#### Minimum length on Non-reversing segment

mgj21932

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

Blair

BillD
Think of any single loco as a pair of wires hovering, one over each rail, drooping to touch the rail at each point where a loco's wheels do.  In order for a short length between AR sections to do its job, it must be long enough to prevent the wire from bridging the two AR sections.  So before the wire enters the second AR it must leave the first AR.

Now generalize this.  The user must be aware of the longest electrically connected "thing" he runs.  For example if you run a dummy loco but wire its wheelsets to an adjacent engine, then that becomes the longest "loco".
That defines the shortest non reversing section between AR sections.
Hope this helps.
Blair

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932@...>
Date: 2021-05-22 22:30 (GMT-05:00)
To: w4dccqa@groups.io
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

Blair

BillD
When incorporating sections of your loop, you must also ensure that those sections are longer than your longest train.  Otherwise the problem just moves.
Think you may have done so, but just clarifying.

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932@...>
Date: 2021-05-22 22:30 (GMT-05:00)
To: w4dccqa@groups.io
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

D B

Blair,

Thanks for the confirmation. And your analogy caused the light bulb to go off.

I had assumed that the loco‘s body/frame meant that the front and rear wheels were electrically connected (as you aptly describe), but I was not sure why that would matter.

I was thinking that the AR-wired segment from which the train had departed (AR-wired segment 1) would have electrically “aligned“ with/matched “polarity” with that of the non-reversing segment as the loco crossed the gap and the AR circuit did its magic. Then the same thing would happen as the front wheels of the loco passed over the gap to the next AR-switched segment (AR-switched segment 2). I assumed that would simply cause AR-switched segment 2 to align with the non-switched segment and, by extension, with AR-switched segment 1 (which was already aligned with the non-switched segment).

What I didn’t consider, however, was the potential for the electrical continuity of the loco (and, as you point out, any electrically connected cars, such as a wired tender) to transmit the electronic message of a “short” experienced by the loco’s front wheels crossing the second gap (between the non-switched segment and AR-switched segment 2) to the AR circuit controlling (protecting) AR-switched segment 1. That AR circuit would either not know what to make of the confusing message or, more likely, would try to switch the “polarity“/alignment of AR-switched segment 1 “back” (after all, that’s what AR circuits DO). We’re talking very short time intervals here. Such that the switching action of the first (rear-most) AR circuit would have “rebounded” back to the second (forward-most) AR circuit, and a “war” between AR circuits initiated.

In effect, by spanning the length of the non-switched segment, the electrically connected elements of the loco cause the two AR-wired segments to no longer be electrically separated by a non-switched segment!

Ergo, the minimum length requirement for the non-switched segment.

Now that I’ve “got it“ (thanks Blair), I hope others do too, as I don’t believe the rationale for the rule (as opposed to the rule itself) had been made clear by the previous discussion on this forum.

Bill D
N&W Steam Only

D B

Blair,

I thought I had taken your advice about the taking into account the lengths of the loop sections, but as I considered your second message, another issue arose.

Does it matter that I have turn-outs in my AR-wired segments?  Because the turn-outs, and the sidings to which they are connected, are all wired to the same AR circuit, or are electrically insulated from any adjacent non-switching segment, I assume the required minimum length of the AR-wired segment can include the turn outs and track segments in between.

Is that right?

Bill D
N&W Steam Only

On May 23, 2021, at 7:29 AM, Blair <smithbr@...> wrote:

BillD
When incorporating sections of your loop, you must also ensure that those sections are longer than your longest train.  Otherwise the problem just moves.
Think you may have done so, but just clarifying.

Sent from his Galaxy

Blair

Bill
It gets tricky to answer this with confidence without a sketch to refer to; in addition, I'm working from my phone and don't have my groups login, so even if you post a sketch, i won't be viewing it or responding until tonight at the earliest.  Sorry!
Blair

Sent from his Galaxy

-------- Original message --------
From: D B <1932mgj2@...>
Date: 2021-05-23 09:05 (GMT-05:00)
To: w4dccqa@groups.io
Subject: Re: [w4dccqa] Minimum length on Non-reversing segment

Blair,

I thought I had taken your advice about the taking into account the lengths of the loop sections, but as I considered your second message, another issue arose.

