Wayne Roderick, 3rd Division, PNR, NMRA (life)

08/24/98 rev 01/13/07


Malfunction Junction (MFJ) is where it all started in 1965 as my wife and I finished construction of our Pocatello home. The unfinished basement was inviting and I built some benchwork in one corner. It was not haphazard and this certainly wasn't our first model railroad, but it was the "big one", dreamed of and planned in the years prior. The railroad name of Teton Short Line was picked, but the name Malfunction Junction was still in the future to be suggested by an air traffic controller friend.

The general layout of MFJ comes from Dr Roy Dohns Victoria on his Victoria and Northern featured in MR Aug '62. Dr Dohn published several articles from 1958 to 1968 and then we lost track of him. That 1962 article was an inspiration for several features including the idea of having no legs under the layout. We have no legs under the layout- it is suspended entirely from the walls including the big center section. Thank you Roy.

To this day, I have never found a better yard design that I feel gives the mostest (operation) for the leastest (space) and so in 1978 when it was time upgrade and re-track it, I kept the original plan. Thanks again Roy.

MFJ Trackplan

figure 1

The yard was originally hand laid using salvaged nickel silver, code 100 rail spiked on homemade redwood ties. The switch machines were made from 48 volt telephone relays salvaged from a local scrap dealer at 8 cents/pound. The stacks of contacts made power routing easy. The turnouts used bench fabricated #6 frogs, the same as the main line- The mainline still uses them. It was cheap all the way, but it worked well.

In 1978, we relaid the yard in code 70 rail on store-bought ties and fabricated the turnouts to use 24" radius without regard to frog numbers. Spiking was done with a Kadee spiker, an amazing machine. This greatly increased the usable track lengths. We also replaced a pair of #6 switchs with a double-slip eliminating an ess curve in a most critical area. These improvements, and many more, control schemes and other ideas, we'd like to share with you. Perhaps you'll find some ideas for your own division point and service facilities. We are not going to talk about the scenery, as there is a wealth of information available on that topic.



In the scaled plan above or in the scaled control panel below, lets identify the tracks and their functions. Note that they are numbered on the right side from 1-10 counting toward the top. They're also shown in a photo.

MFJ Panel

figure 2

Track #1 is the Mainline. You'll see a West Bound Branch breaking away to take short trains up the hill to the Termite Timer Line (TTL). The control panel has occupancy lamps for the main line blocks adjacent to the yard limits.

the ARRIVAL/DEPARTURE (A/D) tracks, Tracks #2 and #3, might be considered as the most important tracks on the railroad because their length dictate the size of trains that can be reasonably handled and so greatly influence the size of passing sidings and ultimately the traffic handling capacity of the railroad. At the West end, find the double-slip switch (DSS) that accesses these tracks so conveniently from either the DRILL or the MAIN. This is a hot-spot of activity and must be reliable. You can't buy a good enough DSS for this- you have to build it! Details Teton Short Line double slip switch. As the name implies, these two tracks are where trains are assembled for DEPARTURE and disassembled upon ARRIVAL. The yard switcher (Goat) uses the DRILL track to sort cars on tracks 4-9 and then pulls blocks of cars to assemble on a DEPARTURE track. The goat has to select a caboose from the upper left caboose track and tack it on the built-up train, so he needs to be able to run around either end. Same situation for getting the power from the roundhouse. Obviously, the the A/D tracks must be double ended. The double-end and access to the main AND drill uses up a lot of linear space and whats left over determines the train sizes. We ended up with 10-12 car trains after the 1978 re-construction.

the COACH track, #4 is a handy place to park the CEO's personal business car near the TSL station. The remainder of it can be used for general switching work.

the TEAM track, #5 serves the TSL freight depot. A TEAM track is used by customers that do not have their own stub or service track and is essential to successful freight operations. You need a TEAM track at every town even if its only the end of a stub used for other purposes, such as it is here.

the YARD tracks, #6 thru #9 are used for sorting cars.

the SERVICE track #10 is a multi-function track. It can be briefly used for sorting, the far end is dedicated to a customer, it is the LEAD track for the a few customers on the local INDUSTRY, it is part of the runaround tracks needed for getting on the proper end of a caboose, and most important of all it is the LEAD into the roundhouse. It's very dynamic so don't plug it up!

