Kato Unitrack turnouts - position feedback

pdavidson Dec 6, 2018

  1. pdavidson

    pdavidson TrainBoard Member

    81
    27
    3
    My N scale layout is all Kato Unitrack. As far as I know, there is no built in method to monitor turnout position to provide feedback on the actual turnout position following a command to change position. I am interested in doing something (I do not know what yet) to monitor the turnout position so I can use that info in my operations.
    Since my turnouts (60 each including 6 double crossovers) are already laid and wired, I am interested in any potential solution I can come up with that would work with turnouts that are already installed because I have no intention to pull up the turnouts to make any changes. Without a fix, I must rely on faith that a turnout successfully made the changes that I commanded. So far, that has not been a big problem because my Unitrack turnouts work very well. However, I still want to make this change if possible.
    Any suggestions?
     
  2. Hardcoaler

    Hardcoaler TrainBoard Member

    10,679
    44,879
    142
    As you know, there're no available electrical contacts to use within the machine to absolutely confirm turnout position. I plan to use a capacitor-discharge circuit in concert with LEDs, all controlled by DPDT toggle switches.

    The late George Ray Stilwell, Jr's circuit is here:

    http://www.trainboard.com/highball/index.php?resources/g-ray-stilwell-bcd-circuit.153/

    Rob Paisley has an excellent website with many applications. His Kato turnout control circuit is here:

    http://www.circuitous.ca/5ToggleSingle.html

    I've messed around with these on the bench and both work well.
     
  3. pdavidson

    pdavidson TrainBoard Member

    81
    27
    3
    Thanks. The low amount of response I have received indicates the lack of a viable feedback method for a Unitrack turnout. The above links provide info on how to drive a turnout but do not provide any positive proof that the turnout has actually moved to the new position.
    Before I installed my turnouts, I played around with methods to modify the turnouts to provide an electrical indication of the turnout position in order to drive an LED on a control panel. Although I was able to find or contrive various ways to do that, none were very practical (or economical) to do for a large number of turnouts. My main current objective is to get feedback on the actual physical status of a turnout. Even though I can command a turnout to change position and my DCC system thinks it has been done (and reports it as a done deal), that is not proof that the turnout is actually in the correct position. It can fail to move due to several reasons such as a bad connection, erratic turnout or trash in the rail. I do have one idea on how to monitor and report the position of the manual switch but even that is not 100%.
    Thanks to all who bothered to read this. In the meantime, I will keep chipping away at this problem
     
  4. Hardcoaler

    Hardcoaler TrainBoard Member

    10,679
    44,879
    142
    You're thinking like a prototype railway signal engineer, insisting on a failsafe indication that the turnout movement is complete. Nothing wrong with that whatsoever. In fact, I think it's very cool.

    Just thinking out loud, if you retained the power routing feature on the Kato turnouts and built a circuit to sense current at the normal and diverting routes, you'd have fair assurance that he turnout has indeed moved. This would be tough to execute at a passing siding however, as a turnout at the east end set for the diverging route would feed current into the sensing circuit at the west end and perhaps give a false indication of turnout position. You could add rail gaps to avert this problem, but gaps would give rise to wiring complexity you hoped to avoid with DCC.

    I removed the back from a Kato N Scale No. 4 Turnout some time ago and took a photo. Perhaps a link could be attached to the throwbar at the extreme right and be used to move an electrical switch or something an optical sensor could read?

    2017-11-19 Kato #4 Turnout Inside View.jpg
     
  5. RBrodzinsky

    RBrodzinsky November 18, 2022 Staff Member TrainBoard Supporter In Memoriam

    5,685
    2,780
    98
    The biggest problem is that the positioning switch within the body will move, even if the points don’t move completely (I.e. something blocking them). And, of course, the OP’s comment on not wanting to disturb the already installed track.

    Of course, this is the case with most model switches, one can know the position you’ve set the motor to move the switch to, but never know whether it truly moved.

    If one is using DCC to control the switches, then one can use JMRI to monitor position. I guess the real question is what do you want to do with this information? Signals? Automation?
     
    Hardcoaler likes this.
  6. pdavidson

    pdavidson TrainBoard Member

    81
    27
    3
    The N scale Unitrack #6 turnouts I am using are totally different design from the Unitrack #4 pictured above. However, the #6 has multiple moving pieces that are crammed into a very tight area and is a challenge to mess with. I have had one totally disassembled and tried out several mods to get feedback. It is doable but I chose not to do that due to the number of turnouts I am using. Soldering a wire to the brass plates or anything else in the turnout is tedious. It is even worse with a double crossover. Even then, getting a correct electrical reading may give a false physical reading if the points have some trash in them.
    I am DCC and I am using JMRI and I have built control panels in both LE and PE/CPE. However, here again the control panel is showing the results of sending a command rather than the actual physical situation.
    This is not a "must do" project for me but it is a hill I would like to climb.
     
  7. pdavidson

    pdavidson TrainBoard Member

    81
    27
    3
    Hardcoaler…..forgot to mention....you are on the right track concerning using the black throw bar to create a signal that would (should) confirm the turnout position. The challenge there is that the solenoid driving all the part is not very strong. If you attach any kind of switch to the throw bar, it may prevent the turnout from working. However, doing something similar is what I am looking at. The #6 throw bar is slightly different from the #4 but has the same button protruding on the track side. I am looking to see if there is a way to use that throw bar movement to act like a switch without creating a lot of resistance. That would allow me to make a mod from the top side without removing the turnout from my layout.
     
