Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Auto Reverser PSX-AR

11521 views
78 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Thursday, December 15, 2016 7:15 PM

Randy, I am assuming that the solid state PSX-AR will perform both functions simultaneously. But, let's assume that there is a delay from one function to another. In that case, two sets of gaps would be required at either end of the reverse loop. The track between each set of gaps would maintain a fixed polarity. If that proves to be the OP's problem, we can provide a wiring and gapping solution for him. I drew one up that seems to work, at least on paper.

Rich

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Friday, December 16, 2016 5:58 AM

richhotrain

 

 
rrinker

If you use a Frog Juicer to power the frogs, and the gaps on the frog rails are also the gaps that isolate the reverse loop powered byt he PSX-AR, you now have back to back reverse loops (the Frog Juicer is effectively an autoreverser for the frog polarity and automatically matches the frog polarity to the route the train is taking through the turnout, vs using switch machine contacts which could be wired backwards, meanining instant short until you get the feeders right). Both products use fast acting electronic detection and switching for the short circuit condition, and this is a possible source of trouble.

 

 

Since you are using a Peco Electrofrog, the frog is already powered so there is no need for a Frog Juicer. But, recall, the Peco Electrofrog is a power routing turnout. Depending upon the route selected, both rails beyond the frog will be powered with the same polarity. So, the two inner frog rails need to gapped. These same gaps can be used to isolate the reverse loop.

 

Rich

 

Rich: By "So, the two inner frog rails need to gapped" do you mean modifying the turnout to make it "DCC friendly" as per Allan Gartner.  If so I am afraid that that is not an option in N scale. Our turnouts are too small and I for one do not have the skills to attempt such a modification. And if you use a Tam Valley Frog Juicer the modification is not required. However as pointed out a Frog Juicer is unlikely to work with the PSX-AR as it will be in conflict.

On the surface my problem does look like a wiring issue. However I have (1) removed all but one of the feeder wires (there was 3) to the reversing loop thereby eliminating any wiring issues there; (2) replaced the original turnout and the Peco motor with a new set (after checking with my CDU that the point motor worked) and (3) I have moved the insulating gaps back a little bit to give the Turnout plenty of time to change after the Loco crosses the gap. 

Still no joy. This is what happens: when the power is turned on the turnout changes to “Normal” (proving that pins 1 & 3 are correctly wired). The Loco proceeds around the loop but once it crosses the insulated gap it shorts and stays that way until you manually change the turnout. Once you do that it moves off ok after a few seconds. In reverse the same thing happens. Just once it did move the turnout after a number of seconds. Not happened since.

As far as I can tell the polarity from the main into the reversing loop on normal is the same and that is not an issue. And after all there are only 4 wires involved - 2 to the turnout and 2 to the track.

So my only conclusion is that it is an N scale issue as the turnout needs the HO "DCC friendly" modification . . .

Is there anyone out there using a PSX-ARSC with unmodified N scale turnouts?

I really do want to thank everybody who has responded to this tread, especially Rich, Randy and bavrail. I have learnt a lot but for now I am going to give in and use a pushbutton to control the turnout and let the PSX-AR control the reversing loop, which it does very well.

 

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Friday, December 16, 2016 6:10 AM

Dennis, you do not need to mess with the turnout to gap the inner frog rails. Just use plastic rail joiners on those two inner rails as opposed to metal rail joiners.

Rich

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Friday, December 16, 2016 6:23 AM

richhotrain

Dennis, you do not need to mess with the turnout to gap the inner frog rails. Just use plastic rail joiners on those two inner rails as opposed to metal rail joiners.

Rich

 

Rich, I misunderstood. Of course I have insulated the two inner frog rails (you have no choice with DCC but to do that with every turnout, or the trains will not run). I have also insulated the reverse loop from the turnout at both ends. I still can only conclude that it is a N scale problem as other people are not having my issues. I have also swapped out the PSX-ARSC so that is not at fault. 

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Friday, December 16, 2016 6:26 AM

You mention that when you let the PSX-AR control the reversing loop it does very well.  So, I assume that you have disconnected the two J9 wires on the PSX-AR and then the operation of the reverse loop works just fine - - - from either direction.

Rich

Alton Junction

  • Member since
    July 2006
  • From: The Villages, Fl
  • 59 posts
Posted by bavrail on Friday, December 16, 2016 7:25 AM

Denver

 Still no joy. This is what happens: when the power is turned on the turnout changes to “Normal” (proving that pins 1 & 3 are correctly wired). The Loco proceeds around the loop but once it crosses the insulated gap it shorts and stays that way until you manually change the turnout. Once you do that it moves off ok after a few seconds. In reverse the same thing happens. Just once it did move the turnout after a number of seconds. Not happened since.

 

Dennis

I have a real dumb question. After the trains goes into the reversing loop is there power to the piece of track between the second set of gaps and the turnout ?

