Guy,
Once it's set up, the PSX-AR just does its thing reliably. Getting back to your original question, a single PSX-AR will handle this job.
Although I don't have Loconet, since I set up my CCTV camera so that it can see the PSX-AR's LEDs at what would be Switch A in the diagram, I can even monitor it, mainly to confirm I've managed to derail something...which is fairly rare, considering it's HOn3 that's down a helix underneath the main part of the layout.
Mike Lehman
Urbana, IL
Rich,
In fact it is not a ''loop with the siding inside of it''. It is a double track siding located at the North end of a End to End layout that I use as a reversing loop. Either track can be used as the reversing loop.
As it is a End to End layout, no train is allowed to go thrue the loop without stopping. I only use the reversing loop to reverse the trains between operating sessions. So the need of the PSX-AR is not mandatory. As noted by Randy, the use of the turtoise to change polarity is all I really need as the trains are still when I change polarity. The PSX-AR would just be a more convenient way to manage the traffic through the loop, I think.
Guy
Modeling CNR in the 50's
I'm not disparging it, I'm just saying I can do darn near the same thing for a $1 relay and NEVER get a short instead of using a $70 device that waits until there is a short and fixes it. And I really wish they would quit claiming they have feedback, they do NOT, you need some sort of input device for your DCC system to get feedback - none of the PSX devices have Loconet or ExpressNet or NCE Cab Bus.
I know all about Tony's products and how they are SO superior to the Digitrax relay devices (gets posted here ALL THE TIME) yet my PM42 breaker works JUST FINE with sound locos - including articifically loading up one section with ALL of my sound locos at the same time and deliberately shorting it. And it connects to Loconet so I can control each section manually and/or view the status even if the board is up under the layout and I can't see the LEDs.
On a complex piece of trackage, where trying to figure out all the possible combinations of some puzzle trackwork like a multiple double slips - forget about it, use the electronic device. For a simple loop or wye, totally not needed.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
rrinkerWorks fine if you always want to make each loop trip run the opposite of the last one. If that fits your operating plan, fine.
Also not so. You can still align the turnout via either a pushbutton or DCC command to traverse the loop in whatever direction you choose.
In case you'd like to actually learn about the product before you continue to disparage it, here's a link:
http://www.dccspecialties.com/products/pdf/man_psxar.pdf
I have the right to remain silent. By posting here I have given up that right and accept that anything I say can and will be used as evidence to critique me.
All of which makes me wonder how the OP plans to use the loop with the siding inside of it.
Since it is a single track coming into the loop, an exiting train is obviously headed in the other direction. What does the rest of the layout look like?
Will an entering train move into the siding so that another train may pass through, perhaps from the other side of the controlling turnout?
Inquiring minds want to know.
Rich
Alton Junction
Works fine if you always want to make each loop trip run the opposite of the last one. If that fits your operating plan, fine.
rrinker You have to throw the switch ANYWAY to get the train out of the loop - otherwise you will derail on the misaligned points! There's no more manual operation involved than there is if you hook up an autoreverser.
Not so. My PSX-AR, when it detects the short, reverses the loop "polarity" AND aligns the switch to the track the train is coming from.
You have to throw the switch ANYWAY to get the train out of the loop - otherwise you will derail on the misaligned points! There's no more manual operation involved than there is if you hook up an autoreverser.
As for doprouts - no, that's another reason to have the contacts operate a relay. As soon as the Tortoise contacts leave the section of the circuit board inside that is connected to the terminal energizing the relay, it will drop out and chance the phase. Even before the contact crosses the dead spot and touches the other side. And when throwing the switch the other way, the relay will not flip until the contacts touch inside the Tortoise, after they cross the dead spot. In either case, instant contact change from one state to another, with no delays.
As for goofing up and forgetting - yes, it will short, and you'll also likely be on the ground. That's why you have power districts, and if you aren't making your wiring robust enough for the quarter test to work, an autoreverser isn't going to work reliably either.
