Connecting Rokuhan turnouts to a decoder

fredz May 8, 2012

  1. fredz

    fredz New Member

    9
    0
    5
    I now use my Rokuhan turnouts in non-power routing mode. This works fine, but when 2 turnouts are not set the same way, and a train goes over one, the track gets short-circuited. Would connecting both turnouts to the same turnout switch (so to the same "button") solve this? Or does this only work on very simple layouts?
     
  2. Garth-H

    Garth-H TrainBoard Supporter

    986
    52
    23
    Unless you have purchased the newest version from Rokuhan turnouts R39 and R40 and converted them or modified the original version of L110 turnouts Rokuhan turnouts are not non power routing by design and as delivered, they are power routing. R39 and R 40 turnouts can be configured for non power routing by following instructions and inserting two screws in the proper locations in the base plate. The original turnouts R006 and R007 can be modified to be non power routing by modifying the turnout according to Rokuhan Video or instructions on David K smith's web site on Rokuhan z scale track.
    In power routing turnouts there has to be an insulator in the track between two frogs facing each other on a piece of track. So, if you have a passing siding off the main with a turnouts at each end and you only turn one of them to the passing siding you will have a short unless there is an insulator between the two frogs. If the two turnouts are both set for the passing siding there is no short and trains will run through. In power routing the track power to all turnouts has to come from the un-switched entry point to turnout. If you power the siding separately and run an engine towards turnout, if the main and the siding are not the same polarity, it will short at the frog.

    Rokuhan turnouts are single coil machines so a DCC decoder that is designed for twin coil will not work. So for me it has to say suitable for Kato unitrack single coil or Bi-Polar turnouts. Then you have to consider the voltage being used. Rokuhan turnouts use very low current 9vdc so 20vdc pulses are not ideal. So far I have heard of good results with ESU decoders.
     
  3. David K. Smith

    David K. Smith TrainBoard Supporter

    1,211
    0
    22
    Garth, I don't think this is correct. If a pair of power-routing switches are used for a passing siding, and each is thrown to a different track, it won't short, because the leg to which the switch isn't thrown is dead (open connection). This is how you can park a train on a stub-end siding with a power-routing switch. You don't need any insulated joiners anywhere except in special circumstances.

    Now, if you have non-power routing switches ("power everywhere") and you have the same setup with two switches and a siding, the polarity of all the tracks is the same, and everything will work. What won't work is the frogs and points, because they are the only parts in the switch that will change polarity, and that's where a train will short if it's going against the switch. Also, it's the power routing switches that must have track power fed from the points end of the switch; otherwise, you can wind up with shorts or dead tracks that shouldn't be dead.

    Here's the page showing the various scenarios: http://jamesriverbranch.net/clinic_4a.htm

    Incidentally, I will be revising this page to cover DCC issues in more depth, because contrary to what most claim, it's actually better to use power-routing switches as opposed to the power-everywhere type, for many reasons.
     
  4. tjdreams

    tjdreams TrainBoard Member

    376
    317
    18
    OK I'm new to DCC and not that good with wiring But
    If the DS64 decoders puts out 20 v and the Rokuhan switches work on 9v Would wiring 2 switches together work? (9v+9v=18)
    Or would i still run the risk of burning out the switches?
     
  5. David K. Smith

    David K. Smith TrainBoard Supporter

    1,211
    0
    22
    I should think a dropping resistor would do the trick. Wiring switches in series would be awkward, since you won't always have pairs of switches to operate in tandem.
     
  6. CSX Robert

    CSX Robert TrainBoard Member

    1,120
    193
    28
    This is not true for all power routing turnouts. Some power routing turnouts have the frog and the inside rails of each leg as one electrical unit and they switch the power to them to one rail or the other depending on which way the turnout is aligned. These types will cause a short as described above. Some power routing turnouts have the frog insulated from the inside rails of each leg and they cut the power to the inside rail of the leg the turnout is not aligned for. These types will not cause a short as described above; although they will cause a short if the frog is powered and a train approaches from the leg the turnout is aligned against.

    I believe the Rokuhan turnouts are of the second type. They appear to be very similar to Kato Unitrack turnouts ,and when the Kato's are power routing, they are of the second type.
     
  7. David K. Smith

    David K. Smith TrainBoard Supporter

    1,211
    0
    22
    This is true. For older Rokuhan switches and newer ones in power-routing mode, the short rails past the frog are isolated from it, and are only powered when the route is aligned with it; otherwise it is dead (open circuit). This provides a tremendous advantage for DCC layouts: locate a rail gap a little more than engine length from the rail leading from the frog. Beyond this gap, install a drop to the DCC power buses. Now you have a switch that you cannot run against, and will also not create a short. (The only way this could fail is if you have a double-header, and the second loco has enough traction to push the leading dead engine into the frog.)
     