Are you sure you are getting a short after the train gets to the second set of gaps ?

Almost sounds to me as if there is no power on the other side of the second gap and when you move the turnout the track between the turnout and the second gap gets power.

We use all peco insulfrog turnouts and have no problems. 

 

WS

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Friday, December 16, 2016 8:04 AM

bavrail
 
 
Denver

 Still no joy. This is what happens: when the power is turned on the turnout changes to “Normal” (proving that pins 1 & 3 are correctly wired). The Loco proceeds around the loop but once it crosses the insulated gap it shorts and stays that way until you manually change the turnout. Once you do that it moves off ok after a few seconds. In reverse the same thing happens. Just once it did move the turnout after a number of seconds. Not happened since. 

Dennis

I have a real dumb question. After the trains goes into the reversing loop is there power to the piece of track between the second set of gaps and the turnout ?

Are you sure you are getting a short after the train gets to the second set of gaps ?

Almost sounds to me as if there is no power on the other side of the second gap and when you move the turnout the track between the turnout and the second gap gets power.

We use all peco insulfrog turnouts and have no problems.  

I suspect that Dennis may be out touch during the day so let me take the liberty to jump in here.

One thing to recall is that on Dennis' layout the turnout that controls entry and exit to the reverse loop is a Peco Electrofrog, not an Insulfrog, so there is an issue with the inner frog rail opposite the active route having the same polarity as its adjacent stock rail. That may well be at the heart of the problem.  From Dennis' description of his wiring, it seems that all of the rails are powered at all times.

What I would do is to disconnect the two wires from the J9 port on the PSX-ARSC and then run the locomotive into the reverse loop by entering the loop via the straight through route. Then, stop the locomotive before it reaches the gaps at the exit point of the loop. Manually flip the turnout point rails and then restart the locomotive. If it makes it through the turnout without a glitch, then we know that the reverse loop is wired correctly and that the turnout is not at fault. If the locomotive does not proceed through the turnout, then it is a wiring problem.

If the locomotive does make it through the turnout without a glitch using the previous procedure, then it is likely a gapping problem associated with the use of an Electrofrog turnout.  There is a relatively easy fix for that problem.

I would rule out the PSX-ARSC as the problem because of the testing procedures that Dennis has already undertaken.

Rich

Alton Junction

  • Member since
    July 2006
  • From: The Villages, Fl
  • 59 posts
Posted by bavrail on Friday, December 16, 2016 3:49 PM
Good Idea

WS

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Friday, December 16, 2016 4:41 PM

richhotrain

 

 
bavrail
 
 
Denver

 Still no joy. This is what happens: when the power is turned on the turnout changes to “Normal” (proving that pins 1 & 3 are correctly wired). The Loco proceeds around the loop but once it crosses the insulated gap it shorts and stays that way until you manually change the turnout. Once you do that it moves off ok after a few seconds. In reverse the same thing happens. Just once it did move the turnout after a number of seconds. Not happened since. 

Dennis

I have a real dumb question. After the trains goes into the reversing loop is there power to the piece of track between the second set of gaps and the turnout ?

Are you sure you are getting a short after the train gets to the second set of gaps ?

Almost sounds to me as if there is no power on the other side of the second gap and when you move the turnout the track between the turnout and the second gap gets power.

We use all peco insulfrog turnouts and have no problems.  

 

 

I suspect that Dennis may be out touch during the day so let me take the liberty to jump in here.

 

One thing to recall is that on Dennis' layout the turnout that controls entry and exit to the reverse loop is a Peco Electrofrog, not an Insulfrog, so there is an issue with the inner frog rail opposite the active route having the same polarity as its adjacent stock rail. That may well be at the heart of the problem.  From Dennis' description of his wiring, it seems that all of the rails are powered at all times.

What I would do is to disconnect the two wires from the J9 port on the PSX-ARSC and then run the locomotive into the reverse loop by entering the loop via the straight through route. Then, stop the locomotive before it reaches the gaps at the exit point of the loop. Manually flip the turnout point rails and then restart the locomotive. If it makes it through the turnout without a glitch, then we know that the reverse loop is wired correctly and that the turnout is not at fault. If the locomotive does not proceed through the turnout, then it is a wiring problem.

If the locomotive does make it through the turnout without a glitch using the previous procedure, then it is likely a gapping problem associated with the use of an Electrofrog turnout.  There is a relatively easy fix for that problem.

I would rule out the PSX-ARSC as the problem because of the testing procedures that Dennis has already undertaken.

Rich

 

Hi Rich: you are right, I am in Australia, which is why I generally write in your mornings. It is 8:30am here right now and i have not been out to the railway [in my garage] this morning. However I have already tested the turnout manually as you suggest and there is no problem - the Loco sails through.

