DC-wired track logic issues

[Santa Fe Jack]

Greetings, all.

As I build my layout, I've been intending all along to go with DC, and perhaps change over to DCC at some point in the future. I have carefully selected the turnouts to be DCC-friendly, for example.

But I have this wiring logic problem that I'm wrestling with: I have three (possibly four) current tracks, with different parts of the layout wiried to different transformers. There are two main lines (big circles, essentially), plus a yard lead off the inner main line, and the possible fourth circuit for the roundhouse and engine runaround. You can see the layout topology in the first graphic at http://ho.tauxe.net

The two main lines, that will be controlled by different transformers, have a double crossover. If the crossover is switched to "straight through", then there is no problem, and the two main lines can be run independently. But when the crossover is switched to cross over, then somehow the two main lines need to be coordinated in both direction and power. I was considering two options, neither of which is particularly satisfying:

1) Wire the power to one of the main lines so that when the crossover is crossed, the main line power also changes to be powered by the same transformer as the other main line. Advantage: The two main lines are powered identically (from the same transformer) so crossing over is guaranteed to be smooth. Disadvantage: At the moment of switching, if the switched main line is not already coordinated (manually) to match the other main line, then there will be a nasty instantaneous change in power (and potentially in direction!) that could be bad.

2) Keep the two main lines powered independently, and trust the operator to make sure that they are in sync when crossing over. Advantage: No potential power jump when flipping the switch. Disadvantage: If the two lines are not coordinated when the crossover is attempted, the engine crossing over will encounter both circuits, and if they are not matched, bad things can happen.

To me, the disadvantages are severe in both cases. They both require "manual" coordination of the power packs by the operator. An inexperienced (or inattentive) operator could result in bad things happening, like either a sudden reversal or at least change in power in case 1, and a short in case 2.

So far, I have not figured out a way to "idiot proof" the system. I could make the little warning sign, but that's not really idiot proof. Well, I do have one idea: Make a little bar with two holes in it that slips over the two control knobs of the dual power pack, forcing the two knobs to be in the same position. This does not force the direction to be the same, however...

I expect that most of you will advise me to go DCC, as that would obviate all these problems. But I have at least one old engine from 1959 (a Gilbert NYC 4-6-4) that I do not expect I can convert to DCC, though I really do not know much about that.

So, what I need is some ideas on clever circuitry or perhaps mechanical controls (like the bar?) to force the syncronization of the power packs before allowing the crossover to be switched to cross over mode.

Any ideas? HELP!

 

[60103]

Jack: what Most of us would suggest is that you not permanently link any throttle to a particular piece of track. If you divide the layout into blocks, you can then hook a switch (say a rotary switch) to each block and then select a throttle to control it. (called cab control) I would say you need at least 2, preferably 3 blocks in each loop.
the other trick requires that the crossovers have switch machines with extra contacts. One set of switches, say the ones in the outer loop, are powered through the contacts in the switch machine and these pick up power from either the inner block or the outer block as required. are you going with the double crossover or two singles? There should be a block division in the middle of the crossovers and between the crossovers.
Clear as mud? Let me know.

 

[Santa Fe Jack]

Well, now, David, that is most interesting. I had not even thought of it that way.

I see what you mean about having many blocks and each of them being able to powered by any particular controller. That's a cool idea.

So, here's what I've got:

The two loops have a double crossover. It's a Walthers Shinohara code 83 (as is all my track) 948-8812, which is supposed to be DCC friendly and may therefore have wiring implications. I suppose I need to experiment with how this behaves electrically.

The switch machines are the Tortoise green ones, and they have what amounts to two SPDT switches (or one DPDT, in effect) on the board. These could be used to route power.


I have two dual power packs, so four indpendent controllers.

After what you have said, I am envisioning giving each of these a color. I will have a minature version of the layout (1:900 scale, actually, that will appear as a park train), equipped with switches, for operating all the turnouts. It would be super nifty to have each block of track in the mini layout light up in the color of the controller that is currently operating it. But, even if that is a bit too much to pull off, that is the "map" that one has to have in his head to operate the layout, I suppose. But wait - that's what the rotary switches are for! D'oh!

So - I need a four-way selector switch for each block, and these could be color coded to the controllers, and would live in the mini layout. That's going to be crowded. But I get ya. (Right?)

