MTH 90-Degree Crossing Design Flaw?

Please let us know how your troubleshooting progresses.

We had a prewar 30 degree crossing at the TTOS mall layout years ago. It was American Flyer O gauge and the center diamond was a huge piece of phenolic. It was so big in fact that what ever roller was crossing it would be dead for too long of a time, causing the occasional stoppage of the train.

We ended up putting a diamond of sheet metal on top of it connected to the center rail.

Hi TRR,

I looked at the track plan and see the curved approach to the 90° cross. The SD-70 has two collectors that appear to provide sufficient spacing to compensate for the hot or the wheels being insulated momentarily.

I have not had one in my hands, so I don't know what MTH does underneath electrically in their cross. Would you post a photo of the bottom? You may have to solder some jumpers to provide connectivity through the cross.

The other setup I don't like about MTH RealTrax is that only one outside is connected to common. I have been soldering a jumper to the common contacts on a layout that I am restoring, so that both outside rails are connected to common.

perhaps, if you put a lock-on on a track on the loop inside of the cross with the opposite outside rail to the first lock-on. Then run wires from the first to the second, color matched , of course.  The 90 may be isolating that inner loop.

Hi RoyBoy and Carl,

Thanks for your suggestions.  Last night as I was shutting the track down I was thinking I may need to add a wire somewhere but hadn't figured out how to handle the perpendicular connection requirements. A diamond shape on top is a great solution and one that might be easy to test.  I'm off to work but will disassemble, explore, and photograph the underside tomorrow when there's better light than when I get home.

I did not know about only one common being connected In MTH track. I wonder if that's why my Lionel engine derails?  So far the MTH shuts down at the cross going in let's say the West-East direction of the layout, where there is little curve in the track.  I'm 75% sure that the Lionel engine is fine there but it jumps the track/shorts going in the Sough-North direction just after the switch. The engine is shorter but given it's pulling a tender, the overall engine lengths are comparable.  Since there's so much going on in that part of the track, I'm not sure the cause.  For example, the switch seems to have a high-spot (the frog?) and engines/cars jump a bit. Maybe with the curve and cross it misses the rail with the ground. Or maybe it just derails, period.

Also, Carl when you say that you believe the two collectors on the MTH should be far enough apart that the plastic spot on the cross shouldn't matter, does that mean that the rear one is in play even going forward?  Based on looking at the undersides of both  engines, I just assumed they were directional -  the front one for the positive connection forward, the rear one for the positive connection moving backward.

Obviously, I'm new to this so when you say a "lock on" do you mean I need to add another transformer or just a second set of wires from the Z-500?  I ask because at the inner loop there's a left hand switch and I've been having trouble with it. Some tests this AM are making me suspect low voltage.  That's a topic for another post as I work out the kinks in my first layout, meanwhile I need to buy a multitester!

Thanks again, gentlemen, will get that photo as soon as possible And keep you posted. Pun intended.

Tomlinson Run Railroad

I have the same layout and have found that the turn out can not be placed directly connected to the 90 degree cross so I moved it back by one 031 curve to your left and the problem was solved. My sd 70 must have just enough of a pick up roller spacing to not make contact when going thru the curve and turn out together. For what its worth my dad had the very same problem and he fixed it by moving the turn out one track piece back from crossing. 

Humm ... interesting. That's the direction that my SD70 usual works but the steam engine doesn't.  Your solution should be easy enough to try when I get home if you mean that you or in this case your dad just swapped the switch with the curved track next to it.  That is, going left to right the sequence is switch, curved section, 90-degree cross. Doing so won't require buying another curved section but assumes that the switch and 031 curve match length-wise.

Thanks!

Tomlinson Run Railroad

The two pickup rollers  are intended to be wired together internally. One of them rolling across an insulator (especially a 90-degree crossing) shouldn't stop the locomotive.