What is your "relatively easy fix"?? Given that bavrail has no problems using an insulfrog turnout I am convinced that it is a problem with the frogs rails. I guess I could just buy an insulfrog turnout.

I am going to test the turnout with a frog juicer this afternoon. Will let you know how I get on.

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Friday, December 16, 2016 5:27 PM

The first diagram below illustrates the straight through route into the reverse loop. Notice the red and blue color rails to indicate polarity.  The polarity of the reverse loop matches the polarity of the straight through route. The circles indicate gaps.

The diagram shows the addition of two powered track sections between the reverse loop and the turnout. The purpose of these two track sections is to provide a length of track to give the turnout time to change routes before the locomotive reaches the turnout. Note the polarity of these added track sections since that is critical to the operation without a short.

As the locomotive exits the reverse loop, the PSX-ARSC detects a short created by mismatched polarities and reverses the polarities inside the loop.

In the second diagram, the locomotive proceeds into that added section of track and the PSX-ARSC changes the route of the turnout.

The next time that the locomotive approaches the reverse loop, the turnout is set to divergent so that the locomotive will enter the loop with the same polarity as the divergent route.

As the locomotive exits the loop at the bottom of the diagram, the PSX-AR will detect a short created by mismatched polarities and reverse the polarities inside the loop and change the route of the turnout to straight through.

See if that works.

Rich

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Friday, December 16, 2016 6:41 PM

richhotrain

The first diagram below illustrates the straight through route into the reverse loop. Notice the red and blue color rails to indicate polarity.  The polarity of the reverse loop matches the polarity of the straight through route. The circles indicate gaps.

The diagram shows the addition of two powered track sections between the reverse loop and the turnout. The purpose of these two track sections is to provide a length of track to give the turnout time to change routes before the locomotive reaches the turnout. Note the polarity of these added track sections since that is critical to the operation without a short.

As the locomotive exits the reverse loop, the PSX-ARSC detects a short created by mismatched polarities and reverses the polarities inside the loop.

In the second diagram, the locomotive proceeds into that added section of track and the PSX-ARSC changes the route of the turnout.

The next time that the locomotive approaches the reverse loop, the turnout is set to divergent so that the locomotive will enter the loop with the same polarity as the divergent route.

As the locomotive exits the loop at the bottom of the diagram, the PSX-AR will detect a short created by mismatched polarities and reverse the polarities inside the loop and change the route of the turnout to straight through.

See if that works.

Rich

 

Let me get this straight. In effect I divide the return loop into 3 separate isolated blocks and deliberately reverse the polarity of the two new blocks closest to the turnout. OK I will try.
Dennis

 

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Friday, December 16, 2016 9:17 PM

Denver

 

 
richhotrain

The first diagram below illustrates the straight through route into the reverse loop. Notice the red and blue color rails to indicate polarity.  The polarity of the reverse loop matches the polarity of the straight through route. The circles indicate gaps.

The diagram shows the addition of two powered track sections between the reverse loop and the turnout. The purpose of these two track sections is to provide a length of track to give the turnout time to change routes before the locomotive reaches the turnout. Note the polarity of these added track sections since that is critical to the operation without a short.

As the locomotive exits the reverse loop, the PSX-ARSC detects a short created by mismatched polarities and reverses the polarities inside the loop.

In the second diagram, the locomotive proceeds into that added section of track and the PSX-ARSC changes the route of the turnout.

The next time that the locomotive approaches the reverse loop, the turnout is set to divergent so that the locomotive will enter the loop with the same polarity as the divergent route.

As the locomotive exits the loop at the bottom of the diagram, the PSX-AR will detect a short created by mismatched polarities and reverse the polarities inside the loop and change the route of the turnout to straight through.

See if that works.

Rich

 

 

 

Let me get this straight. In effect I divide the return loop into 3 separate isolated blocks and deliberately reverse the polarity of the two new blocks closest to the turnout. OK I will try.
Dennis

 

 

No matter how it is wired the turnout always switches to Reverse as it leaves the lower block and enters the loop (not good if there cars behind the Loco) and when the Loco gets into the same block as the turnout the turnout switches back to Normal and the Loco freezes.

Any more ideas - it almost worked this time.

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Friday, December 16, 2016 10:52 PM

This arrangement should work if wired correctly. 

When the loco proceeds on the straight through route of the turnout, the PSX-ARSC will be unaffected as the loco crosses the gaps into that new track section and remain unaffected as the loco crosses the gaps into the loop. That is because the turnout, the newly added track section and the loop are wired the same way (the polarities match). Only when the loco crosses the gaps to exit the loop will the PSX-ARSC be affected since the polarity of the loop will not match the polarity of the newly added track section between the loop and the divergent route of the turnout.

The fact that the turnout is changing its route from straight through to divergent as the loco enters the loop indicates that the wiring is incorrect.

