Train-Li Track Products

(remember: links to sub-pages at the bottom of this page)

click to jump to the bottom


(Note: Train-Li has many more products than track, check them out! )


Train-Li is in the US, run by Axel and his wife, both very enjoyable and "up front" people. Axel has many years of experience working with LGB, and "knows" all the ins and outs of products from Germany and Austria and Switzerland. Train-Li supplies many products, including being the sole importer in the US of the Train Line45 products made in Switzerland.

Besides the track products highlighted here, they have many detail items, figures, etc. 

New "R7" switch

I have not tried these yet, made by Train Line 45 of course. Available in brass and nickle plated brass. Appear to be same excellent construction.

There are some initial questions about fuses for powering the frog, but Axel is intending on replacing the one-time fuses with self-resetting breakers.

The power is routed to the frog by a reed switch, which is triggered by magnets on the moving throwbar assembly. This reed switch is more substantial than the ones you see for triggering sounds. Since it is glass, you can see significantly larger contacts.

I'm not sure what "R7" means, because LGB never made a R7, The actual curvature is very close to 12.5 feet. This makes it like "R4" in LGB parlance. In LGB track, R5 was 15' diameter track, and the R5 switch had a #5 frog.

There is something that Train-Li did that is very smart, in my opinion, trying to fill the LGB "gap", and also make a switch for "normal" designs.

Train-Li radically shortened the length of the track on the diverging route, after the frog. What does this do? It allows you to use the switch in two different ways:

  • Add curved rail right after the frog, and you can emulate the "curved" switches of LGB and others, where you CAN place the switch in a continuous curve.
  • Add straight rail after the frog, and you have a "traditional" turnout, that has a fixed angle diverging from the frog.

The frog "number" is between 5-1/3 and 5-1/2, not a #6. 


Replacement frogs for the Aristo Wide Radius switch

The original Aristo frogs had several problems. In 2009, Aristo updated the design, but it still needs some improvement. It is an all plastic frog, subject to wear on the frog point. See the section on "Aristo Track".

The Train-Li unit is all stainless steel, no plastic at the wear points. The consistency of the wing rail flangeway widths and depths is far superior to the stock plastic frog. Of course, you pay for quality.

Here's a picture of the Train-Li frog by itself:


So the frog point, flangeways and wing rails are all stainless steel, and very well made. The flangeway depth is 3.01 to 3.02 mm. The new Aristo frogs are 3.18 to 3.28 depth, and vary quite a bit (I measured 20 of them).


Here is a view of the underside, note the tapped hole so you can feed power to the frog: (a stainless screw is provided)

Here is a picture of it installed in an Aristo WR switch:

The small clamps are provided with the kit. The guardrails in this picture are stock, Axel just painted the top edges silver to match the SS rail, so the "clamps" are just decoration here.

R.J. DeBerg reports that the clamps DO help hold the frog "down in place" better than with just the stock mounting screws.

Note: this kit does not cure the problem of the guard rail flangeway widths. The guard rail flangeway widths should still be set to 0.106", I use shims, see my Aristo Wide Radius switch page .

I have not installed mine yet, will report back here when I do. The conversion kits come with a nice instruction manual in color.

Drop in frog insert for Aristo #6 switches

Train-Li makes a drop in insert for the #6 frog. See the Aristo track section under the #6 or the #6 wye to see more details on these switches.

The #6 Aristo frog flangeway depth is very deep. I measure 5.97mm in several places. This would be fine if the geometry of the frog was correct. Unfortunately it is not, and because of the poor design, wheels will "drop into" the frog throat. This results in excessive wear, derailments, noisy and bumpy operation. Since the frog is made of poorly plated pot metal, this is a problem.

Train-Li makes a stainless steel insert to fit into the Aristo #6 switch.

With the insert in place, I measured exactly 3.00 mm. flangeway depth, which is the minimum specification per the NMRA, and not coincidentally, the depth of the Aristo Wide Radius switch frog.

I purchased mine when Train-Li first made them. I've used them since 2008, and this is written in 2010. 

The ends of this insert have a short bevel on them, it's a bit abrupt, extend the bevel to about twice the length, since the flanges of many locos and cars will contact it.

The inserts fit perfectly into #6 left and right switches. They do not fit into the #6 wye very well, but if you place it into the wye frog and use a screwdriver and light hammer, it will spread to fit the frog ok. It's a lot better than with no insert. There will be some gaps at the frog throat. (This is because the Aristo "#6 wye" is NOT a #6, but more like a number 3.)

Unfortunately, the Train Li insert turns the switch into a flange-bearing frog. So if you have very deep flanges, you will still get a "bump" (but no damaging drop into the frog or frog point wear). 