---PCJ

Tomlinson,  specifically both rollers are supplying power to the engine full time and have no relation to what direction the engine is traveling.  Actually the main purpose of having two sets of rollers is to bridge gaps where there is no power so the engine will keep going.  It is possible on some switches (crossovers) for both rollers to be on a dead spot however.  This problem has lessened by some more expensive engines having two double rollers, i.e. there are four rollers on the track.

Dennis

Thanks to you both. Obviously, I assumed that the rollers were connected individually to the forward and reverse motors.  My engines are still too shiny and new for me to take a screw driver to them and explore their inner workings.  But I know that day will come :-}

TRRR

Hello TRR, I have attached a copy of where I placed my turn out on my lay out, with one 031 curve between the crossing and the switch.

TRRR,

It seems the questions have been answered about the rollers (collectors) for the center rail. I like the track change idea.

No, not another transformer, just another connection with a lock-on on the inside loop and opposite of the other. That will provide better track power distribution and connect the other outside rail if you place it opposite.

Lets leave the Lionel engine issue and the switch operation issue aside until you get the SD70 running correctly.

I am not a fan of MTH Realtrax even though my 7 x 9 layout uses it almost exclusively. I have a MTH City of Denver tin plate train that will not go through an O-31 switch without shorting out the track due to single powered truck of the engine and unique design of the center rail pickups. I can not run that train on my layout. All the switches have guide rails that are so sloppy that the truck wheels can move such that they bump into the plastic frog. It is not enough to derail but makes a bump in the trains motion. A great percentage of the track section rails misalign because there are no pins in the rails to keep the rails aligned. Careful filing on the rails to eliminate both vertical and horizontal misalignments were necessary. On some of the points of some of the switches I had to open the switches up and adjust both the mechanism and the points themselves. The anti-derailment feature on all seven of my switches was disabled by the extra length of the anti-derailment switch section of track. Shortening those track sections so they were not touching the adjoining track section re-instated the anti-derailment feature.

I have a couple of three foot sections of Atlas Flex track used to smooth out a couple of track sections that had trains waltzing when using standard sectional track. Those section are now a smooth directional change. That is not a Realtrax problem just the result of my choice of my own track layout. But that Atlas track runs smooth and should have been my choice from the beginning along with better switches.

The problems with the Realtrax crossover does not surprise me at all. Realtrax as a sectional track system in my opinion is a failure.

LDBennett

14 May 2015: Just added the pictures with captions.  The underside of the 90-cross section is probably the most interesting.  TRRR

 

Scottransam, thanks for the drawing. I'll attach a photo of the change I made last night. Very similar and I actually like how it changed the space and may allow for a freight yard build out. (I like the way you extended out around the basic layout.)  

 

As Carl/Moonman has noted there are likely multiple issues here.  So, separating the cross and switch track sections will help identify what's going on.  I'll add separate responses too in order to clarify each. This one's for the SD-70 and the cross.

 

#1: The 90-degree cross section. Attached is a photo of the underside. There's copper latticework connecting the three rails in both directions of the cross arms. Notice there is a single jumper wire going from what I'll call the south arm of the cross to the east arm (when viewed from the track side). It is clamped to the center rails. The SD70 stalls moving from the south to the north.  

 

Using my Bethlehem Steel ruler :-) on the top of the track, I see that the center plastic square is 1/32 inch lower than the plastic pieces that surround it.  So, there's a little low point for the roller. The overall impression looking at a side view is that the track where it joins the center square bulges upward on the four sides and then the center piece drops down that 1/32 inch.

 

When the engine shutdown again last night at a slow speed, the front roller was just about to leave the plastic square in the middle of the cross. The rear truck was about 2/3 through the curved track section attached to the bottom of the cross. The truck had swung out so far left that only the right-most metal hanger that holds the roller to the underside was touching the center rail.  So, it looks like there is a case for a moment at least when both rollers have no power.  Mannually pushing the engine forward a short distance powered it up again as it now had contact. I've recreated the roller positions on the track. (Sorry my camera is really cheap but my backyard trees are nice!)

