends to the underside of the rails.

I used them like this for many years.

My recent loss of function in my hands has made me leave N Scale, so I no longer have examples to show, but I'll happily answer questions if this isn't clear.

Don

Hi Don, thanks.

Yes, you could add the jumpers from the outside.  Space is tight (if you want the crossover to lay flat).   There is just a little bit of room as follows.



Yes, that would work.   But it still does not address possible poor connection at the folded tabs.  Nor does it address the loose connections under the rails I discovered.

[u18b]

Those are really excellent pictures that I am bookmarking in case I ever need to open one of these.

But, I just provide power feeders to all 4 tracks connecting to this turnout, and never have any problems.  In effect, they are doing the same thing as your internal jumpers.  I would think you could do the same with external jumpers, with less risk of messing up the innards.  But, if you really were having problems with the press fit of the tabs used to route power as the turnout is thrown, then this procedure does address that nicely.

One comment about your 2-wire-only feed to power your whole layout:  The slow downs and speed ups that you were experiencing are a good indication that you need more feeders to your loops.  Yes, these jumpers will smooth out the differences by removing the electrical gap that separated 2 sections of rail in a path, and even increase the voltage a bit by providing 2 parallel paths to every point on a loop, but I expect that there will be substantial voltage drops at the farthest track distance from your feeder wires, because the track and rail joiners are not going to conduct as well as a bus wire.  At least that was the effect on my 2-loop 9' x 3-1/2' test layout made with Unitrack.

Even if the resulting slow-downs are not noticed or not annoying, the voltage drops may (or may not) interfere with your overcurrent protection.  I suggest that you at least do the "quarter test" and short your rails at various locations around all of your loops to make sure that your power source trips its connection to the rails.  If it does not, then you should add some more feeders until it always does.  Otherwise, an unnoticed short on the track could cause a hot spot that can damage track and/or equipment.  The hot spot does not even need to be at the location of the short - it might be (mostly) on the loosest rail  joiner between your feed point and the short.

Maletrain,
Thanks for your suggestions.

Here is my current setup.  With two modded crossovers next to each other, every track is powered now by two incoming wires.




Here is what I'm trying to avoid.   Here is an old jumper I used before the mod to power the back rails.   But it is ugly.  And it can be an obstruction for trains passing by.
Making my mods eliminates this.



Thanks for the warning about safety.... but my Prodigy system works perfectly.   It shuts off fine at the farthermost point from the power.

[Carolina Northern]

Ron,

I understand what you are saying about the loose fitting and possible loos of connectivity. I probably never saw those problems for two reasons.

The double crossover has been out a long time. there may be some wear in the molds, making things not fit as well.

I may never have seen it because I used Neolube on all my rails for weathering. The conductive nature and slight lubricating action may have masked or prevented that problem.

Mine worked well as long as they were perfectly flat. Any twisting or unevenness will hang them up.

I've always attributed that to Kato's exact engineering. Those coils are just strong enough to do their job, no wasted power. Wonderful if everything is correct, but sometimes it's nice to have a bit of overkill.

I always used Ray Stillwell's BCD circuit to throw my turnouts. I still haven't found anything better. I still use it now that I'm doing HO.

Just a note, even in HO, Kato's #6 is a better turnout than the #4 equivalent.   

Don