This is pretty simple stuff, no more complex than the fancy electronic solution. Now, if you want a complex solution to a simple problem - that would be like wiring the layout for MZL with DC, so you can just drive the train and not the track. All thouse relays and contacts just so you cna concetrate on the throttle only - connecting a DCC system, even with feeders every 3 feet, is much simpler. The advantage of the relay, both for loops and frogs (or just the contacts, in the case of frogs - it's not like 10 locos can all be sitting on one frog, creating a need for high current contacts) is that in both cases you eliminate the short, not just correct it. An autoreverser waits for the short, then tries to fix it by switching phase. Hopefully it does, otherwise, time for the circuit breaker to work. And so do the Frog Juicers - they wait until a wheel touches the frog, it if's a short, it flips the frog phase in hopes of correcting it. The reason the all-electronic ones seem 'better' is because they are correcting a fault, which has to happen as rapidly as possible to avoid an interruption. 1ms for a relay vs .01ms for the electronic one may make an actual difference here. But if you are never causing the short anyway, 1ms for the contacts to move and re-engage is plenty fast enough not to get any noticeable drops.
carl425 SNIP Yeah but... The whole premise of DCC is to automate all this stuff and relieve us from all the DIY tedium. Your scenario of using the tortoise contacts to reverse the loop still leaves you with the job of having to manually operate the switch. SNIP Just because it's possible to DIY it doesn't mean that you should. How much other progress on the layout is not happening while you design, shop, assemble, test and debug these little inventions? It's all about the value of your time and the aspects of the hobby you find interesting. Do you want to be a model railroader or an electronics tinkerer? SNIP
SNIP
Yeah but...
The whole premise of DCC is to automate all this stuff and relieve us from all the DIY tedium. Your scenario of using the tortoise contacts to reverse the loop still leaves you with the job of having to manually operate the switch. SNIP
Just because it's possible to DIY it doesn't mean that you should. How much other progress on the layout is not happening while you design, shop, assemble, test and debug these little inventions? It's all about the value of your time and the aspects of the hobby you find interesting. Do you want to be a model railroader or an electronics tinkerer? SNIP
Just for the record, my Chama staging loop still uses the manual turnout switches. I know I could use the PSX-AR to automate the turnouts, but I prefer manual control. I can monitor everything via CCTV, so I can throw them manually just like I do elsewhere on visible track.
Ultimately, for control systems for hidden track, you should choose the one that works best for your particular application. There is no single solution that fits all circumstances, so knowing your options in a case like this is always good.
rrinkerI would (and plan to) use a relay in all cases, just to make the switching faster and avoid any dropouts. $3.38 on ebay gets me TWO DPDT relays with 12V coils that draw less than 40ma, and have contacts rated at 10A ($1.38 for the relays and $2 shipping). So less than $2 per reverse loop. And a PSX-AR is how much?
The whole premise of DCC is to automate all this stuff and relieve us from all the DIY tedium - "operate the trains, not the track". Your scenario of using the tortoise contacts to reverse the loop still leaves you with the job of having to manually operate the switch. If you get distracted, you get a real short that takes out a hunk of the layout. My reverse loops, being quite unprototypical for a mainline, are hidden. When I send a train into the loop, I just want it to turn around and come back. I'll gladly pay the $54 for the PSX-AR for that convenience.
Just because it's possible to DIY it doesn't mean that you should. How much other progress on the layout is not happening while you design, shop, assemble, test and debug these little inventions? It's all about the value of your time and the aspects of the hobby you find interesting. Do you want to be a model railroader or an electronics tinkerer? Either choice is fine, but I'd rather spend the cash and have the time to do something I find more interesting.
Besides, having something that has been a pain to deal with since the invention of model railroads work without intervention is really cool!
Thank all.
This was very informativ.
PSX-ARs were ~$40ish when I bought mine but it's been a few years.
rrinker Yes, I'm cheap. But also very DIY.
Me, too. But electricity and plumbing are instances when I often find the technical solution saves enough time over the DIYing to make buying out the frustration factor and the cost of my time level the playing field. In this case, the numbers may not quite support that conclusion, but close enough for me.
Also, what happens if you accidentally cross the gaps with contact-driven reversing? I suppose whatever circuitbreaker it's on shuts it down? But could also sit and fry a bit if the quarter test wasn't applied to verify wiring suitability. Same thing applies to the auto-reverser of course, but once again it depends on your comfort level with the particular solution.
I do wonder about the contact driven reversing not suffering from the small dead spot in the middle of the two sets of contacts on the Tortoise. Doesn't that carry over to the relay, too? I've never noticed any such stutter from my PSX-AR sections, so to me it seems sorta a win for the PSX-AR on that, although no direct experience with the contact-driven method to hear what happens with sound locos passing through it.
I would (and plan to) use a relay in all cases, just to make the switching faster and avoid any dropouts. $3.38 on ebay gets me TWO DPDT relays with 12V coils that draw less than 40ma, and have contacts rated at 10A ($1.38 for the relays and $2 shipping). So less than $2 per reverse loop. And a PSX-AR is how much?