 

Solutions? Jumper wire or metal plate?  I'm holding off on adding a lock-on to the inner loop (thanks for the further explanation) because that requires ordering yet another part. I want to get a multitester this weekend so I can better observe what's going on volt-wise and where in the track first.  

 

Tomlinson Run Railroad

** RATS! My photo uploads are failing and I have to head out now. The text should explain things well enough. Will try this evening and then post my discoveries about the switch and both engines now that it's been separated from the cross. **

Problem #2 I added a picture in the post above of the new track configuration that is similar to what scottransam/his dad had success with. It separates the 0-31 right-hand switch from the dead plastic spot on the 90-cross. Here's an update on how the MTH SD70ACe and Lionel 0-8-0 engines and my MTH bump-and-go trolley ran two nights ago.

Traveling clockwise (my intended direction for this layout), the engines both derailed at the switch every single time. They usually didn't before or not that often -- even with the cross and the 'S' curve next to each other.  The shorter length trolley ran several times successfully before eventually derailing at the same switch.  I was suspicious of the S-curve because this revised layout still has one and I read on this forum that they can be problematic.  (The other switch is OK and it doesn't have an S-curve.) But, get this, the engines run just fine counter-clockwise.  That is, entering the base of the switch before it splits off, works.  Entering from the curved side derails.  Previously, I had tested the track joints in both directions with a large, long box car -- there's only a slight "jump" at this track joint at the plastic frog (?).  So this behavior was unexpected.

Looking at the track connection on the curved (bifurcated) side, I see the usual sloppy MTH track engineering that LDBennet describes.  Some of the rail just doesn't quite line up and there are gaps between the metal. But more importantly, the black metal frog in the middle of the switch is higher than the rails.  An engine will likely jump up here and you can see scrapes across the paint already.

LDBennet, I'll have to reread your post a couple of times to make sure I understand what you did to solve this -- some of it is beyond my current understanding.  In the meantime, I will get out the needlenose pliers and see what I can do to improve the curve-side connection.  I'm not ready to file down the metal frog yet but have certainly had to take steel wool to other track sections to prevent derailments.

Having Lionel and MTH ovals from RTR sets to compare, I concluded that Lionel track connections were better.  However, I went with the MTH due to size constraints and because I preferred how the smaller MTH radius curves created empty spaces for planned scenery. But graduating to this layout added the complications of crosses and switches to what was already problematic.  Oh well, at least counter-clock wise runs :-). 

[By the way, it's really awkward writing about derailments of toys when there's been such a serious real derailment in the news this week. I'm sure it and the victims are on the minds of many of us.]

Tomlinson Run (or not! :-) Railroad

TomlinsonRunRR:

Unfortunately the solution is to not use MTH Railtrax sectional track. I don't have the problems you face with derailment but have had some engines slide sideways in the switches and short out the track, tripping my circuit breaker. I just don't run this engines anymore. I kind of miss the arcs and sparks from the short through the switch for that engine.  :-)  :-)

I did grind down the black center track glob in the frog of the switches to minimize the bumping. I never did anything with the guide rails through the switches that allow the trains to slide over enough to cause a vertical bump in their travel. I just recently ground all the plastic track rails and guides down to the level of the rails on all switches. The trains seem to run better (??).

I have way too much invested in the Realtrax to discard it. I just file the poor track rail alignments (both vertical and horizontal) and live with it. The result is acceptable for all but the one tin plate City of Denver which I only use around the Xmas tree. I chose Realtrax because it would allow my desired track layout in my limited space. I'm sorry now.

The MTH experience has been 80% good and 20% bad for all of their products which I have almost exclusively on my layout. The Realtrax is the worst but I have had other problems I easily fixed. I find it annoying that the design of the track and the trains in several case have conflicted. MTH stuff should always run on MTH track but that is not always true, it seems.

