A few more Unitrack turnout wiring questions

[nik_n_dad]

Hello.

 

In my previous post about connecting 2 loops and using DCC, KenS made a great point about using insulating joiners and feeders to ensure that things don't get shorted.  That posting got me to thinking about other cases with power routing turnouts.

 

In the diagram below I've drawn a couple of track configs that we will have on the layout.  We're using n-scale kato #6 power routing turnouts and DCC.  Here's my assumptions for insulated joiners and for places I need to attach power (DCC).

 

Example 1: "Yard"

 

Insulated joiners at A, B, E, F

Power at C, D

 

Example 2: "Siding or Connecting 2 Loops"

 

Insulated joiners at H, M, K

Power at G, I, J, L

 

Example 3: "Passing Siding"

 

Insulated joiners at O, P, S, T

Power at N, Q, R, U

 

Does this sound right?  Am I adding in too many insulated joiners or power connections? 

 

Are there issues with power in the 2nd example where the diverging routes connect ("M")?

 

Thanks In Advance!

 

[CaptOblivious]

It all depends on what you will be doing with the structures.

 

Rule of thumb: Same booster, no isolation. Different boosters, full isolation.

 

Let's take the single cross-over first, because it's easier to work with: You shouldn't need any isolation, if all four tracks are powered from the same booster. However, you should be sure that both loops get power on both ends (i.e. at points I G J and L), which is easy to ensure if this is embedded in a loop of track. I run power-routing cross-overs like this all the time. (The problem with the earlier design in the other thread was that it would use one L-hand and one R-hand turnout; since the electrical design of the turn outs is a little weird, there was some call for isolation under certain circumstances. This one uses only L-hand turn outs, so there is no issue.) Of course, it should go without saying that you must make sure that both main lines are powered by the same booster! You should never need isolation at J and K except in weird circumstances though (where the crossover is at a power district boundary, for example, where L and J are wired to different boosters and I and G are wired to different boosters.).

 

The yard: If you will only ever encounter movements on one leg of the yard at a time, no isolation is needed: The turnouts will route power to whichever leg is selected, and anything on an unselected leg simply won't get power, which is fine. If you intend for multiple simultaneous movements through the yard (e.g. a loco working the top two legs while through trains pass through the bottom leg) you might consider your isolation scheme.

 

The passing siding. Same thing: no isolation is necessary. The turnouts will route the DCC signal to whichever siding is selected by the turnouts. That said, you will want to implement your isolation scheme only if you want a train standing on one siding to have its lights on while being passed by the other siding. In which case your scheme is correct.

 

So, for these simple cases, your isolation schemes are largely correct: What you haven't quite understood yet is when isolation is called for. In general, if all tracks are being powered by the same booster, there is no need to isolate them. Only when the DCC signal is coming from distinct sources do you need to think about isolation.

 

note all bets are off if you plan on implementing detection or automated control. In that case, your schemes are not nearly complicated enough!

[KenS]

To an extent, the answer is "it depends".  What  you've sketched out will certainly work, although in some cases it may be overkill. As before, I'll comment in the context of Kato's #6 turnout, which you're using.  The #4 is slightly different, and as Capt O. pointed out, other brands may behave differently (routing both rails, or not routing) and require the exact design you've given, or require additional feeders to the frog itself.

 

In (1), all you really need to gap to be safe are the rails facing the frogs at A, B, E, F. Then you need to feed those same rails (the two inside rails) at C & D (as well as both rails somewhere left of A/B and the frog rail right of E/F).  And, even that is unnecessary if you're willing to trust the "power routing" function and the yard sidings are dead ends or simple sidings (as in #3). In that case, one feed left of A/B with no insulators will work (but see the comment at #3).

 

In (2) you've got it right, although again it's only the frog rails that need to gap.

 

In both (1) and (2), you may need to gap both rails at the left or right end if there is any chance that beyond what's shown there's a "reversing section" (i.e., if, for example, the two tracks on the right are both connected in a loop, so the upper rail of one becomes the lower rail of the other).