In terms of wiring, the turnout should be wired with feeders that are connected to the main bus.  The two newly added track sections should also be wired to the main bus, taking care to match the polarities shown in that upper diagram. The loop should be wired the same as the straight through route of the turnout, but the feeders to the loop should be connected to the output side of the PSX-ARSC. The input side of the PSX-ARSC should be connected to the main bus.

In terms of gaps, all 8 of those gaps should be plastic rail connectors so that the loop and the two newly added track sections are isolated from each other and isolated from the turnout.

Rich

Alton Junction

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Saturday, December 17, 2016 5:43 AM

To make the area of your layout perform in the way that you described it, it seems that the wiring of the loop does not match the polarity of the straight through route of the turnout.  Also, the wiring of the newly added track section between the loop and the divergent side of the turnout seem to be reversed from my diagram. If so, that is the problem.

Reverse the loop wiring and the wiring of that upper track section between the loop and the divergent side of the turnout.

Rich

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Saturday, December 17, 2016 6:00 AM

richhotrain

This arrangement should work if wired correctly. 

When the loco proceeds on the straight through route of the turnout, the PSX-ARSC will be unaffected as the loco crosses the gaps into that new track section and remain unaffected as the loco crosses the gaps into the loop. That is because the turnout, the newly added track section and the loop are wired the same way (the polarities match). Only when the loco crosses the gaps to exit the loop will the PSX-ARSC be affected since the polarity of the loop will not match the polarity of the newly added track section between the loop and the divergent route of the turnout.

The fact that the turnout is changing its route from straight through to divergent as the loco enters the loop indicates that the wiring is incorrect.

In terms of wiring, the turnout should be wired with feeders that are connected to the main bus.  The two newly added track sections should also be wired to the main bus, taking care to match the polarities shown in that upper diagram. The loop should be wired the same as the straight through route of the turnout, but the feeders to the loop should be connected to the output side of the PSX-ARSC. The input side of the PSX-ARSC should be connected to the main bus.

In terms of gaps, all 8 of those gaps should be plastic rail connectors so that the loop and the two newly added track sections are isolated from each other and isolated from the turnout.

Rich

 

Rich: IT WORKS! I found my wiring error (or two) and now the PSX-ARSC works flawlessly. Thank you very much for all your time and effort. It has been a long week. With your permission I will publish your drawing on my web site (denversrailroads.com) once I have experimented with the length of the two newly added track sections. I will also send the link to Tony’s Train Exchange, who have yet to reply to my follow up email, and suggest they send it out with any orders. Your solution is quite different to the wiring they suggest in the manual.
Originally I had wanted to buy 3 Jack Wabbit’s to automate 12 turnouts but in view of our dramas and the fact that the manual states that the Jack Wabbit out of box is configured for Tortoise and not snap coils I am going to give that idea away. A diode matrix and push buttons is much easier.
Have a happy Christmas.

 

Dennis
  • Member since
    May 2015
  • 37 posts
Posted by Denver on Saturday, December 17, 2016 6:04 AM

richhotrain

To make the area of your layout perform in the way that you described it, it seems that the wiring of the loop does not match the polarity of the straight through route of the turnout.  Also, the wiring of the newly added track section between the loop and the divergent side of the turnout seem to be reversed from my diagram. If so, that is the problem.

Reverse the loop wiring and the wiring of that upper track section between the loop and the divergent side of the turnout.

Rich

 

The big wiring mistake (which I missed at first because I was checking all the track wiring) was that I had the turnout connections to J9-3 and J9-1 back to front. Once I corected that I found my track wiring error quite easily. Thanks once again.

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Saturday, December 17, 2016 6:08 AM

Awesome!

So it now works flawlessly?

Rich

Edit Note: Oops, I missed the first of your two most recent posts. Congratulations on your success. You did well to stick with it until the problem was solved.  Yes, feel free to use the diagram as you see fit.

I am attaching a diagram that I drew up this morning showing the wiring connections in more detail.

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Saturday, December 17, 2016 6:10 AM

richhotrain

Awesome!

So it now works flawlessly?

Rich

 

Yes, flawlessly.

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Saturday, December 17, 2016 6:24 AM

Dennis, I missed the first of your two most recent replies when I asked that question.  Check back on my edit to see my further reply.

I can't help but wonder if those two newly added track sections are really necessary now that you have corrected the PSX-ARSC wiring. My turnouts are controlled by Tortoises so I use the simpler PSX-ARs, not the PSX-ARSCs. But the PSX units should be fast enough to change the turnout routes instantly so that a short will not occur.

At this point, you may decide to leave well enough alone, but it would be interesting to experiment by removing those two newly added track sections to see if it works without them.

Rich

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Saturday, December 17, 2016 6:47 AM

richhotrain

Dennis, I missed the first of your two most recent replies when I asked that question.  Check back on my edit to see my further reply.