LDBennett

TRRR,

The 90° cross looks ok electrically. All of the rails around the center block are connected. The jumper is there to provide the option to isolate the pieces of the cross if there are two different power districts crossing one another. The jumper should be there for the configuration that you are using.

One caution, move the jumper so that it doesn't lay across the common contacts to avoid chafing the wire and having a short.

A 3 axle truck on a small radius track looks like a tight fit. The collector frame or roller may be slightly touching the outside rail at the point where the engine stops on the 031 curve.

If it is only the case of the roller dropping off of the center and not touching the outside while the front is on the insulating cross block, then the curve is still too small for this engine.

What is the part number for SD70 engine?

TRRR,

I tried to reconfigure the track arrangement to see if you can get better running. It is a close fit on the left with the roadbed and a slight force fit of the track joint. You need another 4.25" straight.

Moonman/Carl,

#1 Thanks for looking at the underside of the cross, the suggestion for protecting the jumper (it's on a carpet in the living room), and a possible reworked track to fix the curve problem.  The MTH RailKing SD70ACe is a PS3, item number 30-4231-1E that came with a RTR set.  I think the problem is that the rear roller is dropping off the center rail at the exact moment the front one is on the plastic as shown in the side view photo reenactment. It only happens at slow speeds. Higher speeds provide enough momentum to carry the train over the cross (although it does jerk upward slightly).  So, RoyBoy's suggestion of a diamond shaped metal piece on top of the plastic seems like a good solution assuming that it won't short out anything in the underwiring and I can afix it so it doesn't move when the rollers go over it.  I thought of glue, a copper pin, staples, most of which seemed like bad ideas.  Even extending each arm to have a narrow piece and bending each piece down 90-degrees to force it into the slot in the plastic where the center track enters the plastic.  Any thoughts?

#2,Regarding the curve with the RH switch sparking and derailing with both engines, the new layout adjustment (see photo above) actually was much worse!  The SD70 fuel tank also kept hitting the switch apparatus on the inside curve of the switch. Even after I moved the light to the outside, the low and long tank still bumped the shorter plastiC side.   So, I went back to the original layout that you posted from the MTH catalog. After some filing and testing 'till 1:00 AM, the engines all run clockwise without sparking or derailing.  Yeah!  I found things improved with the engines when I added cars but only up to so many (3).  Originally, I was trying to keep testing simple by only running the engines until I got them to run without problems.  On this layout, however, that logical troubleshooting approach was actually part of the problem.  the engines and curves apparently needed some rear weight.

#3 problem with this layout still needs to be dealt with, which is the left-hand switch "buzzes" and gets stuck if I try and switch it manually or with the remote.  It works just fine manually and with the remote when I remove it from the layout and apply current directly.  (Fortunately it doesn't have a high frog problem.)  I had a similar buzzing problem once with the right-hand switch when I used a 6-12 V transformer with my trolly and the voltage was too low. The switch needs 10V to operate.  So, as one of you suggested, it is possible the inner loop isn't getting the 10V the second switch requires.  I will get a volt meter this weekend and see if I need to order another lock on.

I appreciate yours and others' patience and help getting my first non-RTR oval working.  It was great to finally get back to running trains and thinking about what I can add to suggest scenery, stops, and so on.  Just two more tweaks, and I think I'll be there for good. (Or at least until my cleaners come and need to vacuum the carpet :-)

Tomlineon Run Railroad

TomlinsonRunRR:

I had a similar problem (buzzing switch motor). The anti-derailment feature of MTH switches uses the inside rail on both the straight and curved sections down from the frog as an electrical switch to detect a train when the switch is set the wrong way. The farthest most end of both those short rails that form a "V" in the switch need to NOT touch the adjacent track section rails. That is they should be electrically isolated. MTH makes those track too long and they can touch the adjacent rail and the switch thinks it is a train and tries to change the switch. I had to shorten the "V" section rails in each of my seven switches to get them to work correctly. When the wheel of the engine rolls across the gap in the rail, it completes the circuit to the switch motor and the switch changes to avoid a derailment. They work perfectly if those "V" rails are isolated.