 

In (3), the question is whether or not you want or need the unused track to be live at all times (e.g., to keep car lights lit on a parked train) or if you want it to be unpowered when the switches are thrown against it (e.g., for storing trains not in use, or just because power isn't needed). If you're willing for it to go dark, then the power routing of the switches will be sufficient for the inner rails, and the outer rails are continuous (in the Kato #6 turnout you're using) and will always have power, you just need a feed at N or U. If you do take the power-routing approach, you need to ensure both switches are always thrown to the same track, or you'll have all rails live, and be depending on the frog insulator to keep the voltage apart, which could be bridged by a wheel, causing a short.

 

Note: most DCC guides would argue for more conservative wiring, as you illustrate. And that's needed with many switches, which don't reliably power route. I've used Kato #6s for several years, and have never had a problem with them correctly power routing.  Also, if you're going to glue the track down in a semi-permanent manner, you many want a more robust design (some people even solder the unijoiners). But if this is just snapped-together unitrack, with the ability to unsnap it if you need to fix something, then you can use power-routing (if you want), and I'm describing options with that in mind.

 

If you want the main to always be live, you should insulate its rail facing the frogs at O & S, and feed it at C.

 

If you want both to always be live, then insulating both frog rails at O, P, S, & T and feeding then at Q & R is necessary.  Note that because the outer rails are continuous, feeding both rails at N will provide power at U (or vice versa). This is also the method I'd use if I planed a permanent or semi-permanent layout.

 

Also, note that unijoiners don't always make good contact, and are only rated for 3 amps.  Since that's somewhere between 5 and 10 normal trains, the latter isn't likely to be a problem in most layouts. But it's a really good idea to feed the track every 6 feet (2m) or so, even if the track topology doesn't require it. Insulators aren't needed in that case, unless you want to have separate circuit breakers or boosters.

[CaptOblivious]

Have I misinformed about the cross-over? I've never had to isolate my crossovers before: My Tomix turnouts have built-in isolating gaps, as part of the power-routing function. Do Kato turnouts not have this feature?

 

I'm beginning to think that perhaps we could all (well, maybe except Ken, who appears on top of things) a worked-out wiring diagram of a turnout in situ to see the dangers of not isolating…

[KenS]

I'll note that Capt O is right, when describing Kato's crossover.  I was describing a crossover made from facing #6 switches, which I believe is what you're illustrating.

 

Additionally, his comments on isolation are mostly correct, although I'll argue (as before) that you don't want hot rails of different phase separated by only the frog insulator, which is why you should be isolating the rails leading from the frogs.

[KenS]

Have I misinformed about the cross-over? I've never had to isolate my crossovers before: My Tomix turnouts have built-in isolating gaps, as part of the power-routing function. Do Kato turnouts not have this feature?

 

No, the Kato crossover isolates the inside rails on all four switches. It just needs to be fed from both sides (and as you note a loop will do that for you). But that's not what's illustrated in #2. And I was answering without assuming a loop existed.

[KenS]

I may be being overly-conservative.  I just measured one of my #6 switches, and the gap at the frog is over 1mm (about 1.5 or a bit less). That's not likely to be bridged unless one of the trucks has derailed.  Still, I'd insulate to avoid that case anyway.

[CaptOblivious]

I didn't mean the integrated double-crossovers (I don't own one), I meant single crossovers constructed from discrete turnouts. The gap on my Tomix turnouts is 1 or 2 mm, FWIW.

[CaptOblivious]

Just to interpose: Nik_N_Dad: your questions look innocuous on first glance, but as it happens, they are really probing! I'm learning a lot from these discussions, where I had just assumed in the past that the answer was very simple. As soon as I finish this damned dissertation draft (deadline is today!), I'm going to look at my turnouts very very closely and start a discussion about them.

[nik_n_dad]

I posted this as we're hoping to start gluing track down tonight (but more likely tomorrow)...and yeah yeah, we need photos!

 

This will be the 3rd layout we're doing.  I'm really not the train guy, Nik is (we don't know where he got the train bug, but since it's his hobby, it's now my hobby, and I'm not complaining).  Point being, I'm still learning through this.