I can't help but wonder if those two newly added track sections are really necessary now that you have corrected the PSX-ARSC wiring. My turnouts are controlled by Tortoises so I use the simpler PSX-ARs, not the PSX-ARSCs. But the PSX units should be fast enough to change the turnout routes instantly so that a short will not occur.

At this point, you may decide to leave well enough alone, but it would be interesting to experiment by removing those two newly added track sections to see if it works without them.

Rich

 

I have just sent an email to Tony’s Train Exchange with the conclusion – “After adding the two additional track sections the PSX-ARSC now works flawlessly.  It seems that the PSX-ARSC needs the additional sections to give it time to change routes.” If your additional tracks were not needed I would never have had a problem in the first place as originally I only had 4 wires and I tried every permutation. It looks to me as if these DCC Specialists boards (PSX, Jack Wabbit, etc) are designed for the slow moving Tortoises that you are using and have been modified to work with Snap Coil machines and in the case of the PSX need time to react to a change instruction.
The wiring error I mentioned was to your new configuration. It was not an error made days ago and never corrected. Remember I have tried many different wiring options during this last week.
I am going to reduce the length of the additional sections from their current 2’6” to about 9” but I certainly do not plan to remove them.
Dennis
  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Saturday, December 17, 2016 7:01 AM

Let us know how Larry Meier responds to your email. I've got a feeling that he will say that the additional track sections are not needed. I wonder if he will suggest a timing change on the PSX-ARSC?

Rich

Alton Junction

  • Member since
    July 2006
  • From: The Villages, Fl
  • 59 posts
Posted by bavrail on Saturday, December 17, 2016 5:58 PM

Denver

 It looks to me as if these DCC Specialists boards (PSX, Jack Wabbit, etc) are designed for the slow moving Tortoises that you are using and have been modified to work with Snap Coil machines and in the case of the PSX need time to react to a change instruction.

 
Dennis
 
I don't think the above statement applies. We use two PSX-ARSC's. One has gaps about 2 inches and 7 inches from the turnout. The other has gaps 3 inches and 6 inches from the turnout. 
They both reverse polarity and throw a peco turnout before the engine gets there.
 
Cheers

WS

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Saturday, December 17, 2016 9:00 PM

 See, i was right after all, you need 2 sets of gaps or else it all just keeps shorting out. The diagram Rich made is exactly what I said on the first page. Without power routing turnoyts, like Atlas, it would work the usual way, just one set of gaps at the entrance of the loop and one set at the exist, the PSX-AR would handle the track polarity and throw the switch to the proper route. When there is a power routing turnout, that extra section and a second set of gaps at the turnout solve the problem of the PSX-AR reversing the loop polarity only to have the new point position short it again, making the PSX-AR reverse the track and again throw the switch, causing a short again.. over and over. Assuming the polarity reversal and switch throw happen exactly at the same time, it would appear as just a short and the breaker function of the PSX-AR would just cut power and give up. Given that there is some circuitry involved and it's not exactly the same on the part feeding the turnout control and the part feeding the polarity control, it's nearly impossible for each to happen perfectly simulataneously (it's the slight delays - an extra transistor switching, or a logic gate, that leads to race conditions in electronic circuits if you aren't careful with the design), it would probably just oscillate back and forth before finally shorting for good and killing power to the loop.

 If you look at Rich's diagram and follow the power around, assuming the train enters the loop straight across the bottom (because that's the way the power at the turnout is drawn), when the gap between the loop and the upper short section is bridged, the PSx-AR will flip the loop, making the inner track blue and the outher track red, matching the short section. It will also throw the switch, making the diverging frog rails blue - matching the short section and the loop. Perfect.

Now try it without that short section. Train enters loop at bottom. frog rails are red, as is the inside rail of the loop. So far so good. Now it crosses the gap. The iside rail actually already matches due to the power routing of the Peco turnout, but the outer rail it blue on the loop and red on the turnout, so the PSX-AR flips the loop, now the inner loop rail is blue and the outer is red. And signals the switch machine to operate. Now, which is faster, the PSX-AR's electronic sensor, or the solenoid switch motor? Because until the switch motor operates, there is NOW a short between the inner loop track and the diverging frog rails.

Since there are issues using slower relay-based autoreversers downstream of a PSX circuit breaker, I can asusme the PSX-AR uses the same ultra fast dectection circuit. And a solenoid switch motor is, let's face it, a relay. My money is on the PSX-AR activating again, but now we're back to the original condition of a short between the outer loop rail and the turnout. Or, if the turnout still completes its throw, still a polarity mismatch on the inner rail (which has flipped back to red) and the diverging frog rails (which are now blue).