I would advise checking your flanges on your rolling stock, and filing them down if they are way in excess of 3mm. For reference, all Aristo locos have flanges about 0.124", 3mm is 0.118". USAT locos have much smaller flanges, but the rolling stock has deeper flanges.

I have these inserts in all of my #6 switches, and what a difference! Highly recommended. $10 each.


ProDrive - DCC switch motor:

Available from Train-Li is the Train-Line45 DCC switch motor. It is powered from the track, and has a small motor to move the switch points. These are new in early 2009.

There are 4 CVs to program, and they must be programmed on the programming track, and only in limited modes. More on programming modes later. They are stationary/accessory decoders.

There are about 3 firmware versions, each with slightly different requirements.

I use an NCE system, some systems will not program these at all, and some other systems seem to have no issue (more about this later). NCE supports Direct, Register, and Paged mode CV programming in service mode.

Programming: (first 2 generations) 

  • First, you need to use register or paged mode on the programming track to program these decoders. They will not program on the main. Some systems must be directed to use register or paged. Overall you will have most success using paged mode.
  • Zimo systems try direct programming and if it does not work, drops back to what appears to be paged mode.
  • NCE controllers work well with STD mode, which will select register or paged depending on the response from the decoder, I recommend this mode.
  • Again, they will NOT program on the main, and you cannot use the normal "direct" "CV" programming.
  • For the first 2 generations, you needed to put a load on the 2 terminals inside. Initially, I used a DC lamp that drew 60 millamps at 12 volts. A friend and I tried a lamp that drew half that and it did not program. A lamp, though a traditional load, is really a bad idea, because it is just used for resistance to load down the programming output. Running a light bulb at an unknown voltage does not let you know what current you are drawing. I did some more experimentation with resistance values and used several of these units. I settled on a median value that succeeded on each unit I tried, 50 ohms. I tried many values, and the "friendliest" unit started working at about 750 ohms. This is a pain in the butt.

Programming: (third generation) Firmware version 24 and above

For the latest generation, no load is required, but there is an additional problem, it won't program on most US DCC systems. It took some figuring out, but apparently European systems traditionally have a solution.

These decoders come stock in "analog mode" where they respond to DC, and NOT to DCC. Internally this is by setting the address to zero. You will know the decoder is in "Analog mode" because it will "fidgit" back and forth when on DCC.

To be able to control and program the decoder, you first have to change the address to something other than zero, the big problem is doing this using standard NMRA programming.

Basically, you need to briefly interrupt the connection to the programming track right after the command is issued to get ANYTHING changed. (It turns out that most European systems DO this, in fact some, like the Zimo have programmable timer for the interruption).


My technique is to use the STD mode on an NCE cab (you can use paged mode too). This mode tries to read the decoder manufacturer ID, and the decoder version, and then goes to do some basic configuration, and the first item is setting and activating the short address.

Start the process, and let it time out on reading the ID and version.

Now you get to the point where you want to set the short address, and then activate it. You can enter any address from 1 to 127 for the short address, it does not matter.

Press enter after the short address. Then you will be prompted to enter 1 to activate it.

Just after hitting enter, briefly interrupt the power to the decoder.

This will set SOME address into CV1 and set CV29 to recognize the short address.

Now exit and re-enter the STD or paged programming modes, you will not be able to read the manufacter id, let it time out, but you will be able to see the decoder version.

Now you can set and activate the short address. You can also set the other CV's (35, 40, 46) as desired.


  • On Aristo turnouts, the motor will only attach to one side, and the default setting on the direction CV (CV40) is zero, which is right for left hand turnouts. Set CV40 to 4 to reverse the direction to be correct for right hand turnvouts.
  • CV46 gives you control over the speed of motion, 12 is the slowest speed, 44 is the highest speed.
  • Sometimes the address (CV1) changed after programming, so it's best to check it again after programming the other CVs.
  • CV35 is new, it will send power to the two internal pins to light a lamp when it receives an address. Set CV35 to the address you want it to react to. Most people will set it to the same address as the decoder, i.e. CV1.
  • You must disconnect power from the unit for 10 seconds after setting CVs, although CV1 seems to take effect immediately.


  • Check the throw of the unit before bolting up to the turnout, to make sure you have the right direction.
  • CV1 is the address, from 1-255, default is 1
  • CV35 is the address of the "light output pins" goes on and off in relation to turnout position. I believe you need firmware version 10 or above for this function. (Cliff set it to 99, why?)
  • CV40 is the direction, values 0 or 4, 0 is default.
  • CV46 is speed, 12 (slowest) to 44 (highest)

Note, latest version of firmware seems to be version 25


Below is a picture of an early unit. Along the top edge you can see the 2 wires that go out through the case for connection to the track.