Does it matter that I have turn-outs in my AR-wired segments?  Because the turn-outs, and the sidings to which they are connected, are all wired to the same AR circuit, or are electrically insulated from any adjacent non-switching segment, I assume the required minimum length of the AR-wired segment can include the turn outs and track segments in between.

Is that right?

Bill D
N&W Steam Only

On May 23, 2021, at 7:29 AM, Blair <smithbr@...> wrote:

BillD
When incorporating sections of your loop, you must also ensure that those sections are longer than your longest train.  Otherwise the problem just moves.
Think you may have done so, but just clarifying.

Sent from his Galaxy

Don Vollrath

A short section of fixed polarity track between two AR sections needs to be only as long as it takes for the leaving or entering AR controllers to recognize that a polarity correction is needed, and do so as an offending truck with multiple connected axles and wheels rolls through. So making the fixed polarity section only as long as a multi axle passenger or diesel truck should be sufficient.
Note that once the “leaving” AR section has flipped to agree, it is Essentially out of the picture as that same truck crosses from the fixed polarity section and begins to enter into the next AR section, and causes it to correct that polarity if necessary.

DonV

Don Vollrath

Oh, I forgot Bil. With steel locos the answer is contingent on how the power pickup is arranged on the loco and tender. One must ensure that an AR corrective “short” occurs on the leaving section before those axles and wheels begin to roll into the next AR section. To cover all situations I would make the fixed polarity track section at least as long as the connected steam drivers of the loco.
A corrective short needs to occur only on one track With good AR controllers.
Operating speed of your trains may require a longer section of fixed polarity.

DonV

Jim Betz

Bill D,

Here is something to think about.  The rule that a reversing section must be
longer than your longest train isn't just a good idea - it's a "rule".  There is a
caveat to that rule that goes "... there are times when you might run a train
longer than your reversing section and it will work and you will be tempted
to ask "what's the big deal?".
If the train bridges both ends of the reversing section - at the same time - it
will create an electrical problem that your auto reversers do not know how to
deal with - it is possible that the train will "stutter its way past the problem"
but that looks really ugly.  It is best to avoid it.
IF your layout is so small that your auto reversing section doesn't work
right - then the correct/first response is to cut down the length of your longest
train(s).  In other words - avoid the problem.  Yes, I know, I want to run
long trains also ... but not so long that they are longer than my reversing
section, staging (storage) tracks, or my shortest passing siding.  Well, OK
I agree that doing a saw by is fun and you can avoid it by not having two
trains that need to pass arriving at that siding at the same time.  But even
the real RRs considered doing saw bys a problem and avoided them if
possible.  So should we.
The physical layout (such as the length of storage tracks, passing
sidings, and reversing sections) is best treated as a "given" and then

Here's a last consideration - if you ever intend to have more than one
operator on the layout at the same time ... then if you assume that
person is not super smart and can not learn complicated stuff EASILY
(such as saw bys) then the layout will run ... wait for it ... better for YOU.
For example, let's just think about having the other operator being your
6 year old grand.
- Jim
- Jim

whmvd

Blair,

What you describe is safe. No question. But it will also, usually, mean that you make a longer intermediate section than necessary. The 'bridging' you describe is indeed the tricky bit, but: only when the AR action HAS NOT YET TAKEN PLACE.

Consider a loco that pushes a lighted long wheelbase 4-axle wagon through the reversing sections, and all axles of the wagon have power pick-ups. Also assume that the polarity on entering is the worst it can be.

A = fixed polarity section
B = AR section 1
C = fixed polarity intermediate section
D = AR section 2
E = fixed polarity section
| = isolators
L = loco
W = wagon, with P being power pickup

Initially B has opposite polarity/phase to A, D and E (which are fixed), but the same as C.

1. Pushing loco moving to the right.
LLLLL-WPPWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

2. A<->B boundary bridged, AR B switches, so A, B, D and E now have the same polarity/phase.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

3. Wagon spans the A<->B boundary. No AR action. All further bridging of A<->B now irrelevant.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

4. B<->C bridged, AR B switches.
LLLLL-WPPWWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

5. C<->D bridged, AR D switches, so B, C, D and E now have the same polarity.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

6. C<->D still bridged, No AR action.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

7. Nothing bridged, wagon spans entire section C, but, crucially: there is no problem, because there is no short!.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

and so on - the rest is trivial..