There may be complex situations where it's easier to just let some fancy device figure it out, but I prefer to avoid the short, not correct it. Same with frogs. EASIEST way to power frogs is with something like the Tam Valley Frog Juicer. DCC in, and hook a wire to the frog. It figures everything else out. But at $80 for 6 frogs. The relay they sell for the various servo controllers is $32 for 8. $4 per frog vs $13+. OK, I have to figure out the proper polarity for the bus wires to connect to the relay. It's a 50/50 shot. and they do use screw terminals, so if it's bckwards the first time, it only takes a few seconds to swap. Depending on how I mount the servos, it may be even cheaper to just use a microswitch, but $4 per is pretty cheap.
Yes, I'm cheap. But also very DIY.
rrinkerThe loop, especially a simple one, can only hold one train, maybe two.
Depends on how long the loop is. In my case, it's long enough and the trains are short enough that I can stack up to six trains, three on each track. But this is a case where if you are switching power with the Tortoise you probably want a relay to handle the actual switching to avoid hi currents across the Tortoise contacts. Many of my narrowgauge locos have sound, so there could be more draw than just the loco crossing the gap imposes. I think this is also where using the PSX-AR may be just as easy and almost as cheap as trying to use the Tortoise contacts.
Yup, absolutely nothing. You can flip the polarity back and forth under the running train all you want and nothing happens. Go ahead and try it - hook a piece of flex track to your DCC system with clip leads. Start the train going forward. Unhook the wires (don;t touch the throttle or anything). Clip them back on the opposite way. As soon as power is applied, the train will start moving again - in the same physical direction as before. If it was moving left to right, it will still move left to right.
This is one of the keys to understanding DCC. It is the commands sent to the decoder that tell it what to do, not the polarity of the track. In DC, direction is controlled by the track polarity - which is why if you take two F units and put them on DC track back to back, they work together and move the same direction - one running 'backwards' as in, away from the cab. However, if you put two (independent, let's ignore consisting and so forth) F units on DCC track, back to back, and tell them both to fo in the forward direction, they will move apart, away from each other, becuse 'forward' is the direction of the cab. Unless wired incorrectly (there is an NMRA standard for DC loco wiring), you can;t have a cornfield meet with DC locos, they'll never run towards each other (in the same block). But since the polarity of the track power means nothing in DCC, it's quite easy to play Gomez Addams and smash trains together if you don't pay attention.
It greatly simplieifres reverse loops, since you can change the polarity under the moving train. No more synchronizing the cab and mainline reversing switches to keep the train moving smoothly. Once the entire train, or at least all power pickups (locos, lighted cars) and metal wheels are past the gaps, you can flip the polarity under the running train without a blink and run it out the exit of the loop with no hesitation (obviously you need to line the switch).
Technically, it's not 'polarity' but rather phase with DCC, since what it is is a square wave AC signal. If you make a reverse loop without gaps, you'll have a piece of track where the wave is at the most positive side touching a rail where it is at the most negative at the same time - this is a short. Put in the gaps, and you can flip this square wave upside down right under the decoder - it is not the positive and negative parts of the wave that tell the loco what to do, but rather how long it stays either positive or negative. Short sections of positive or negative are the 1 bits, long sections are the 0 bits, and those 1's and 0's make up DCC packets which include address information and command information being sent to specific decoder addresses. Barring some special cases, the signal is positive just as much as it is negative, so the positive half and the negative half contain the exact same information - which is why it can be flipped over with impunity, the data remains the same.
The reverser is making changes that a DC person would call reversing polarity, but the decoder doesn't see it that way. In fact, the decoder wouldn't care if the reverser kept changing polarity every second under the locomotive, provided the right reversal takes place while a metal tire is bridging a gap. That bridging is when the short takes place due to the conflict in phase between two meeting rails.
The reverser senses the conflict. The conflict might be as the train enters the loop, but it might be when it leaves the loop...depends on what the reverser did to the rails last time it switched them. It doesn't switch unless there is a conflict, and that takes place wherever the conflict is. So, if a train entering the loop does not cause a short/conflict, the PSX senses no fault. The train trundles around the loop, but momentarily causes the short at the far end gaps, whereupon the PSX reverses the rails.