You can see the small motor and the decoder board. On the lower right edge of the board you see a row of 5 solder pad, just above the 2 left ones is the 2 pin header where you connect an 18v lamp if you have programming trouble. This header is also where the upcoming optional light will connect. I was not able to get this output to light the 18v bulb.

The picture below shows the throwbar assembly removed from the motor and you can see the offset "crank" on the end of the motor that moves the throwbar.

The picture below shows the throwbar removed. There are 2 springs (not shown) on the throwbar that allows some "give" in the movement and keeps things from breaking if there is a jam.

The picture below is a closeup of where the 2 pins (track power/signal) come out of the housing. They are siliconed to keep moisture out.

Below is the newer version. There is an additional 4 pin connector that allows manual (pushbutton or reed switch) control of the switch. There is also an electrolytic cap that allows the operation of the switch drive with half wave AC as with the old style LGB switch controllers. The waterproofing of the servo is now dark blue and appears to be a dip or conformal coating.



Boehler switch motor, and Aristo-Craft "clone"

Boehler makes a slow motion switch motor, and Aristo-Craft has made a blatant clone of it.

I got the following technical points from Axel of Train-Li, and after reviewing the points and the pictures, I agree. In my opinion, you get what you pay for, and the lower price of the Aristo clone is from "short cuts" in the cloning, and cheaper materials and parts.

1. Motor armature wiring:

Boehler/Train-LI-USA has enameled coated copper wire for durability and heat insulation
Aristocraft has a clear coat copper low durability, poor heat insulation. This coating should never used for motor armatures.

2.Internal contact switch

Boehler/Train-Li-USA has a sealed Delrin package with high interior side walls to keep the gold coated contacts in place.  The contacts are plated with a thick layer of gold on the circuit board as well as the contacts themselves to ensure uninterrupted operation.

Aristocraft has a cheap half shelf plastic with very limited sidewalls and steel contacts which lead quickly to contact problems on the contact surfaces of the motor circuit board.

The picture below shows the Train-Li contacts: (note the plastic "walls" between them that keeps them straight and aligned)

Here is the Aristo contacts. Notice how the contacts can "skew" in the housing?


3. Two translation gears

Boehler/Train-LI-USA: High quality Delrin gears with self “greasing” ability hence the entire Boehler drive does not contain any grease

Aristo uses cheap low tolerant plastic that require a lot of grease.  The grease itself can easily (especially in the summer) melt on seep into all ports of the internal drive, inhibiting electric contacts, and potentially quickly destroying the motor.  And once grease is out the the gears the gear might just quickly strip.

Here's the Aristo, all plastic gears except the motor, and not nylon but something else:

The Boehler gear train looks better, and you can tell they are real nylon gears:

4.The chassis design

Boehler/Train-LI-USA uses two half shells that have definite sidewall overlap to prevent water from coming into the drive.  The two half shells use a rivet to be held into place, therefore no damage can be done to the chassis.

Aristocraft: uses two half shells with less than perfect side wall design and the two shells are screwed together which leads to easily breaking the chassis when tightening the screws when factory assembled (you can see the bottom boss is already split)

5. The switching shaft

Boehler/Train-LI-USA: Both shaft ends are housed in a guided brass bushing – keeping them perfectly align within the chassis.

Aristo: The two used brass bushing have no outer rim which keeps them in location, hence there is no guidance which leads to play within the chassis

Here's the Boehler:

Here's the clone:

The "centering spring" that allows the travel of the drive to match the travel of the turnout is visible. There have been several reports that the Aristo spring had bent under pressure (from running through a switch the wrong way) and the spring did not recover and now the unit does not throw in one direction all the way.

6. Manufacturing process

Boehler/Train-LI-USA drives are made and assembled in Germany to high quality standards.

Aristo manufactures in China with low quality standards, hence the cold solder as seen here on a brand new motor connector


You get what you pay for, and I believe all the short cuts that Aristo took will lead problems in operation and reliability.

Given summer heat the bad insulation of the motor armatures wiring will either shorten the life of the motor drastically or will in some instances create even meltdown.



Train-Li imports some high quality track and switches from Germany. 

They have brass, nickle-plated brass and stainless steel (I have not seen the SS stuff yet).

Like many Train-Li products, an easy way to think of them is what LGB would have done in quality if they took their products "one notch higher".

Here are pictures of an R2 and an R3 switch. Notice the quality, and the improvement over LGB in the metal frog insert?

Weather Underground PWS KCACARLS78