  8. NScaleKen

    NScaleKen Permanently dispatched

    157
    128
    5
    I was just thinking about this very problem, and this thread pops up.

    I drifted towards current sensing as the best solution for the same reasons discussed, that the points making electrical contact would be the most reliable sensor of the points moving to a closed or open position. I am thinking of going the isolators and unfortunately more complex wiring route. I have used isolators to make trains stop before they entire a switch from the wrong direction. Kato #4 switches with the power routing configured to be on. Last section being driven by the switch so locomotives are unpowered unless the switch is in the correct position and powering the proper line out of it. Since I am just making my layout design now, I can consider having this standard so it protects from shorts (though I do have PSX4 circuit breaker panel), and also have current sensors that tell JMRI which track is powered and thus the switch position.

    Does anyone see any problems with that method? Needs to be power routing switches only, are Kato #6 only non power routing? can they be changed to be like a #4 and selectable routing? And then what would be used to sense the voltage? could a current change based occupancy sensor be just assigned as a turnout sensor in JMRI and when it sends a signal that is interpreted as sensor 1 or 2 on a 2 sensor feedback setup in JMRI? Come to think of it, would ignoring the power routing but just using a sensor like that in conjunction with the switch flipping be enough? a timer of some sort so JMRI or the sensor knows when to look and when to ignore fluctuations in current, to discriminate between the switch throwing and trains going by?

    I guess I will be giving it a try, wouldn't mind trying to set up occupancy detection on a siding and I guess I could see how the device works and think about it more then. Am I missing something basic about all this?
     
  9. NScaleKen

    NScaleKen Permanently dispatched

    157
    128
    5
    as far as switches causing resistance, a leaf spring of thin copper would do and should not prevent the throw bar closing. Like the material used for the contact strips in locomotive frames. Edit: another option is a ramp and a switch with a roller on the end. I have some like this for 3d printer end of travel signalling, the roller rides on the ramp and allows less force to close the switch. I dont use them that way, actually bought the wrong type, but that is apparently what the little wheel on the small end stop switches I got is for.
     
  10. mrtinvan

    mrtinvan TrainBoard Member

    34
    8
    9
    Might take some experimentation, but a reflected IR sensor might be able to detect the position of the black throw bar, without ever having to touch it. Might have to paint it silver or something, and hide the sensor in a switch shack.
     
  11. Hardcoaler

    Hardcoaler TrainBoard Member

    10,679
    44,879
    142
    I think this method holds some promise. The Kato turnout offers an exposed finger tab to move turnouts manually and by drilling a hole in it, an arm could be mounted to move an activate an IR sensor.

    The prototype actually senses that points have closed tight and assuring that result on a model might be difficult. Activating the turnout on a model might flow current or activate a sensor, but a small gap between points and stock rail could still present.

    For my road, I'll be content with a capacitor discharge circuit and indicator LEDs. It's been my experience that once a model turnout is tested in place and the action is reliable, I don't often have trouble with it. My crews, passengers and neighboring communities will just have to suffer with anxiety as trains make their way long the line. :)
     
  12. warnerj01

    warnerj01 TrainBoard Supporter

    130
    40
    16
    Are your switches isolated after the Frog? If not, would it be possible to isolate the track after the frogs and add feeders? If so, I have an idea. I'm short on time, so I can't add more.
     
  13. pdavidson

    pdavidson TrainBoard Member

    81
    27
    3
    Jerry,
    Short answer is "it depends" on which Unitrack turnout you are using. I am using the N scale #6. That means the frog rails are turned on/off depending on the active route. One is always dead and the other is hot. I think the #4 is different.
    Paul D
     
  14. Pirateo40

    Pirateo40 New Member

    3
    4
    1
    Old thread, but in case anyone is following - I just came up with this circuit to provide positive feedback on the turnout position for Kato. Also LED indicators, and using the Stilwell capacitor throw: https://photos.app.goo.gl/LYce6BdgGsx2Sj6PA
     
    RBrodzinsky and drbnc like this.
  15. Hardcoaler

    Hardcoaler TrainBoard Member

    10,679
    44,879
    142
    That's pretty neat! Very clever and it should be highly dependable. Though, because he uses a momentary contact toggle, the LED position indicator doesn't stay lit. Nonetheless, there's another Stillwell circuit (and one by Rob Paisley) that uses a DPDT toggle that keeps the position indicator LEDs lit.
     
  16. Pirateo40

    Pirateo40 New Member

    3
    4
    1
    The hall effect keeps the LED lit showing the actual position of the turnout, no need for DPDT switches. Even if the turnout is manually thrown the LED will indicate the position. It can drive two LEDs, one trackside and one on a control panel
     
    Hardcoaler likes this.
  17. Hardcoaler

    Hardcoaler TrainBoard Member

    10,679
    44,879
    142
    Oh, I see. The video shows only a red LED lit, so I was a little confused. A really neat application there!
     
  18. Pirateo40

    Pirateo40 New Member

    3
    4
    1
    Yeah, I was lazy and only set it up to show one LED. Probably should have shown the other one too. It just requires a 2nd hall effect mounted under the other magnet
     
    Hardcoaler likes this.
  19. traingeekboy

    traingeekboy TrainBoard Member

    5,677
    580
    82
    I was planning a layout in a different scale, but had the same problem.

    I am eyeing these toggles as part of a control panel.

    [​IMG]
    They are Fleischmann 6900.
     
    Hardcoaler likes this.

Share This Page