Switch motors are only suppose to be energized momentarily. When they are continuously energized, as would be the case with the too long rails, the switch motors over heat and burn up. They buzz if there is a continuous voltage on them. They have been know to catch fire.

LDBennett

This problem has lessened by some more expensive engines having two double rollers, i.e. there are four rollers on the track.

Some of the less expensive RailKing uinit have this also.

I have a pair of early 90's Weaver locos that lose power by being on a switch gap and a UCS gap at the same time, but am not aware of any double rollers being available.  Woerkaround is not to go too slow, and PS2 keeps on going.

Recap: the inside switch at the bottom of this layout makes a buzzing sound if you try to switch it either manually or with the remote when the track has power. Its position cannot be changed.  It only buzzes it you try to switch it.  And, previously, if you added power to the switch by itself outside of a layout, everything worked fine. (I didn't retest that this time.)

I bought a multimeter today and did some simple testing. Perhaps this new data will explain to one of you what's going on with the left-handed switch on the inside loop. I also took the back off the switch, the switch block, and the remote switch. Didn't see any touching wires but one little metal piece that is visible when assembled had a little corrosion that I tried to remove. 

1. Based on LDBennett's comments, I noticed that the track pieces that move were the same length on the switch that works as the one that doesn't.  The Vs seem to touch a railing or not when they are supposed to, depending on how I set the switch manually.

2. BAD: If I manually set the switch to be straight when the power is off, turning the track power ON immediately changes the switch position to curved!  I never noticed this before. Hopefully it's unrelated to my removing the back and replacing it.  But it's clearly not good.

3. Continuity seemed fine and as expected by this novice. I tested from the track below the switch to the switch itself, and also from the track below the switch to the track above the switch.  

4a. I disconnected this part of the inner loop from the right arm of the 90-degree cross. Viola! The manual and the remote switches worked perfectly!  

4b. So, I replaced the cross with a 10" section to connect it to the curved side of the switch. The switch lamp wouldn't light up. It appeared dead. BUT, there was continuity from the track section below the switch to the 10" section above the left arm of the switch. The voltage was 19.1 however. The transformer is supposed to be 18 volts. Next, I tested the continuity of the straight arm-side of the switch. Touching what should be negative or the unused rail produced a high-pitched tone (not the usual continuity buzz) or the silence that I expected since the switch in between the two leads appears dead. This high pitched tone is what I get when I touch the red and black test leads to both positive rails. But I was touching the negative/unused rails on the track beyond the switch. Weird. Note that this arm of the switch goes to a siding only.  So, although the switch gave no readings, track on all three of its arms had power, and the track section above the supposedly switched off straight side was acting as though the grounds had positive power.

5. I ordered a second lock on, but the track near the switch had about 18 volts, which should have been enough to throw the switch had it been working properly.  (In fact, it DID throw the switch unexpectedly when I first turned the power on, as already noted.)

So, is there a short or something?  Or, is it time to contact MTH support?

Thanks,

Tomlinson Run RR

TomlinsonRunRR:

I did a poor job of explaining the anti-derailing problem.Since I am not able to post a picture, for whatever reason, I'll try explaining it a bit differently this time.

Imagine the switch is a "Y". The track rails that need to be isolated from the adjacent track section correspond to the inside of top of the "Y". I am not talking about the points but the two fixed rails  on the inside. They must not touch the rails of the track section next to the switch. There needs to be a gap in those two rails to electrically isolate them.

When you attempt to change the switch position manually or electrically then the switch will buzz and not be able to be changed. It means those two short track sections are touching the rails of the adjacent track section. The switch thinks it is that a train is entering a switch set the wrong way. Isolate those two rails and I think you'll be ok. All it takes is a gap of 0.010 inches.

