Protecting the coils in old twin-coil switch machines - DIAGRAM ADDED (1st post)

[lionelsoni]

But, if you release the momentary switch, the entire circuit should be open at that point, regardless of where the turnout contacts are.  Do you perhaps use control rails, which, unlike the momentary switch, could try to throw the turnout indefinitely?

[nickaix]

Bob, thanks for the clarification. It helps a lot.

You are right: I do use momentary switches to throw the turnouts. The burnout problem comes when something prevents a turnout from throwing completely. The contacts are supposed to cut off power to the coil after the turnout has thrown fully. The turnout in the diagram has already thrown fully; had something prevented it from completing its motion, then contacts 2A and 0A would still be touching each other and the coil would stay energized. Or if the contacts are out of adjustment, the same thing could happen even if the turnout throws properly. Not having to worry about that would be another advantage of the diodes.

[lionelsoni]

There is one power supply powering multiple lamp-capacitor circuits.  Each capacitor, in turn, powers as many turnouts as you want it to.  As the number of turnouts that might be thrown by a single capacitor increases, so must the capacitor size.  On my layout, I use a single capacitor for each cluster of related turnouts that are likely to be thrown at the same time.  This also simplifies the wiring, because I use nearby accessories like lighted billboards and bumpers for the lamps, keeping the lamps, capacitors, and their wiring all in the same area of the layout.  I use control rails to throw turnouts automatically, so I don't want to share capacitors between these clusters, since throwing turnouts in one area might leave a capacitor discharged when it is needed to throw a turnout automatically elsewhere on the layout. 

The diodes do keep the coils isolated.  They form a simple "or" gate, which throws the next turnout in line regardless of which way the present turnout is thrown.  I locate them right at the terminals of the present turnout, so only one wire goes on to the next turnout.

Yes, this would replace your contact scheme.  However, much of the wiring would be the same.  For example, your wire from XW to the next turnout's RW or NW is equivalent to my wire from A to one of the next turnout's Bs.

The most straightforward cluster of turnouts on my layout is the throat of an 8-track yard.  The controllers connect only to the rank of 4 turnouts that immediately adjoin the yard tracks.  Each of those 4 has a 2-diode gate wired directly onto its terminals, with a wire going directly to one coil of the appropriate turnout in the next rank.  Those 2 turnouts also have 2-diode gates wired directly to their terminals, with wires going directly to the coils of the final turnout that adjoins the yard lead.  All 7 turnouts are powered from one capacitor (10 millifarads, if I remember correctly) charged through a number 57 lamp from the single DC supply that serves the entire layout.  Throwing any of the first rank turnouts thus aligns all three turnouts needed to get a path between the yard lead and the selected yard track.

I'm a little puzzled about exactly how your scheme works.  If you use non-momentary SPDT switches to throw turnouts, as your diagram suggests, what happens when two upstream turnouts or one upstream turnout and an SPDT switch both try to throw a downstream turnout in opposite directions?  It seems like you would need to use momentary switches, much like Lionel controllers, or pushbuttons to avoid buzzing motors.  But that in itself would prevent the burnout problem, unless you deliberately held the switch for a long enough time.

[nickaix]

Bob, thank you for your reply. Am I right in thinking that everything in this diagram to the right of the capacitor gets repeated for each turnout? It looks like the diode is there to prevent both sides of the coil from energizing at once? And that in this way of doing things all the turnouts will operate simultaneously?

If I AM understanding this rightly, then I guess this scheme would completely replace the system I am currently using with everything wired through the moving contacts on the switch machines. This thing with the diodes looks more elegant; but my turnouts are already wired in and (since I repaired that one machine) working reliably. I think I would prefer not to rewire everything if there is a simpler way to protect what I already have installed.

(And please pardon all the questions! My grasp of electrical matters is, at best, tenuous. I'm trying to learn, though.)