 

Our first layout suited its purpose, but with Nik being 5 at the time, and me not knowing anything, and the world of books pretty much ignores kato unitrack, I fought all sorts of weird shorts and missing power from spots (like around the kato double crossover).  When we switched over to DCC, it really was too much work to consider fixing it all (ok, and we wanted a bigger layout)

 

Our 2nd layout is really a temporary thing we did to support DCC while we were planning on building "the big" layout.  It's basically mom's dining room table (thanks mom\sweetie) with a board made from 2" foam with mulitple loops, sidings and mini-yard.  I mostly went down the path of overkill on insulated joiners, but also discovered that the kato feeder\joiners aren't the greatest.  I also think in a few places around the turnouts we have had shorts. Nothing serious.

 

Which brings us to today.  I'm learning through trial and error, but have found you guys to be top-notch.  I originally thought last night that my posting may have been naive, but I guess this stuff is indeed a bit more whacky.

 

If it helps, I can throw together a simplified track diagram of what we're doing for context.

 

Thanks again for the help.

 

Captain-  Again in your debt.  Should I throw a giftcard to a brewpub in the mail to you or some decoders  ;-)

[CaptOblivious]

How big a gap does the Kato insulating Unijoiner introduce? Or, e.g., Atlas insulators? Is 1–1.5mm not enough of a gap?

[David]

I'm not sure if I'm on the right track or following all this right, but wouldn't the pickup system in most Kato cars (where power is collected from all 4 wheels on each side and carried on a brass strip through the length of the car) make any discussion of the length of the gap made by the unijoiner moot unless that gap was over 10cm? It seems that if a Kato car drives over the gap, it's going to bridge the two sections of track one way or another, so if a car is driving over that section, that section better have the switches/power routing in the right configuration or there will be a short.

 

See my doodle where the live track (red) and the insulated joiner (black) do little to prevent the pickup system (yellow) from bridging the gap.

[KenS]

How big a gap does the Kato insulating Unijoiner introduce? Or, e.g., Atlas insulators? Is 1–1.5mm not enough of a gap?

 

A unijoiner separates two rails that have the same phase. When a car bridges them, and any car with multiwheel pickups will bridge them, the two separate power sections (circuit breakers, boosters) will be cross-connected, and that's not a problem (although it could impair the function of the circuit breaker if something shorted one side elsewhere at the same time, but nothing will avoid that problem).

 

The risk with the frog is when the two rails that meet at the frog don't have the same phase.  That won't happen with a stub or double-ended siding off the same line.  In those cases the only reason to feed separately is to avoid losing power when the switches aren't routing. But it can happen with a crossover when one one switch is thrown and the other isn't, or it can happen with any diverging switch when there is a loop or other reversing section (e.g., a wye) involved.

[CaptOblivious]

Ken: but what when adjacent sections of track are out of phase? E.g. reversing sections? Isn't that just like the frog case?

[KenS]

The difference is that you wouldn't build track where only a unijoiner separates a phase difference, unless one of the two sides is a "reversing section" (i.e., it's connected to a booster or circuit breaker that will "instantly" reverse phase when a short across the unijoiner is detected).

 

You are correct that this is no different in principle from an accidental short that trips a circuit breaker (standalone or the one in the command station or booster) to disconnect all power, such as running the wrong way into a switch set against you.  In theory, a circuit breaker will always trip fast enough to prevent problems.

 

However, that's dependent on several factors, one of which is the distance (voltage drop) from the circuit breaker to the short, another is the type of load.  From my reading (which, admittedly, is mostly warnings from people trying to sell me their solution) circuit breakers can fail to trip due to excess voltage drop (long or inadequate wiring), or heavy load on the switched track (i.e., multiple sound decoders or lots of bulb-lighting on trains) causing the breaker to be close to the trip point to begin with.  So, while you have to depend on a circuit breaker in a reversing section, you will likely have only one train on it, and short wiring to it, while in the general case there are a lot more variables (i.e., more sources of risk).

 

Frankly, there's probably a bit of snake-oil in many of those dire warnings, and I won't claim I've understood all the arguments I've heard made.  On the other hand, I have heard (second hand) of actual failures, so there is some degree of risk there.

 

You can, and should, design to mitigate those risks, with shorter, heavier gauge wiring (and checking voltage levels with something like the RRamp meter). Not to mention keeping breaker-protected sections short to limit the load within them.  But why run a risk you don't need to?  If you can avoid having two opposite-phase elements closely adjacent (side-to-side, not end-to-end as in the unijoiner) then you can reduce the risk there.