Later, a train comes along and enters the loop in the reverse direction as the previous train. The PSX corrected the conflict at that time between those rails, so there will be no conflict in the reverse direction for this second train. However, there must be the inevitable conflict when the new train crosses the far side gaps because there is where a conflict must be, logically.
gmpullman So what happens to the train as this polarity changes on the rails under the engines' wheels? That's the part I'm a little corn fused about... Maybe rail A and B can be quickly swapped and the decoder won't know the difference?
So what happens to the train as this polarity changes on the rails under the engines' wheels?
That's the part I'm a little corn fused about...
Maybe rail A and B can be quickly swapped and the decoder won't know the difference?
Well, yes, I understand the DPDT contacts on the Tortoise. I was wondering what Randy was refering to when he first mentioned using them. I mean, you have the Tortoise that operates the points on the main loop switch.
So, you enter the loop depending on that switch set for, lets say the diverging left hand route. Rail A is happy with + and + on both sides of the gap and rail B is happy with - and - on both sides of the gap. The train continues forward.
The end of the train clears the switch and you have to throw it to the straight route. As the points move the Tortoise contacts change rail A to - and rail B to +
The polarity you really want to change is on the remainder of the layout AS the train is within the loop. At least that's the way it used to be done back in DC days.
I'm just askin'
One of my reversing sections is part of an engine treminal and I have a dozen or so engines sitting there. When I move through there the PSX-AR switches and I see a quick blink of the engine lights but I've never had any of them moving at the time.
Just wondering? Ed
Well, you need both to actually change polarity - but one set is enough to trigger a relay to do the actual reversing. The Tortoise contacts tend to not like switching high currents, plus they can make conact with the opposite side before they have physically move the points off the prior stock rail. And there is a bit of a dead spot in the middle, which can cause a stutter, or a sound dropout.
gmpullman There's something I'm missing about using the Tortoise contacts to reverse the loop polarity. Of course, as the train enters either leg of the loop rail A must be ++ and rail B must be -- now once the train is fully in the loop, you are going to throw the switch and the Tortoise is going to change the polarity on what? The loop is occupied by the train. You don't want to change that polarity.
There's something I'm missing about using the Tortoise contacts to reverse the loop polarity.
Of course, as the train enters either leg of the loop rail A must be ++ and rail B must be -- now once the train is fully in the loop, you are going to throw the switch and the Tortoise is going to change the polarity on what? The loop is occupied by the train. You don't want to change that polarity.
Yes, you change the loop polarity. What do you think the PSX-AR does? With DCC, you change the polarity under the running train, NOT the main line like you do with DC. Polarity of the DCC signal has no bearing on the direction of the train. Forward is forward, reverse is reverse - relative only to the loco, not to the track power. This is also why autoreversers don't work for DC - they change the loop polarity, so the train would approach the mismatched gaps, short - the reverse would trip, and the train would instantly move the opposite direction.
You CAN still do it the old way, and reverse the rest of the layout, but it's pretty pointless - plus you need to switch potentially a lot more current if you reverse the entire rest of the layout - on a big layout you could have dozens of trains running. The loop, especially a simple one, can only hold one train, maybe two.
Of course, as the train enters either leg of the loop rail A must be ++ and rail B must be -- now once the train is fully in the loop, you are going to throw the switch and the Tortoise is going to change the polarity on what? The loop is occupied by the train. You don't want to change that polarity. You want to change the polarity of the remainder of the layout since you just made a westbound train out of an eastbound train and it is still moving forward.
From what I see the train will continue around the loop but when it encounters the main line rails you will have the short. Or am I missing something?
The old Atlas controller had an east-west switch. The trick was that the polarity DIDN'T change while the train was in the reversing section but you changed the polarity of the rest of the layout (or the cab that your train was assigned to) in old DC days.
I have three reversing sections on my layout. 2 use the PSX-ARs and one has a Digitrax AR-1 and they all work "seamlessly" in the background without even thinking about them.
Just curious, Ed
Another alternative is to use a PSX-AR and forget about the Tortoises on the two turnouts inside the loop by just using Peco Electrofrog turnouts. You can throw the points with your finger and keep the inner loop unpowered when the points are set for straight through.
If the Tortoise for A is already controlling the loop polarity, you don;t even need the PSX-AR. In fact this is the PREFERRED way to do it - rather than have a short automatically corrected by the device, prevent the short in the first place. To avoid dropouts of sound, and also to keep the Tortoise from handling too much current, have the Toirtoise contacts drive a relay.
Even the gaps at the bottom can stay, and Tortoise B would control which track was live, same as it does now. But all power feeding this must come fromt he relay on Tortoise A.
