Is anyone using Rob Paisley's signal circuits with his 2011 circuit boards on a layout with Circuitron Tortoise turnout controllers?
I have installed a 2 track crossing circuit using a dedicated power supply and have found out that I get a nuisance activation of the signal crossing devices each time I activate a turnout Tortoise. Have been working with Rob tirelessly to resolve this but have yet to have any luck. I have tried two separate dedicated power supplies for the Tortoises with same result.
I am interested in your results and if you have these systems on a layout without any problems I am interested in the power supplies you have used.
And if you had similar issues what did you do to resolve them.
Thanks for your replies.
John
I've got a couple of Rob's grade crossing circuits, and I love them. One drives just flashers and the other drives both flashers and physical crossing gates. They are controlled by optical detectors between the rails.
Since you say you "activate" a Tortoise, I assume you mean you are using that Tortoise to flip switch points, right? I do that, but because I've got crossing gates I also have a Tortoise dedicated to the gates, which is controlled by the detector board. Which one causes the trouble? Also, is it a particular Tortoise, a group of them, or all of them?
Do you use the same power supply to drive the detection circuit that you use to drive your turnout Tortoises?
I run all of my Tortoises off a small 8-volt supply. That's all that supply is responsible for. The detectors, on the other hand, are powered by a 12 VDC 4-amp supply that also powers structure lights, street lamps and a few Miller Engineering animated signs. Because the detection circuits are some distance apart, they happened to each end up on a different power supply.
Are you using optical detection for your grade crossings? Are the optical detectors between the ties, as recommended, and are they close to the turnouts being activated when a problem occurs? Have you tried varying the room lighting to see if that makes a difference?
My guess, pending any more information, is that you're running the Tortises and the detection circuit boards off the same supply, and that it's inadequate for the job. A Tortoise uses slightly more power when it's moving the points, and if the supply is marginal for the job, the detector circuit may be seeing the slight drop in output of the photodetectors and interpreting that as a train.
It takes an iron man to play with a toy iron horse.
Mister Beasley
I have copied your response (thank you) here and will reply to your points with Bold print.
I've got a couple of Rob's grade crossing circuits, and I love them. One drives just flashers and the other drives both flashers and physical crossing gates. They are controlled by optical detectors between the rails. I also plan to add a 2nd crossing with just flashers as you have done. So far I have only installed the Flasher & Gate circuits and they indeed are activated using optical sensors provided by Rob.
Since you say you "activate" a Tortoise, I assume you mean you are using that Tortoise to flip switch points, right? Correct . I do that, but because I've got crossing gates I also have a Tortoise dedicated to the gates, which is controlled by the detector board. I as well am using a Tortoise to drive the gates in the same way you did. No problem that I can detect due to that Tortoise, but in fact if the circuit is already being activated I am not sure I would be able to tell. When the other Turnout Tortoises cause an activation, I see no impact by operation of another Tortoise. The only way to clear a nuisance activation is to hit the STOP button in Rob's circuit. Which one causes the trouble? Turnout Tortoises Also, is it a particular Tortoise, a group of them, or all of them? Every Tortoise on the Layout and some are up to 8 feet away from any Signal Circuit wiring.
Do you use the same power supply to drive the detection circuit that you use to drive your turnout Tortoises? No they are independent Supplies and each is dedicated to its own function. The Signal Circuit suppy is definitely regulated, and I believe the Tortoise one is too.
I run all of my Tortoises off a small 8-volt supply. My Tortoise supply is 12 vdc. That's all that supply is responsible for. The detectors, on the other hand, are powered by a 12 VDC 4-amp supply that also powers structure lights, street lamps and a few Miller Engineering animated signs. My Signal Circuit supply is also 12 vdc and 1555 ma and is dedicated. Because the detection circuits are some distance apart, they happened to each end up on a different power supply. Due to the layout geometry wiring for both the Signal Circuits and Tortoise in some areas pass closely to each other but I have tried to not run them co-linear to avoid coupling.
Are you using optical detection for your grade crossings? I've got a couple of Rob's grade crossing circuits, and I love them. One drives just flashers and the other drives both flashers and physical crossing gates. They are controlled by optical detectors between the rails.
Are you using optical detection for your grade crossings? Yes they are povided by Rob. Are the optical detectors between the ties, as recommended, and are they close to the turnouts being activated when a problem occurs? Yes, interesting point - my Start detectors in all cases are close to a Turnout but only Two of the Turnouts at one end of the circuit are Tortoise powered. The other end is manually actuated. Have you tried varying the room lighting to see if that makes a difference? NO.
My guess, pending any more information, is that you're running the Tortises and the detection circuit boards off the same supply, No and that it's inadequate for the job. A Tortoise uses slightly more power when it's moving the points, and if the supply is marginal for the job, the detector circuit may be seeing the slight drop in output of the photodetectors and interpreting that as a train.
These were all good questions and statements. I have yet to hear back from Rob on some filtering I have tried across the Tortoise power supply output - capacitors. No effect. I now plan to operate some tortoises with a nine volt battery to see what the result is to see if I can determine if it is sensitive to the power supply type. Maybe I just need to get a old lawnmower battery to drive the tortoises. There has to be a reasonable solution. I have parts coming soon to filter the optical sensor circuits in a fashion recommended by Rob.
I hope I get comments from some more thinkers on this issue. Let me know what you think based on my responses, Mister Beasley
Thanks
My next suggestion was going to be the one you came up with - run the Tortoises on 9-volt batteries to see what happens. You can use a small battery like the ones for smoke detectors to do this. Just take it out when you're done.
There is a time and a place for a tin-foil hat, too. Try making up a makeshift metal enclosure for the circuit board, and ground it to see if you can block any stray RF signals. This is a stretch. I can't see what would make this necessary, but it's something to try.