the CABOOSE track on the West end of MFJ is obvious- What's not obvious is that this track has electric power on it at all times for the purpose of charging the batteries in the cabooses (cabeese?) necessary for the Utah Pacific marker lights on all of our fleet.

the COAL track is somewhat unique as the only way to reach it is to take a car of coal onto the turntable with a short switcher. Makes for one more fun operation once in a while.

the IN and OUT tracks can each handle two consists of two units. The tracks are electrically gapped midway with the left end controlled by a toggle switch and the turnout end controlled by the turnout. The differance between IN and OUT is really a matter of the surrounding support stuctures i.e. sandhouse, water, coaling etc.

the ROUNDHUSE has nine stalls plus seven more outside tracks. It along with the turntable is designed to handle 100 foot locos.

Together, these tracks make up a fascinating operations area and initially in the developement of the TSL, was "all their is". We had a mainline that went East to a balloon track for the first year.


Look at the control panel- How do you run this thing with arriving and departing trains, the goat switching and engines being serviced? Well of course, you use command control today- but we didn't do that in 1978 when we re-built MFJ. OK- where are all the toggle switchs on the panel that are obviously needed? They are few and far between, because we let the logic of turnout positions route the cab power. We go way beyond the simple concept "power routing frogs" to make it fun to operate without being a master of toggle switchs. We like to run trains-not panels! When a new-comer learns the yard, I give him a full 30 seconds of explanation and walk away with the words "have fun" and he does.


This is not built to our current standard The old 1978 panel served until 2005. It was built using novel techniques at the time, i.e. such as automobile tire valves for push buttons, layered construction etal. Most important to the long term success was the Logic chips (TTL) to greatly simplify operation. The explanations and techniques still have some value, if you choose to not make the conversion to Command Control.


After nearly 30 years of service and scabbed on modifications, the old panel has been retired. This was the final step in integrating Bruce Chubbs CMRI as our Layout Control Bus (LCB). Some long considered changes will now be very easy. Our new panel has only push buttons and switches that connect to ground and LEDs that source to the +12V buss. The logic described below is gone, being replaced with simple statements in the computer. We used a single CMRI SMINI for the interface. Chubb's SMINI will OUTPUT 48 functions for the panel LEDs and switch machines and INPUT 24 functions for push buttons. The new SMINI is the fourth and final "node" on our growing LCB The SMINI connects directly to the RS485 buss to the computer. The 48 OUTPUT functions were adequate because in most cases the 20 switch machines could share a line with one of the 36 panel LEDs. The panel also has several incandescent occupancy lamps. Even 24 inputs was just a little shy for our use, so the 29 route selection buttons were BCD coded to 8 inputs with TTL 74148s.


2005 I leave with the following, just in case you are "anti computer". You can still do it the hard way like I did back then. There is still value in the discussion but the hardware is sure simpler with the CMRI

Now we get into the fun stuff-no computers here, just simple logic using TTL chips and a few relays behind the scene so it's fun for your crews to operate. You're going to have to do some work to get it simple to run, but it can be a challenging and rewarding task. Perhaps a schematic of the whole yard, plus another of the A/D and SERVICE tracks will tell the whole story to some. Their are many details that you might get an idea from so we'll add a few of those with some words of explanation toward the bottom.

MFJ schematic

figure 3

In this schematic, figure 3, note that their are many LED's indicating turnout position, but only enough push buttons for route selection. The push buttons are the small circles with a number or letter inscribed. The essential toggle switchs are marked with "S" in a larger circle. The lower right one (block 1) is a center off that can select MAIN or YARD cab(s) for power to the main/block24 track. The toggle for the DRILL track works the same. The toggle at the BRANCH selects the BRANCH cab or whatever is connected to Block 24 for a short buffer block between MFJ and the Termite Timber Line. The remaining toggles are simply ON/OFF. The toggle on Track 10 kills all the black colored track extending from it and the two toggles on the left allow us to hold four engine consists on the IN/OUT tracks at the roundhouse. All turnouts are of the power routing kind, enhanced with extra contacts on the switch machines. This simply means that if the turnout is against you, the track is dead!


This is an old trick that I discovered (perhaps others did too) and popularized in the early 70's. It saves a handfull of toggle switchs. It works like this in a double ended track like tracks #2 & #3.