  8. JimVane

    JimVane New Member

    3
    0
    4
    Garth,

    Has Rokuhan or Digitrax solved the 20 volt output problem when using an externally powered DS64 with Rokuhan turnouts?

    Fredz: "For the pulse output mode for use with solenoid and bi-polar turnouts, the voltage is always approximately 20 volts." (using a built-in capacitive discharge unit.)
    I contacted Rokuhan about this, and they say:"our turnout is worked over 9V DC. Maximum probably 15V or 17V DC."


    "Moreover, when using an external power supply with the DS64, which is recommended, the DS64 always takes power from the source with the highest voltage. Eg if you DCC central delivers 10 or 11V, and your auxiliary power source to the DS64 delivers 9V, the DS64 will take power from the DCC central. For the Rokuhan turnouts: the lower, the better:"
     
  9. shamoo737

    shamoo737 Staff Member TrainBoard Supporter

    4,530
    381
    67
    Welcome to Trainboard Jim.
     
  10. Cleantex

    Cleantex TrainBoard Member

    468
    2
    18
    The unique reverse voltage coil of the Rokuhan is very sensitive to over-voltage. I run them with only 7,5Volt and impulse limited to 0,3 seconds. The specific decoder deserves 4 turnouts and is available as a kit : (WDec2-MA-K, 27.90Euro)
    http://www.jokashop.de/
     
  11. JimVane

    JimVane New Member

    3
    0
    4
    Thanks Armand, I'll check it out...

    Jim
     
  12. markm

    markm TrainBoard Supporter

    795
    232
    20
    tj,
    The one thing I would not do is stack the switches in series. While each switch is rated at 9V, if you connect them in series, other electrical characteristics (impedance) come into play. If you connect the switches in series, you'll probably see something like 12V across one switch and 8V across the other (or worst).

    Mark
     
  13. ViperBugloss

    ViperBugloss TrainBoard Member

    84
    18
    9
    Using Zener diodes with Digitrax DS64 Stationary Decoder and Rokuhan switches

    I have been experimenting with connecting Rokuhan switches to the Digitrax DS64 Stationary Decoder.
    I found information on a similar stationary decoder from a different manufacture (http://www.ldt-infocenter.com/dokuwiki/_media/de/anschlussbeispiele/page_1368.pdf).
    This recommends putting two 9.1v Zener Diodes wired head to head in series between the decoder and the switch.
    I am no electronics expert, but from the Wikipedia information on Zener diodes, I am deducing that this circuitry limits the voltage in either direction to 9.1 volts.

    I have tried this and it works. Sending a throw or closed command from the command station activates the switch as required.
    This is just an experiment on the workbench, rather than a long term evaluation of whether this is a suitable solution to avoid damaging the Rokuhan switches over prolonged use.
    I have not been able to measure the exact voltages involved using a simple voltmeter as the voltage discharge is very short, however, I believe that this is a solution to the problem of using the DS64 with Rokuhan switches.

    Rob Pearce (aka ViperBugloss)
     
  14. markm

    markm TrainBoard Supporter

    795
    232
    20
    Rob,
    Basically one Zener diode will conduct at about .6V and the other will clip the voltage to 9.1V. You may want to add a small value series resistor, 50-100 ohm, to limit the current to protect the diodes. The resistor selection might be a bit tricky: you don't want to exceed the power rating of the diodes, but enough to throw the turnout. If you add a resistor, let us know what value works for you.

    Mark
     
  15. markm

    markm TrainBoard Supporter

    795
    232
    20
    Rob,
    I've had a chance to look at the ZPY9V1 data sheet. Looks like for long term reliability, you would want to limit the current through the Zener diodes to 100mA, it looks like you'd want a 27 ohm series resistor.

    Hope this helps,

    Mark
     
  16. ViperBugloss

    ViperBugloss TrainBoard Member

    84
    18
    9
    markm

    Firstly, of course I am talking about connecting Rokuhan turnouts to a Digitrax DS64 Stationary decoder (using the word switches can cause confusion although fortunately not in this case).

    As I said in my previous posting, I have little electronics experience, although I did study physics at school but that was 50 years ago.
    I remember that V = iR, and doing things with Wheatstone bridges.

    Following your suggestion about current limiting resistors, I have been reading up on zener diodes (mainly at "http://www.electronics-tutorials.ws/diode/diode_7.html"), doing a few calculations, and some experiments. I have tried a number of resistors from 10 ohms to 180 ohms, including 27 ohms. However, I have never been able to get the turnouts to operate. The zener diodes I am using are not the ones mentioned in the article I quoted but a similar one ("http://uk.farnell.com/jsp/search/productdetail.jsp?sku=1858630l"). These have a max power rating of 1W, and using the information in the tutorial I calculate the following:

    Max Current = Watts/Voltage = 1/9 A
    Min R = (20 - 9) / (1/9) = 99 ohms. (The DS64 output is 20 volts.)