 

There's no "right" answer (well, there are some things you just can't do, like reversing sections without track gaps), it's really all about risk management, and how much you're willing to trade off the design and construction time (extra track gaps and wiring) to mitigate risks, against your evaluation of how serious those risks are.  I'll admit I probably err on the side of more-work/less-risk. 

 

But I've paid hundreds of dollars for some of these trains, and some of them are near-impossible to replace.  I've seen photos of what happens to a locomotive when a circuit breaker doesn't trip, and they tend to look like they were put on a stove (much melting).  Risk aversion makes sense to me.

[KenS]

Getting back to the original topic of this thread, I decided to check out my various track switches, and see how they worked in terms of power routing.

 

diagram 1: Kato #6 switch.  This switch toggles the entire frog and point rail polarity when it's thrown, and leaves the inside rail of the non-lined section dead.  It's a very robust design, with minimum risk of an accidental short. Unless you run into the switch from the non-lined route, in which case a short is guaranteed when the wheel hits the frog. Or if your trucks are out-of-gauge, and the back of a wheel hits the point rail.

 

diagram 2: Tomix C140-30.  This is one of Tomix's "spring-switch" turnouts (very nice for tram layouts).  Because of this, the point rails and frog are separated, and a split-frog design is used. This ensures a train running "wrong-way" into the switch can pass through without causing a short.  The downside is that opposite-phase elements are closely adjacent.

 

diagrams 3 & 4: Kato's #4 has two modes of use, controlled by screws on the underside that control how power is routed to the frog and the two inside rails. Just to confuse everyone, Kato marked the mode most people call "power routing" as "non power routing".  The design uses a separated frog, and point rails that always have the same phase has the adjacent stock rail.  Because it lacks a split frog you still can't run into it in reverse (unless you're willing to de-power the frog, which is an option). But you don't have to worry about out-of-gauge trucks.

 

diagrams 5 & 6: Kato's crossover is wired much like their #6 switch, and shares it strengths (and weaknesses). Because the rails (R1, R3, B2 and B3 when straight, R1, R2, B1 and B2 when crossed) are gapped in the center, feeds are needed on either side.

 

Both the Kato #4 (set for "non power routing") and #6, as well as the Tomix C140, can be used to create a siding that is isolated from the main.  And all it needs for constant power is a feed (ditto for the main) if you aren't worried about the close proximity of the two inside rails near the frog (I think we've beat that horse to within an inch of its life, so I won't say more on that). So, referring back to the original diagram, one feed at N or U is required if you're willing to let the unlined track go dark. If not, additional feeds at Q & R are also needed (for the two Kato switches, only the inner rails need feeds; for the Tomix both would).

 

Similarly, a yard ladder of stub tracks can be fed from a single feed at the end, left of TO1 in the original diagram (and additional feeds on each track would be needed if you want constant power). And the Kato #4 always makes both routes live if you set it for "power routing", so you could make a yard ladder with it without additional feeds, and all tracks would be live at all times.

 

Only if you have a diverging switch that loops back on some other switch in such a way that phase reverses do you absolutely need track gaps. And then you need a reversing circuit too.

 

The TO3/TO4 crossover, if made from #6 switches (or the #4 set to "non power routing" or the Tomix), is not quite the same as half the Kato Crossover. When switched, the through route goes dark, but when lined straight the through route has power. So it only needs to be fed from one side (if you're willing for the through route to go dark when the switch is thrown). And a similar crossover made with the #4 (set to "power routing" to make both routes live at all times) avoids even that problem. With any switch, there's no requirement for gaps between the facing points like the ones built into the Kato crossover (assuming you get the two loops wired the same).  On the other hand, they do no harm.

 

If two of the tracks from the TO3/TO4 crossover were connected to a loop (e.g., the two ends on the right), you'd have a reversing section, and you'd need both a reversing circuit and track gaps between all four rails and the loop.  The Kato crossover would need gaps on the two outer rails, but provides the gaps for the two inner rails.

 

You can add more gaps and feeds to isolate things to avoid potential problems or to simplify locating problem sources when things don't work (and I would), but that's a matter of taste or design style, not a requirement.