[lionelsoni]

   ---lamp-----------------
   |        |       |     |
   |        |     coil   coil
   |        |       |     |
   |        |       |     |-<diode--
  +|        |       |     |        |
  DC       +|       |     |        |
 power  capacitor   |-----+-<diode----A
supply     -|       |     |
  -|        |       |     |
   |        |    B----o o----B
   |        |          ^
   |        |          |
   |        |          |
   -----------------------C

A--to B of next turnout

B--from A of previous turnout or control rail

C--common (optional if control rails are not used)

I would start with a 16-volt DC supply, number-53 lamp, 5 millifarad capacitor, and 1N4001 diodes.  Use a bigger capacitor if the capacitor doesn't throw the turnout fully.  Use a number 57 lamp if the capacitor doesn't recharge quickly enough.

[nickaix]

Please see the 1st post for the requested diagram.

[nickaix]

Thanks for the replies. I had wondered about some kind of capacitive discharge system, too. The timing circuit also sounds like a neat idea.

I am away from the trains today, but I will post a diagram of my setup next chance I get.

Thanks again for the help!

[lionelsoni]

It is not a requirement that all the turnouts be supplied from one capacitor.  That is an advantage of using diode logic to select the turnouts to be thrown.  Without more information about how the present turnouts are sequenced, we cannot say just how capacitive discharge should be applied to the existing scheme, or whether diode logic combined with capacitive discharge is a simpler, more effective, and of course burnout-proof way to go.

[lonewoof]

If this system is powering more than one switch machine in sequence, a capacitor-discharge supply won't work. The first coli would discharge the capacitor, then there'd be no power for the next one(s).

I think your best bet would be a timer circuit (easily enough done with a 555 IC) that would apply power for a limited time (say 2 or 3 seconds) that would be long enough for all the machines to operate, but would not keep power on if one machine hung for some reason.

 

[lionelsoni]

I think it might.  The idea is that the coils are powered from an electrolytic capacitor rather than a conventional power supply.  The turnout motor draws current for only a fraction of a second, no longer than needed to move the points.  When the controller contact, control rail, or whatever opens the circuit, a DC supply recharges the capacitor in a couple of seconds through an incandescent lamp.

However, without knowing exactly how your turnouts are interconnected, I can't say whether capacitive discharge is compatible with the way you're doing things.  On my layout, I also throw multiple turnouts with a single switch or control rail; but I use diode logic rather than any contacts on the switch machines (mostly Lionel O27 turnouts).  There is no sequencing--they all throw simultaneously.  But there is no problem with a peak power supply load, since the energy comes from the capacitors rather than directly from the power supply.  For example, I have an 8-track classification yard whose 7 turnouts are powered from a single capacitor recharged through one lamp.  I use 8 contacts in 4 controllers to designate which track I want, and 3 turnouts throw at a time.  The logic uses 12 diodes, which are located at the turnouts' terminals.

[balidas]

You'd probably want s fuse that resets itself, but I'm wondering if the capacitive discharge circuit Bob Nelson recommends would work for you.

[nickaix]

On the family layout, we have a number of old Kemtron and PFM twin-coil switch machines. They are powered off the ZW, at about 16 volts. These have contacts which are made or broken when the switch machine operates. They're wired in tandem, so throwing one turnout will immediately throw any others in the yard that are not lined properly for the train to pass--but they are not thrown simultaneously. Each one throws a split-second after the one before it. Power for the coils is routed through the contacts, so that power is cut when the switch throws, and then shunted to the next turnout in line.

Sooo, I had one machine which would sometimes not throw all the way--which meant that the coil was always powered. Boy would that thing get hot in a hurry! I have since fixed it, but I am lucky I did not burn it out.

So I am thinking I need something like a fuse or a breaker that will not trip just by throwing a turnout, but will trip if one of the motors continues to draw current for more than about 15 seconds or so.  Is there any such animal?  Or are there better ways to protect these things?

As requested, here's a diagram of how these are wired. This is from Linn Westcott's book "How to Wire your Model Railroad." 

For operating multiple turnouts in tandem, the wire XW is connected to either RW or NW on the next machine, depending which way you want the next turnout to be thrown. I have 6 machines daisy-chained together this way, so that each operates the next.