When you throw a turnout, does it do anything else? Does it activate position-indicator LEDs, or is there a power-routing turnout that turns on power to an engine?
Have you tried running the detector circuit on batteries?
Try taking one photo-detector at a time out of the circuit. Since you're doing this double-track, you probably need to jumper over it since a detection actually drops the voltage to zero. You could have a sensitive detector, I suppose, although once again it's a stretch. I'd also make sure there's a good, strong light source for every detector, just to make sure it's not something silly like you standing in front of the light to push the control pane button and inadvertently tripping the detector.
I have decided to run some tests on the systems to see the impact of various configurations. I have published two documents in the link below to Dropbox..Stick in your browser if you are interested in spending a litte (or lot) of time helping me sort this out.
https://www.dropbox.com/sh/d4kr3e21cc0s757/AACr8FTsASFbE7KXHMx4wRKQa?dl=0
The numbered Turnouts are those that are powered with Tortoise units. The Signal circuit boards are basically located just to the left of Turnouts 11 and 12. We have used one set of the SPST switches in Turnouts 11 and 12 to control the operational logic of the Signal circuit to keep the signals off for any trains headed to the yard as determined by Turnout positions for 11 and 12. So that explains the interest in modifying those two turnouts' configurations for the tests. What I found very interesting is the difference in results by just switching the Tortoise power supply from an AC/DC adapter to a 9 vdc battery. Yet there are some subttle differences in the other configurations. I am still struggling with this and anyone's help is certainly welcome. In the chart of results the x's mean that 11 and 12 were not actuated since they had no power attached. The N' was one circumstance where it appeared that if the tortoise was actuated each way quickly no upset occurred but if there was some time inbetween actuations them the Crossing signal circuit was upset. (go figure).
556 (and 555 single) put quite a large spike on the supply rail when they switch states. I see Rob's circuit calls for a single 10uF cap. I have found a 47uF and .1uf in parallel do a much better job of keeping the supply rail clean. The 47uF takes care of the bulk of the spike while the .1uF picks up the high frequency leftover. Both need to be as close to the 555/556 Vin pin as practical.
False triggers are usually one of two things - noise on the supply rail -or- your input (trigger) wiring acting as an antenna. Try the above filtering to clean up the supply rail and do something to shield the input wiring i.e. twist the wires; use shielded cabling with the shield grounded; or even temporarily wrap tin foil around the input leads and ground it.
Here you can see the filter cap pair utilized on one of my own 555 circuits. It works well.
Alan
Freelancing the LK&O Railroad
Multiple caps are almost always better for power filtering, since the frequency response of the cap is determined by its value.
If that doesn;t work - the spikes may be being induced on the runs from the sensors, not via the power supply. Given that it's throwing OTHER Tortoises that trigger the crossing, it's probably not power spikes by the crossing controller. A small (.01uF) cap across each of the lines to the sensors may help. The ot NCE Switch-Its had a similar problem if the wire run to a local control pushbutton was too long - the input is low impedence and highly sensitive and can easily be triggered by induced voltages. Make sure the power runs to the Tortoises that falsely trigger the crossing circuit don;t run in parallel to the sensor lines for the grade crossing - in fact nothing should run in prallel with the sensor leads to the circuit. Another thing that may help - get a hunk of Ethernet cable and connect the senors via a twisted pair. The twisting alone may be enough to alter the characteristics to reduce the false triggering.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
The discussion of long leads made me think of something. I have my detector boards, signals and detectors all clumped together as groups, although one crossing is on one side of the layout and the other crossing some distance away. The runs from the power bus to the board, the detectors to the board and the board to the signals themselves are relatively short. If you've instead centrally mounted your two circuit boards and run long wires to the detectors and signals, that could contribute to the sensitivity of the systems to interference.
I am really happy with all the responses I have gotten on this thread, and I need to figure out my sequence for trying them out. I think I will first start with making sure I have the maximum separation possible between the sensing leads for the signals and the Tortoise circuitry. I will then attempt the rather simple approach that circuitron recommended by applying back-to-back zener diodes across the tortoise input. Paisley has suggested and I am getting parts for some filtering on the sensing leads themselves. And there are a few other things that have been suggested that I think I will have the opportunity to try. It will probably take me till mid next week to get through a lot of this stuff and then I will report back on what I have found out.
Thanks again, John
I'm taking a few days at the beach myself. My browsing time may be limited, so please don't think I've abandoned you.
We love a good mystery here.
Well here is the answer to the problem. After much testing, reconfiguring and contemplation all the signs pointed to the system becoming upset by EMI spikes caused by the Tortoise system. My signal system and the tortoise system are completely independent and electrially separated all the way to the 110 wall outlet. However, tortoises not even close to the system wiring were causing upset. Fact was that due to signal system needs the photo sensors for the signals on one end of my double track crossing were quite close to tortoises. Additionally I had some logic wires attached to 2 of the tortoises to control the signals via Turnout positions. Removing the logic wires decreased the sensitivity significantly but not entirely. Tells me that worst performing Tortoise motors send the signals down the wires.
So I resolved the issue by 1. relocating the sensors closer to the gate crossing to avoid the need for turnout positioning logic ( gives me an undesireablely short time for the gates to actuate at present. ) Photosensors are still close to Tortoise machines. But the final touch was to install an RC filter network to the three board inputs from the photosensors to the Track 2 PWB. Why Track 2? - who knows, but the system was only sensitive when the Track 2 board was connected. ( I tried a second board with same results)
I may get enough energy some day to reconfigure back to my original photosensor location and the associated logic to see if the RC filter corrects it. But now I am tired.
Thanks for your suggestions.