This is not an issue for a manually thrown turnout ("manually" meaning push buttons and a powered switch machine or a true manual linkage like a Blue Point), only if using the auto-throw feature AND with power routing turnouts. The extra sections would be unnecessary with an Atlas turnout because the polarity of the exit rails is always the same regardless of the position of the points.

                      --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Saturday, December 17, 2016 9:27 PM
Well I think Randy has about summed it up. And bavrail is using insulfrog turnouts, which I think explains why he does not need the extra track sections. Anyway it now all works so I am not going to test fate by any further experimentation. Once again thank you all. It’s been an interesting, if time consuming, exercise.

 

Dennis
 
  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Sunday, December 18, 2016 5:02 AM

rrinker

 See, i was right after all, you need 2 sets of gaps or else it all just keeps shorting out. The diagram Rich made is exactly what I said on the first page. Without power routing turnoyts, like Atlas, it would work the usual way, just one set of gaps at the entrance of the loop and one set at the exist, the PSX-AR would handle the track polarity and throw the switch to the proper route. When there is a power routing turnout, that extra section and a second set of gaps at the turnout solve the problem of the PSX-AR reversing the loop polarity only to have the new point position short it again, making the PSX-AR reverse the track and again throw the switch, causing a short again.. over and over. Assuming the polarity reversal and switch throw happen exactly at the same time, it would appear as just a short and the breaker function of the PSX-AR would just cut power and give up. Given that there is some circuitry involved and it's not exactly the same on the part feeding the turnout control and the part feeding the polarity control, it's nearly impossible for each to happen perfectly simulataneously (it's the slight delays - an extra transistor switching, or a logic gate, that leads to race conditions in electronic circuits if you aren't careful with the design), it would probably just oscillate back and forth before finally shorting for good and killing power to the loop.

 If you look at Rich's diagram and follow the power around, assuming the train enters the loop straight across the bottom (because that's the way the power at the turnout is drawn), when the gap between the loop and the upper short section is bridged, the PSx-AR will flip the loop, making the inner track blue and the outher track red, matching the short section. It will also throw the switch, making the diverging frog rails blue - matching the short section and the loop. Perfect.

Now try it without that short section. Train enters loop at bottom. frog rails are red, as is the inside rail of the loop. So far so good. Now it crosses the gap. The iside rail actually already matches due to the power routing of the Peco turnout, but the outer rail it blue on the loop and red on the turnout, so the PSX-AR flips the loop, now the inner loop rail is blue and the outer is red. And signals the switch machine to operate. Now, which is faster, the PSX-AR's electronic sensor, or the solenoid switch motor? Because until the switch motor operates, there is NOW a short between the inner loop track and the diverging frog rails.

Since there are issues using slower relay-based autoreversers downstream of a PSX circuit breaker, I can asusme the PSX-AR uses the same ultra fast dectection circuit. And a solenoid switch motor is, let's face it, a relay. My money is on the PSX-AR activating again, but now we're back to the original condition of a short between the outer loop rail and the turnout. Or, if the turnout still completes its throw, still a polarity mismatch on the inner rail (which has flipped back to red) and the diverging frog rails (which are now blue).

This is not an issue for a manually thrown turnout ("manually" meaning push buttons and a powered switch machine or a true manual linkage like a Blue Point), only if using the auto-throw feature AND with power routing turnouts. The extra sections would be unnecessary with an Atlas turnout because the polarity of the exit rails is always the same regardless of the position of the points.

                      --Randy

 

 

Dunno. Randy, I remain conflicted. 

I understand what you are saying, but as I read what you wrote, I keep asking myself, then why doesn't DCC Specialties say so in their literature. There is no mention of a second set of gaps and an additional short track section in the manual. Although the PSX-AR manual does not include any drawings of the PSX-ARSC being used to throw the points, it does include as a feature "Automates Reverse Loop Turnouts". So, you would think that they would at least mention the issue that we are dealing with here. 

Now, you might ask yourself, then why, Rich, did you draw up the diagram with additional track sections on either end of the loop?  My answer would be that I drew the diagram as a testing technique to solve Dennis' problem. It did not surprise me that it worked because it should work. But, I still cannot help wondering if those two additional track sections are really necessary.

I can understand Dennis' reluctance to now remove the two sections, or at least remove the plastic rail gaps between the two track sections and the turnout, but my investigative mind would not be able to resist such a further test.  I do not use Electrofrogs or the Peco switch motors (I use Insulfrogs and Tortoises) or else I might try it on the workbench.

Mind you, I'm not saying that it will work without those two additional track sections, but I think that it is at least worth a try.  I sure would be interested in what Larry Meier would have to say about this.  Without those two additional track sections, the reverse loop and the turnout are still fully isolated from each other, so it is all a matter of timing.