    Using a 100 ohm resistor in series is one the tests I made but again it was not successful.

    A question that I do have is how the Rokuhan turnout should be wired. The original article ("http://www.ldt-infocenter.com/dokuwiki/_media/de/anschlussbeispiele/page_1368.pdf") shows the turnout wired in series but the electronics tutorial states that the load is connected in parallel to the zener diode. This tutorial is referring to a continuous load but for the Rokuhan turnout there is a brief impulse, so I do not know how this effects things, or how the turnout is wired internally.

    Without using a resistor, my experiments with Rokuhan turnouts in tests on my workbench (actually, my dining room table) are successful. I have wired in two turnouts and setting up a simple route, sees the first turnout change (with a satisfying click), brief pause, then the second. However, your point about long term reliability is a good one. One of the zener diodes appears to have stopped working and was replaced. At under 5p each, this is not a problem, compared with £33.50 for a Rokuhan turnout (UK price), but if there is a failure when wired into a proper layout, determining and rectifying the problem could be irksome.

    I am currently preparing a simple board for a DCC test layout incorporating a reverse loop similar to the layout suggested in "The Digitrax Big Book of DCC", so that I can continue learning about DCC, this time on an actual layout.

    Rob Pearce
     
  17. Garth-H

    Garth-H TrainBoard Supporter

    986
    52
    23
    Rob

    The Rokuhan like a Kato turnout uses a single coil and a rare earth magnet to through the points, so when the coil is energized it moves the points in one direction , to reverse the direction and move the points in the opposite direction you reverse the polarity on the two wires. so if the red wire is plus and black wire is minus and that moves the points to straight through route then with minus on red and plus on black the diverging route will be chosen. The manual controls use a lever or toggle switch that is spring return to center off, so you only get a momentary pulse of power to the coil.

    I = V/R so to find R it is R =V/I or 20v at .25 amps or 20/.25=80 ohms change current to .5 amps 20/.5 = 40 ohms change the current to 1 amp then R = 20 ohms
     
  18. markm

    markm TrainBoard Supporter

    795
    232
    20
    Rob,

    As they like to say on Dr. Who: "It's complicated." So question everything I write here, since I can get into techno babble too easily.

    Basically, when you apply power to a turnout, the actual signal seen at the turnout looks something like this:
    View attachment DC Pulse.jpg


    The coil initially has a zero resistance and will try to draw as much current as it can. As it starts drawing current, it also starts developing a magnetic field and "charging" the coil. Sometime during this process, the magnetic grows strong enough and some of the energy stored in the coil will be drained moving the turnout. After that, the coil become saturated and look like a resistor (marked as space header) and the current through the coil looks like Ohm's Law, R=V/I.

    From data Garth has previously posted and from some of my experiments I've done with the Rokuhan turnout switch, this resistance is high enough that a continuously powered coil does not overheat.

    The Zener diodes generally have two power limits: a continuous power limit and a pulse power limit. The continuous power limit is specified based on the power limit of the diode: P= V*I, I= P/V = 1watt/9.1V or about 100mA. This should not be an issue because of the high resistance of the Rokuhan coil.

    The pulse power limit, is that maximum power the Zener will tolerate without damage. This is usually specified as a pulse of about 10ms, which is typical of the time it takes for a coil to charge and self-limit it's current. The datasheet I found for the Vishay diode you mentioned doesn't specify a limit, so I assumed 100mA.
    Looking at 9.1V Zener diodes from other manufacturers, the value ranges from 500mA to 10A. If you exceed this value, it will damage the diode. This is the current we want to limit. So using Garth's values R=(20V - 9.1V)/.5A = 22 ohms.

    Follow me so far? Here's is where it get complicated.

    If the turnout coil doesn't get enough current fast enough, it doesn't develop a strong enough magnetic field to throw the turnout. Reviewing the ds64 document, it's unclear how much power the output can supply. If you are using the recommended 12V 300mA aux supply they mention, it's simple: you can't supply enough current to blow the diodes so no series resistor is needed. It sounds like you're using the track supply, so it would depend on the DCC supply, but I would expect you would need a limiting resistor.

    The back-to-back Zener diode clipper should work either in series or parallel although the usual configuration is parallel. You might want to try the output in the static mode with a 1K resistor instead of the turnout as the load and check the static voltage with a meter in the serial and parallel modes across the outputs.

    Feel free to ask more questions, they're a lot cheaper than the turnouts. Hope this helps.

    Mark
     

Share This Page