If only one turnout (#4 or#9) at the ends of the track is aligned, then the block boundry extends to the far end-BUT if the turnouts at both ends are aligned, then the boundry is moved to the center (oval mark). This is amazingly easy to do with cross connected contacts or slave relays as shown in figure 4. Imagine a train pulls all the way into the track via the aligned switch at one end, then simply by aligning the other end also, the block is split and the yard goat can come in from the rear to do work while you disconnect the power and head for the roundhouse. Each time you do this, it eliminates two toggle switchs.

The second floating boundry is marked by an oval to the left. In this case the position of turnout #3, that is part of the double slip switch (DSS), determines if the A/D tracks get power from the MAIN/block 12 or the DRILL track, again eliminating toggle switchs.

In the goal to minimize panel controls, this is the test:

Is the logical position of a anticipated toggle switch related to the position of one or more turnouts? If the answer is yes- get rid of it!
Lets run trains-not panels.

MFJ relays

figure 4


Back in 1972, when we published the First TSL signalling system, I had made a mistake that would take some work to overcome. That signal system used a NEGATIVE 12 volt supply, My mainline switch machines used 28 volts AC and my yard switch machines (telephone relays) used PLUS 48 volts DC, and there were others. Nothing was compatible! After some research, I decided to adopt the relatively new, but widely accepted standards of TTL logic and strive to make every device on the TSL work by pulling a TTL standard signal of 1.6 ma to ground and to actuate that device by sinking its load current to ground. It didn't take long to swap out the transistors and diodes in the signalling system and feed it from a POSITIVE supply. More work was required to drive the 28 volt AC slow motion switch machines on the mainline from a TTL signal, actuating them with a push button that momentarily sinks 1.6 ma, and it it was a real challenge to drive the 6 AMP rotary switch machines in the new MFJ yard. This effort would pay off in years to come as the computer entwines it's tenticles into the Teton Short Line. The relays you see in figure 4 have a TTL7406 buffer amp driving them, but not shown. The following figures from our files depict how standard TTL can be used to simply the panel controls. You may choose other logic standards such as low-power TTL or CMOS, but I suggest that you pick a standard and work toward it for the long run. Today, as I'm integrating Digital Command Control (DCC), while keeping many conventional options, the task is much easier because of our logic standard.
logic02 Simplest TTL logic for MFJ. The two NAND gates form a flip-flop that stays where it was last set with a momentary button push. Output connections are provided to the switch machine and to the power routing relay(s) shown in fig 4. The machine moves one way while a logic LOW is applied, returning to normal when the logic goes HIGH
MFJ logic12 Not a flip-flop, but a tri-flop with three stable states. Driving two switch machines to select one of three routes. The picture also includes a reset device that initialy sets all the switch machines to NORMAL when power comes on. This prevents the heavy current load if they were to come up random, requiring several machines to work at once.
MFJ logic03 Diode steering with TTL. Now we're really building with Tinker Toys and making things behave the way we want them, always striving to eliminate uneeded panel control. This logic handles the complex at the West end where the double-slip switch (DSS) resides. The DSS uses two machines to control the route in and the route out. One of them, #3, is part of this complex.
MFJ logic04 More fun. We have combined diode steering with two tri-flops and one flip-flop for the LED route indication, then added a couple NAND gates to drive the appropriate switch machines for the sorting yard and the service track lead.
The rotary switch machines in MFJ are ancient (WWII) surplus actuators advertised in the MR mags by TAB for many years. They draw about 6 amps at 12 volts for an instant while they snap rotate 30 degrees and then hold with only 200 ma. There was also a 24 volt version. You old-timers must have boxes of them. Working since 1978, they last forever (if you thought to put a diode arc suppressor on the cutoff contacts). John Lukesh done a complete writeup in the NMRA bulletin June 79, showing TTL interface. We'd done it differant, with overcurrent protection to make it safer, but the results were the same.
If you're not already into the world of Logic, then run down to Radio Shack and grab a book on Logic Circuits- It's a fun extension of our hobby to make our controls friendly, so we can run trains instead of control panels.
MY FINAL ADVICE: Do the LED and turnout logic, but forget all this extra relay and toggle switch stuff for power routing and go directly to Command Control, Do not pass GO, Do not collect $200.

Teton Short Line Home Page