Now, the problem of avoiding a nasty mixup is to remember only that you cannot switch the power to a particular block unless
1 there is no engine on the block
or
2 if there is an engine there, the power must be zero for both the controller being switched from and the one switched to

Oh - and a fifth option would also be userful: no power at all. This is good for the yard and the roundhouse and any other places where engines are stored.

Thanks, David. I now have a whole new perspective on this.

And, I could use some help on understanding the double crossover better.

 

[NYNH&H]

If you want to learn about wiring DC blocks, which was the "standard" way of wiring a model railroad form the beginning of DC trains to the late '90's when DCC took over, check out Andy Sperandeo's Easy Model Railroad Wiring. It starts assuming no knowledge of anything, and builds up to crazy complicated, step by step. Or do the best thing and get a Digitrax Zephyr from Tony's, and a few decoders to go DCC. That is the easier.

 

[60103]

If you pick the rotaries, you can easily get a 6 position with 2 poles and use one side for the indicator lights. It will also offer 2 "off" positions. I'm using these for part of my wiring.
Worst problem with rotaries is when you crank it past another cab that's running a train and your train jerks.
Get "non-shorting" rotaries. These leave the one position before they contact the next one. Shorting rotaries will connect 2 cabs together as they turn.
You may not need the extra complications I described for the crossover, if you can discipline yourself (and your helpers) to set up the receiving block to match the sending.

 

[Santa Fe Jack]

Good point on the "shorting rotaries" problem. What would be ideal, I think, would be pushbutton selectors, if I can find such a thing.

So, now I am pausing in layout development to consider the pros/cons of converting to DCC. Can I even convert an old Gilbert steam engine from 1959 to DCC?

 

[MasonJar]

It may be possible to convert, but you will have to judge whether it is worth it based on the amount of work it will take.

The key to converting any loco to DCC is to ensure that the motor receives power only from the decoder. The means (for example) if the frame is used to conduct power to one side of the motor (common in older locos), you will have to 1) insulate the motor from the frame, and 2) probably change the way power is routed through the loco. This could involve replacement wheel sets (insulated), adding pick-ups, rewiring the tender (if steam), and so on. Not overly complicated, but requiring some careful planning and execution.

If it is a nice loco and a good runner, you may judge the effort worthwhile. If it isn't, but still holds (sentimental) valure for you, maybe a display case is in order...

Andrew

 

[Santa Fe Jack]

It may be possible to convert, but you will have to judge whether it is worth it based on the amount of work it will take.

The key to converting any loco to DCC is to ensure that the motor receives power only from the decoder. The means (for example) if the frame is used to conduct power to one side of the motor (common in older locos), you will have to 1) insulate the motor from the frame, and 2) probably change the way power is routed through the loco. This could involve replacement wheel sets (insulated), adding pick-ups, rewiring the tender (if steam), and so on. Not overly complicated, but requiring some careful planning and execution.

If it is a nice loco and a good runner, you may judge the effort worthwhile. If it isn't, but still holds (sentimental) valure for you, maybe a display case is in order...

Andrew

I guess I need to open 'er up and take a look around. I don't want to get into new wheel sets and all. She's a great runner (very heavy and powerful, with lots of torque even at low speeds) and has great sentimental value as well. This was my first loco from when I was a little kid, and is part of a complete train set from Gilbert, circa 1959 or 1960. Unfortunately, it is missing the front 4-wheel truck and I have not been able to figure out how to fix that. The cars are good, too, though I need to change over all the couplers to Kadee at some point.

 

[MasonJar]

Try NorthWest ShortLines (www.nwsl.com). They offer all kinds of loco mechanisms and may be able to suggest something for you to remotor, regear, and replace missing parts.

While it is a good runner, I have retired my first trainset (a Bachmann CP F-unit and assorted cars) to the display shelf, as it does not fit my era, and I think it would take too much work to get everything up to operating scratch.


Andrew

 

[Russ Bellinis]

One thing to be careful of when designing a dc layout with multiple throttles is to make sure that you never have two throttles on the same bit of track. Fortunately the power transistors in most commercial throttles are pretty forgiving, but if two throttles are wired oposite (such as one set for forward and one for reverse) to each other on the same track, you risk blowing out transistors. Your double crossover needs to have insulating gaps cut into the tracks. It seems most double crossovers available commercially have all tracks on the crossover common with each other. The other thing you can do if you are not comfortable cutting gaps in the rails is to insulate all 4 rails in and out of the crossover and make the double crossover a separate block that can only be powered by one throttle at a time, either through toggle switches or a rotary switch.