Yes, this is a contrarian view, but I see no reason why it wouldn't work. And save the cost of a PSX-AR. This is a 'simple' balloon loop, with the extra siding inside it, which really doesn;t change anything. For such loops, you can avoid the short altogether by controlling the loop track polarity through the switch motor.
In your diagram, I can see that you're set-up is nearly identical to mine, down to the need to convert it from DC. I'll explain a little bit more as to why it makes it easier to operate for staging by making it one way. As far as I can see, other than perhaps a minor relocation of gaps, all the wiring can stay the same.
First of all, my reversing section is ONLY the track with Tortoise A and the arm that reaches towards and through Tortoise B (including both power-routed extensions leading to it), powered coincidentally enough thru a PSX-AR. I'll call the first, "loop" turnout past A as Tortoise C and it is also part of the reversing section.
Tortoise B is also in the reversing section but the gaps for it are however far you make them into the loop with the power routing feature of the contacts.Your diagram shows them relatively short and close to B, but on my loop they're actually about half way around the loop so much longer than in your diagram.
Thus, I used the Tortoise's contacts at Point B end only to switch power on the two tracks that lead into it. That creates the "capture and emitter" half of the loop.
I gapped the loop internally about half way around it. You have probably already done that with your DC wiring in some fashion so that if the Turnouts C and B are set to different tracks it won't short. But gaps are easy enough to move if they need to be adjusted. The main issue is to make the distance between wherever you locate the gaps BEFORE the train gets to B and BEFORE the train enters/exits the loop at A longer than your longest train.
The contacts on Tortoise C are used just far enough to get past the frogs. Presuming that you already have leads to the two tracks leading out of C, simply cutting a set of gaps right past it (if not already present) and hooking up the remainder of the first half of the loop to full time power now that you're on DCC will set it up as the "receiving" end of the loop.
Alternatively, the gaps on the entry end at C could come before that turnout.
It's this paired concept of "receiving and transmitting" that suggests why it's easier to operate it as a one way loop. If you're inattentive to train position once the train has entered either loop track, it either runs into a dead track and stops at the reversing section track gaps, which are dead if B is thrown against it OR B is set correctly and the train is able to move foreward into the reversing section. Thus the power routing contacts at B provide power to the tracks leading into it only when it is set to the track you want powered.
There is one way this setup can be fooled. In coming around the loop track, you don't want to cross the gaps that come before B on each track and leave the loco half in the first part of the loop coming out of C and half in the reversing loop. That or leaving a metal wheel bridging those gaps will cause a short because it is possible then for those rails to be different polarities. Once one is on the habit of moving trains through the loop, this is very unlikely to happen so long as you turn the throttles to zero when parking a train on the loop.
EDIT: One other potential operator error situation would be if you start a train through B and have A set for a train to enter rather than exit the loop. Generally, if your swicthes are mounted next to each other that's avoidable and obvious enough if you do create a track fault.
In my last reply, I indicated that I will probably bridge the gaps North of Switch B then feeding both tracks simultaneously. Beside that, I like to use the blocks inherited from my previous DC wiring. Almost all the insulated blocks are automatically controlled by a Turtoise switch machine, some with toggle switch. The locomotives I don't intend to use during a session can then sit on an unpowered section of track. In fact It saved me a lot of reprogramming when I inadvertently made some programming on "programming track mode" on the main.
If I had wired the layout for DCC in the first place I would have probably made different choices but I am happy with what I get as I get the best of both world.
I don't see why you only want power to one track at a time.
But that aside, you only need one PSX-AR and only two sets of gaps, one set at each end of the turnout that forms the loop.
Everything inside those two sets of gaps is wired to the output side of the PSX-AR.
I understand that:
- I leave intact the gaps East and North of A;
- I bridge the two gaps North of B;
- I wire the PSX-AR on both rails immediately South of B;
- I discard any other wiring.
Thank you everyone.
The two loops are concentric, and don't cause a conflict in phase/polarity where they meet their access points; the two turnouts. You could have this arrangement at each end of a loop of track and you wouldn't need a reverser. So, when already inside the potential conflict zone, the one turnout that forms your reversing loop, there is also no conflict for the two sidings. Therefore, as Carl says above, you only need to worry about the conflict, and that takes place 'west' of the two tracks and their turnouts, where the loop pinches. That is where you need your gaps. The single reverser will reverse all the tracks, including the siding, when it detects a fault.