Rich

Alton Junction

  • Member since
    May 2015
  • 37 posts
Posted by Denver on Sunday, December 18, 2016 6:47 AM

richhotrain

 

 
rrinker

 See, i was right after all, you need 2 sets of gaps or else it all just keeps shorting out. The diagram Rich made is exactly what I said on the first page. Without power routing turnoyts, like Atlas, it would work the usual way, just one set of gaps at the entrance of the loop and one set at the exist, the PSX-AR would handle the track polarity and throw the switch to the proper route. When there is a power routing turnout, that extra section and a second set of gaps at the turnout solve the problem of the PSX-AR reversing the loop polarity only to have the new point position short it again, making the PSX-AR reverse the track and again throw the switch, causing a short again.. over and over. Assuming the polarity reversal and switch throw happen exactly at the same time, it would appear as just a short and the breaker function of the PSX-AR would just cut power and give up. Given that there is some circuitry involved and it's not exactly the same on the part feeding the turnout control and the part feeding the polarity control, it's nearly impossible for each to happen perfectly simulataneously (it's the slight delays - an extra transistor switching, or a logic gate, that leads to race conditions in electronic circuits if you aren't careful with the design), it would probably just oscillate back and forth before finally shorting for good and killing power to the loop.

 If you look at Rich's diagram and follow the power around, assuming the train enters the loop straight across the bottom (because that's the way the power at the turnout is drawn), when the gap between the loop and the upper short section is bridged, the PSx-AR will flip the loop, making the inner track blue and the outher track red, matching the short section. It will also throw the switch, making the diverging frog rails blue - matching the short section and the loop. Perfect.

Now try it without that short section. Train enters loop at bottom. frog rails are red, as is the inside rail of the loop. So far so good. Now it crosses the gap. The iside rail actually already matches due to the power routing of the Peco turnout, but the outer rail it blue on the loop and red on the turnout, so the PSX-AR flips the loop, now the inner loop rail is blue and the outer is red. And signals the switch machine to operate. Now, which is faster, the PSX-AR's electronic sensor, or the solenoid switch motor? Because until the switch motor operates, there is NOW a short between the inner loop track and the diverging frog rails.

Since there are issues using slower relay-based autoreversers downstream of a PSX circuit breaker, I can asusme the PSX-AR uses the same ultra fast dectection circuit. And a solenoid switch motor is, let's face it, a relay. My money is on the PSX-AR activating again, but now we're back to the original condition of a short between the outer loop rail and the turnout. Or, if the turnout still completes its throw, still a polarity mismatch on the inner rail (which has flipped back to red) and the diverging frog rails (which are now blue).

This is not an issue for a manually thrown turnout ("manually" meaning push buttons and a powered switch machine or a true manual linkage like a Blue Point), only if using the auto-throw feature AND with power routing turnouts. The extra sections would be unnecessary with an Atlas turnout because the polarity of the exit rails is always the same regardless of the position of the points.

                      --Randy

 

 

 

 

Dunno. Randy, I remain conflicted. 

 

I understand what you are saying, but as I read what you wrote, I keep asking myself, then why doesn't DCC Specialties say so in their literature. There is no mention of a second set of gaps and an additional short track section in the manual. Although the PSX-AR manual does not include any drawings of the PSX-ARSC being used to throw the points, it does include as a feature "Automates Reverse Loop Turnouts". So, you would think that they would at least mention the issue that we are dealing with here. 

Now, you might ask yourself, then why, Rich, did you draw up the diagram with additional track sections on either end of the loop?  My answer would be that I drew the diagram as a testing technique to solve Dennis' problem. It did not surprise me that it worked because it should work. But, I still cannot help wondering if those two additional track sections are really necessary.

I can understand Dennis' reluctance to now remove the two sections, or at least remove the plastic rail gaps between the two track sections and the turnout, but my investigative mind would not be able to resist such a further test.  I do not use Electrofrogs or the Peco switch motors (I use Insulfrogs and Tortoises) or else I might try it on the workbench.

Mind you, I'm not saying that it will work without those two additional track sections, but I think that it is at least worth a try.  I sure would be interested in what Larry Meier would have to say about this.  Without those two additional track sections, the reverse loop and the turnout are still fully isolated from each other, so it is all a matter of timing.