 

[hiscopilot]

Thanks for posting these questions and answers, I really am learning so much.

 

[Santa Fe Jack]

One thing to be careful of when designing a dc layout with multiple throttles is to make sure that you never have two throttles on the same bit of track. Fortunately the power transistors in most commercial throttles are pretty forgiving, but if two throttles are wired oposite (such as one set for forward and one for reverse) to each other on the same track, you risk blowing out transistors. Your double crossover needs to have insulating gaps cut into the tracks. It seems most double crossovers available commercially have all tracks on the crossover common with each other. The other thing you can do if you are not comfortable cutting gaps in the rails is to insulate all 4 rails in and out of the crossover and make the double crossover a separate block that can only be powered by one throttle at a time, either through toggle switches or a rotary switch.

I think I won't have the problem of two controllers talking to one block, if I wire each block through a switch that selects only one of the the controllers (or nothing).

I think you're right about the double crossover. The outer long rails are continuous, but I've figured out that if I cut the long rail at the middle, then in effect the double crossover will have four separate blocks. This is cool because I could conceivably have one controller cross over one way and another crontroller the other way with no electrical issues.

 

[60103]

Jack: check your crossover to see if there are a string of plastic insulators right across the middle of the crossing. That will tell you that it's insulated into 4 sections, although I don't know about the continuous outer rails.

 

[Santa Fe Jack]

I checked it, and there are many insulated and cleverly cross-wired sections. The piece ways it's "DCC-friendly", FWIW.

The outer rails being continuous, I cut them with a hacksaw and inserted some plastic between the rail junctures. So now it works fine - each of the four leads is independent, so I can have the blocking go straight through, or cross.

I think this is going to work. I just need to find compact 5-way (or more) selector switches that do not "drag" across the intermittent settings. a pushbutton-type would be neatest. That may be tough to find.

60103 - where did you find the 6-position non-shorting roataries?

 

[60103]

Rotary Club

60103 - where did you find the 6-position non-shorting roataries?

Radio Shack ops:ops:ops:

We've got some quite good electonic supply stores on the west side of Toronto. My rotaries came from some years ago; we no longer have Radio Shack in this country.

 

[pgandw]

I just need to find compact 5-way (or more) selector switches that do not "drag" across the intermittent settings. a pushbutton-type would be neatest. That may be tough to find.

Tough to find in these days of soft logic, but the old car radio tuner buttons would be just about right. An auto junkyard would be your best bet. The radio doesn't need to work - you just need the buttons.

my thoughts, your choices

 

[NYNH&H]

Pushbuttons would need some sort of way to "hold", and keep the power going. This sort of sounds like engineer control, another topic altogether. Just stick with some rotaries if you are going to go DC. Or just to the easy thing and get DCC. Check out the Digitrax Zephyr, it is an awesome system!

 

[60103]

One of my friends made something like this from a lot of two-coil relays (sort of like a switch machine) and a diode matrix. A push on one button would engage that relay and disengage all the others (most of which were already disengaged...)

 

[Santa Fe Jack]

The radio buttons (from wihc the computer GUI analygy is derived) is exactly the type of device I need. But after browsing through my Mouser Electronics catalog last night, I can't find anything like them. And if Mouser doesn't have it, it's not availabe except as some sort of custom device, like 60103 is talking about. I'd also want it to be very small.

I believe I'm stuck with the rotary switch idea, and I found the 6-position 2-pole non-shorting rotary switch at Radio Shack. Just one problem that I can't see how to get around: If I'm switching from, say, position 1 to positon 3, I HAVE to go through positon 2, albeit momentarily. In this application, these positions all are selecting different power pack controllers. If there's already a train running on#2, then the block I'm switching will "see" that as it passes over the contacts. This is not good.


The only way around that that I can see is to have something like a little extra disconnect switch that one would turn off before engaging the rotary. This makes the operations more complicated, as one would have to disengage power to the block witht he disconnect, turn the rotary to the desired position, then re-engage power. Hmmm... it's analogous to a manual transmission on a car, with the clutch being the disconnect and the gear being the desired position on the rotary.

Has anyone thought of something simpler?

 

[MasonJar]

Has anyone thought of something simpler?

I hate to point it out again, but DCC is simpler... (duck and cover...! )

You said in your very first post that you want to go to DCC eventually. Maybe it is time to reassess and go with DCC now...? The wiring is simpler, takes less time, and does not involve as many moving parts. PLus you run the locomotive, not route power, when operating...

Andrew