LDBennett

Ah, I get it now, thanks, and I reread your earlier post and it makes sense, too. It's rather late, so what I did to quickly test your solution was to prop the track sections above the Y up off the floor a little bit. Doing so created a small gap between the sections. I turned on the track power, flipped the remote switch, and sure enough, it worked like a charm!  Tomorrow I'll get the files out and fix it permanently.

Probably a red herring, but with no power to the track, if I use my finger to pretend I'm a train approaching the switch from the wrong direction and nudge it, the track that normally worked, switches easily.  The other one doesn't budge.  But with your solution, that detail doesn't really matter, it's just odd.  I agree with you about the less than ideal quality of this track.

Thanks for the great troubleshooting help, LDBennett. I really appreciate it! Have a nice evening.

Tomlinson Run Railroad

First never use steel wool around toy trains it gets tangled in their guts. I'll say that before something breaks, I learned that from some who killed an engine with it.

I have a similar issue with the Lionel fas-track version of the 90 degree crossing, and derailment issues with manual O-36 switches. Sounds like a misaligned joint and possibly the roller getting temporary stuck in the engine, moving it would jar it free, oil it? Also I have found carpet can cause alignment issues, not bad ones, but the deeper the carpet the greater the problem. (Old carpet tends to wear into a mix making it easier to see)

On one set of MTH roller center rail pickups I found the stamped metal frame that holds the rollers was not free to move. I had to "adjust" it with pliers until it would move up and down freely. Just another MTH quality issue. I think Lionel has the same problem since they both are Chinese products. But we could not afford them if they were made in America thanks to the high cost of labor here. I guess we can't win.

LDBennett

Hi Allin,

Thanks for the heads up about the steel wool.  Fortunately, I had located my set of hobby files since that post and have been using them.  I just wiped the surface I filed on the switch with a damp cloth to pick up any tiny filings.  At this point, my various tests indicate that the location of the engine rollers on the long trucks, plus the plastic spot on the cross, and the track curve are the cause of stalls at slow speeds.  Interesting you found a similar problem with Lionel track. Yes, carpet can add challenges. Ironically with this regard, I find the MTH track is less problematic.

Tomlinson Run RR

I just filed the rail as LDBennett suggested to isolate the switch.  It was a little tricky to get at the base of the rail because of the plastic tab. So, I added a tiny piece of electrical tape at the base of the rail. Maybe there's a better way (Dremel tool?) but all I have are mini-files.  The triangular ones work well for filing track because they are stiff.

Here are pictures for future reference.  Thanks again everyone.

Tomlinson Run Railroad

I used a Dremel with a narrow disk to reach all the way to the plastic ties. I had to mod all seven of my switches this way. Another MTH quality control issue for me.

I would add that those short sections of track are controlled by the position of the points (that control the direction the train will go) through linkages internal to the switch. That linkage changes micro-switches to change the electrical state of those two anti-derailment rail segments. They become just an electrical extension of the adjacent rail in the switch direction and a electrical switch to control the solenoid such that the train forces the switch to a position of no derailment. It is all done by the micro-switches and the points.

LDBennett

Ah, another tool for the toolkit someday.  Thanks for the background info/detail on the switches. I did poke around in the innards a little yesterday.  It was a real joy to put my early steam PRR consist through its paces just now and have all the switches work.  it looks great.  Now back to some serious school work!  On my next break, I will setup the SD70 for some modern era fun. Now that everything is working, it turns out that this is a nice little layout with several possibilities.

Many thanks for help getting it there,

TRRR

In case anyone ever stumbles on this old topic, here's an update.  After running my trains over the MTH 90 degree cross for about four months, the raised plastic that originally seemed about 1/32 of an inch too high is now completely and uniformly worn down.  Now I can run my MTH engine over it at the slowest possible speed and it no longer stops.

Yeah! Problem solved.

Tomlinson Run Railroad