Rich

 

Ok, given the amount of time you guys have put into this matter I have done the test. Furthermore I have made 3 videos which I may upload to YouTube tomorrow after I have reviewed them. The first video is of the PSX working ok. I then soldered the gaps and disconnected the track power from the two new sections, which should have resulted in one single reverse loop, which could then be tested. However on powering back up I discovered that the reverse loop had NO POWER at all and the Loco just stopped after the first gap and that was that. So that video was a bit of an anticlimax. At that point I was not prepared to try and work out why there was no power. I have spent hours and hours and hours this week looking for such wiring problems/errors. So I then cut new gaps much closer to the turnouts (as I said I would), powered back up and ran a 3rd video of the Loco performing 100% ok.
Larry Meier has not replied to my last two emails. I am not expecting to hear from him. Go back and read his initial response on 14th December 2016 that I posted, it was full of technical confusion. He told me to “align the switch clear (using the PSX-AR switch address 2044),” which really threw me in a pointless direction. His conclusion was that “I suspect that you have a wiring error.” Tell me something I did not know but also tell me where in the PSX-AR manual it tells you how to wire a snap coil turnout motor and what to do if it does not work.
As yet no one has joined our thread to say that they use Peco switch motors with N scale electrofrog turnouts and that they have had no problems. I find it hard to believe but maybe no one else has.
What I would like to hear from Larry Meier is that he has tested such a combination and proved that the PSX-ARSC works straight out of the box as it supposed to. 

 

Dennis
  • Member since
    May 2015
  • 37 posts
Posted by Denver on Sunday, December 18, 2016 7:44 AM

I have just realised that the reason I had no power was because the polarity of the two extra sections was not the same as the reversing loop. In which case I will not be testing the wiring without the two new sections. I could be back where I started from a week ago! I will also not publish the videos. 

Dennis

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Sunday, December 18, 2016 7:53 AM

LOL

Hey, at least you tried.   Yes

I still want to hear from Larry Meier who has the answer.

Rich

Alton Junction

  • Member since
    September 2004
  • From: Dearborn Station
  • 24,281 posts
Posted by richhotrain on Sunday, December 18, 2016 8:06 AM

Denver

Larry Meier has not replied to my last two emails. I am not expecting to hear from him. Go back and read his initial response on 14th December 2016 that I posted, it was full of technical confusion. He told me to “align the switch clear (using the PSX-AR switch address 2044),” which really threw me in a pointless direction. His conclusion was that “I suspect that you have a wiring error.” Tell me something I did not know but also tell me where in the PSX-AR manual it tells you how to wire a snap coil turnout motor and what to do if it does not work.

As yet no one has joined our thread to say that they use Peco switch motors with N scale electrofrog turnouts and that they have had no problems. I find it hard to believe but maybe no one else has.
 
What I would like to hear from Larry Meier is that he has tested such a combination and proved that the PSX-ARSC works straight out of the box as it supposed to. 

 

Dennis
 

Dennis, I agree with you that the use of a Peco Electrofrog turnout to control the entry/exit to a reverse loop, combined with the automatic reversing of the point rails, is not addressed in the PSX-ARSC manual. I don't see where N scale is part of the problem though. I would bet that HO scale will act the same way as N scale.

Not to defend Larry Meier, but the reason that he mentioned the 2044 accessory address was merely to point out that there is the capability to set the position of the point rails at power up. He also mentioned that the preferred wiring protocol is to match the polarity of the reverse loop with the straight through route of the turnout. Of course, it could be done the opposite way to match the divergent route of the turnout, but Larry was merely trying to simplify and standardize the entire process.

To your credit, you not only stuck with it instead of giving up as some guys do, and your detailed description of the problem was immensely helpful in solving the problem. In a lot of threads, we can feel like we are pulling teeth to get a full statement of the facts and circumstances.

If you don't hear from Larry, I will contact him because we deserve a fuller explanation from the guy who designed the various PSX units.

Rich

Alton Junction

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Sunday, December 18, 2016 4:06 PM

N scale and HO scale Electrofrogs work the same.

I suspect only OONE extra section is needed, not both, because all polarities will already match on the entrance to the loop (left over from the previous passage).

Wait - I take that back - think it through - if the second train now goes int he loop via the diverging leg of the turnout (since the last one came out there), everything will match - frog rails, the short section, and the loop. Now as it comes around the bottom - it will have the exact same problem as when it ran the other way, if the extra section is not there.

I don't know why this is not mentioned in the PSX-ARSC documentation. As I said, it's not an issue with any sort of turnout if not using the PSX to actually operate the switch motor - either you wiull have manually made the change proper to the gaps being bridged, or you will get an uncorrectable short and the PSX-AR will cut all power and the train will stop instead of derailing by running through a turnout set against it. It's ONLY a problem withint the auto-throw AND power routing turnouts. It MAY be ok with an unmodified turnout, since as soon as closed point rail moves away from the stock rail, the entire frog assembly becomes unpowered. But this makes for a rather unreliable turnout. I guess they don't expect anyone to make a trivial reverse loop where you enter one side of the turnout and the track immediately loops back on the other leg of the turnout without any other intervening trackage.

 Again, it's really the automatic part that cuases the issue, otherwise it's not a problem. Or if automates through other means, say an IR detector that trips when the train hits a certain point - if the switch is thrown BEFORE the gap is bridged to trip the autoreverser, it would work just fine. That mimics the action of manually operating the turnout while using the autoreverser to just handle the track polarity.

                                    --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!