And this is why only the minimum size of a TRUCK counts. A longer size becomes necessary only when a truck is not really a truck but a single axle. In other words: the long wagon only has one power pickup at each end (never mind the number of wheels, the number of pickups is what counts). Theoretically, it should STILL work. But in practice, a short on a one-axle truck can be very, very easily missed as it would be just the wheelset closing the gap  in the rails on its own. See what happens in the figures above when the short in step 5 is not seen. After that, spanning the C section (fig. 6) DOES mean trouble, because the B<->C short has not occurred and the reversers now fight - exactly what should be prevented from happening.

One solution (yours) is to make the C section long enough to span the entire wagon, which is fine and works. However, if that is not convenient (and the layout might well dictate that it isn't) then the other possibility is to stagger the cuts in the rails. That increases the chances of the one-wheel short occurring by a lot. This is the only advantage to staggering that I have ever been able to see. But I repeat: there is ONLY a problem in case of one-axle trucks (if you can even call them trucks).

Wouter

On Sun, 23 May 2021 at 12:27, Blair <smithbr@...> wrote:
BillD
Think of any single loco as a pair of wires hovering, one over each rail, drooping to touch the rail at each point where a loco's wheels do.  In order for a short length between AR sections to do its job, it must be long enough to prevent the wire from bridging the two AR sections.  So before the wire enters the second AR it must leave the first AR.

Now generalize this.  The user must be aware of the longest electrically connected "thing" he runs.  For example if you run a dummy loco but wire its wheelsets to an adjacent engine, then that becomes the longest "loco".
That defines the shortest non reversing section between AR sections.
Hope this helps.
Blair

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932=yahoo.com@groups.io>
Date: 2021-05-22 22:30 (GMT-05:00)
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

D B

Wouter,

As usual, most of the detail in what you said went right over my head.   But the credible explanation from “the expert” confirming Blair’s recommendation is reassuring.

In my case the fixed segments are several feet long, but not the length of my longest train; however the AR-switched segments are comfortably longer than the longest train.  Now I have a bunch of re-wiring to do to implement the change.

Thanks for the insight.

Bill D
N&W Steam Only

On May 23, 2021, at 2:20 PM, whmvd <vandoornw@...> wrote:

Blair,

What you describe is safe. No question. But it will also, usually, mean that you make a longer intermediate section than necessary. The 'bridging' you describe is indeed the tricky bit, but: only when the AR action HAS NOT YET TAKEN PLACE.

Consider a loco that pushes a lighted long wheelbase 4-axle wagon through the reversing sections, and all axles of the wagon have power pick-ups. Also assume that the polarity on entering is the worst it can be.

A = fixed polarity section
B = AR section 1
C = fixed polarity intermediate section
D = AR section 2
E = fixed polarity section
| = isolators
L = loco
W = wagon, with P being power pickup

Initially B has opposite polarity/phase to A, D and E (which are fixed), but the same as C.

1. Pushing loco moving to the right.
LLLLL-WPPWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

2. A<->B boundary bridged, AR B switches, so A, B, D and E now have the same polarity/phase.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

3. Wagon spans the A<->B boundary. No AR action. All further bridging of A<->B now irrelevant.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

4. B<->C bridged, AR B switches.
LLLLL-WPPWWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

5. C<->D bridged, AR D switches, so B, C, D and E now have the same polarity.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

6. C<->D still bridged, No AR action.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

7. Nothing bridged, wagon spans entire section C, but, crucially: there is no problem, because there is no short!.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

and so on - the rest is trivial..

And this is why only the minimum size of a TRUCK counts. A longer size becomes necessary only when a truck is not really a truck but a single axle. In other words: the long wagon only has one power pickup at each end (never mind the number of wheels, the number of pickups is what counts). Theoretically, it should STILL work. But in practice, a short on a one-axle truck can be very, very easily missed as it would be just the wheelset closing the gap  in the rails on its own. See what happens in the figures above when the short in step 5 is not seen. After that, spanning the C section (fig. 6) DOES mean trouble, because the B<->C short has not occurred and the reversers now fight - exactly what should be prevented from happening.

One solution (yours) is to make the C section long enough to span the entire wagon, which is fine and works. However, if that is not convenient (and the layout might well dictate that it isn't) then the other possibility is to stagger the cuts in the rails. That increases the chances of the one-wheel short occurring by a lot. This is the only advantage to staggering that I have ever been able to see. But I repeat: there is ONLY a problem in case of one-axle trucks (if you can even call them trucks).

Wouter

On Sun, 23 May 2021 at 12:27, Blair <smithbr@...> wrote:
BillD
Think of any single loco as a pair of wires hovering, one over each rail, drooping to touch the rail at each point where a loco's wheels do.  In order for a short length between AR sections to do its job, it must be long enough to prevent the wire from bridging the two AR sections.  So before the wire enters the second AR it must leave the first AR.

Now generalize this.  The user must be aware of the longest electrically connected "thing" he runs.  For example if you run a dummy loco but wire its wheelsets to an adjacent engine, then that becomes the longest "loco".
That defines the shortest non reversing section between AR sections.
Hope this helps.
Blair

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932=yahoo.com@groups.io>
Date: 2021-05-22 22:30 (GMT-05:00)
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

Blair

Wouter

Not so, because there is no technical reason why, when the second gap is bridged, that the first AR doesn't respond at the same time as the second, which is the same as having the two ARs adjacent with no fixed zone between.  As long as you create a condition wherein the two AR's see trip conditions from the same shorting action, the result will depend on the response times and sensitivities of the devices, resistance of the wiring internal to the loco, and the occasional interjection from Murphy.

Continue backing your train, repeating your diagram, until you place LLLLLLLL across all three - AR/Fixed/AR; additionally, ensure that you acknowledge in your scenario that LLLLL is one electrically contiguous locomotive, not a 4-diesel consist, for example, with independent pickups.  When your loco hits that second AR, you have a short, and yes, the first AR does respond as well as the second one, as that short propagates across the loco to the original AR.  That is the case I described (please revisit my description earlier in the thread), and the case I have encountered before on my own layout.  If you insist on describing the second short only in terms of a wired truck, passenger car, or wheelset that is independent of the locomotive wiring, then yes, you're correct, but not answering the OP question.

Blair

On 2021-05-23 12:27, whmvd wrote:
Blair,

What you describe is safe. No question. But it will also, usually, mean that you make a longer intermediate section than necessary. The 'bridging' you describe is indeed the tricky bit, but: only when the AR action HAS NOT YET TAKEN PLACE.

Consider a loco that pushes a lighted long wheelbase 4-axle wagon through the reversing sections, and all axles of the wagon have power pick-ups. Also assume that the polarity on entering is the worst it can be.

A = fixed polarity section
B = AR section 1
C = fixed polarity intermediate section
D = AR section 2
E = fixed polarity section
| = isolators
L = loco
W = wagon, with P being power pickup

Initially B has opposite polarity/phase to A, D and E (which are fixed), but the same as C.

1. Pushing loco moving to the right.
LLLLL-WPPWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

2. A<->B boundary bridged, AR B switches, so A, B, D and E now have the same polarity/phase.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

3. Wagon spans the A<->B boundary. No AR action. All further bridging of A<->B now irrelevant.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

4. B<->C bridged, AR B switches.
LLLLL-WPPWWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

5. C<->D bridged, AR D switches, so B, C, D and E now have the same polarity.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

6. C<->D still bridged, No AR action.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

7. Nothing bridged, wagon spans entire section C, but, crucially: there is no problem, because there is no short!.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

and so on - the rest is trivial..

And this is why only the minimum size of a TRUCK counts. A longer size becomes necessary only when a truck is not really a truck but a single axle. In other words: the long wagon only has one power pickup at each end (never mind the number of wheels, the number of pickups is what counts). Theoretically, it should STILL work. But in practice, a short on a one-axle truck can be very, very easily missed as it would be just the wheelset closing the gap  in the rails on its own. See what happens in the figures above when the short in step 5 is not seen. After that, spanning the C section (fig. 6) DOES mean trouble, because the B<->C short has not occurred and the reversers now fight - exactly what should be prevented from happening.

One solution (yours) is to make the C section long enough to span the entire wagon, which is fine and works. However, if that is not convenient (and the layout might well dictate that it isn't) then the other possibility is to stagger the cuts in the rails. That increases the chances of the one-wheel short occurring by a lot. This is the only advantage to staggering that I have ever been able to see. But I repeat: there is ONLY a problem in case of one-axle trucks (if you can even call them trucks).

Wouter

On Sun, 23 May 2021 at 12:27, Blair <smithbr@...> wrote:
BillD
Think of any single loco as a pair of wires hovering, one over each rail, drooping to touch the rail at each point where a loco's wheels do.  In order for a short length between AR sections to do its job, it must be long enough to prevent the wire from bridging the two AR sections.  So before the wire enters the second AR it must leave the first AR.

Now generalize this.  The user must be aware of the longest electrically connected "thing" he runs.  For example if you run a dummy loco but wire its wheelsets to an adjacent engine, then that becomes the longest "loco".
That defines the shortest non reversing section between AR sections.
Hope this helps.
Blair

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932=yahoo.com@groups.io>
Date: 2021-05-22 22:30 (GMT-05:00)
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

whmvd

Blair,

I find this very interesting scenario. And I think I've got it right.

You see, no matter what you do from fig. 7 onwards, crossing gaps and straddling or bridging, there are no shorts. Because the polarities (phases) are already the same. The first moment that any AR is going to do anything at all is when the first pickup crosses the D<->E gap.

Maybe you missed the pluses and minuses on the lines below the diagrams? In #7, there are three consecutive minuses for B, C, D. No shorts!

Wouter

On Sun, 23 May 2021 at 20:16, Blair <smithbr@...> wrote:

Wouter

Not so, because there is no technical reason why, when the second gap is bridged, that the first AR doesn't respond at the same time as the second, which is the same as having the two ARs adjacent with no fixed zone between.  As long as you create a condition wherein the two AR's see trip conditions from the same shorting action, the result will depend on the response times and sensitivities of the devices, resistance of the wiring internal to the loco, and the occasional interjection from Murphy.

Continue backing your train, repeating your diagram, until you place LLLLLLLL across all three - AR/Fixed/AR; additionally, ensure that you acknowledge in your scenario that LLLLL is one electrically contiguous locomotive, not a 4-diesel consist, for example, with independent pickups.  When your loco hits that second AR, you have a short, and yes, the first AR does respond as well as the second one, as that short propagates across the loco to the original AR.  That is the case I described (please revisit my description earlier in the thread), and the case I have encountered before on my own layout.  If you insist on describing the second short only in terms of a wired truck, passenger car, or wheelset that is independent of the locomotive wiring, then yes, you're correct, but not answering the OP question.

Blair

On 2021-05-23 12:27, whmvd wrote:
Blair,

What you describe is safe. No question. But it will also, usually, mean that you make a longer intermediate section than necessary. The 'bridging' you describe is indeed the tricky bit, but: only when the AR action HAS NOT YET TAKEN PLACE.

Consider a loco that pushes a lighted long wheelbase 4-axle wagon through the reversing sections, and all axles of the wagon have power pick-ups. Also assume that the polarity on entering is the worst it can be.

A = fixed polarity section
B = AR section 1
C = fixed polarity intermediate section
D = AR section 2
E = fixed polarity section
| = isolators
L = loco
W = wagon, with P being power pickup

Initially B has opposite polarity/phase to A, D and E (which are fixed), but the same as C.

1. Pushing loco moving to the right.
LLLLL-WPPWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

2. A<->B boundary bridged, AR B switches, so A, B, D and E now have the same polarity/phase.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

3. Wagon spans the A<->B boundary. No AR action. All further bridging of A<->B now irrelevant.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

4. B<->C bridged, AR B switches.
LLLLL-WPPWWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

5. C<->D bridged, AR D switches, so B, C, D and E now have the same polarity.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

6. C<->D still bridged, No AR action.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

7. Nothing bridged, wagon spans entire section C, but, crucially: there is no problem, because there is no short!.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

and so on - the rest is trivial..

And this is why only the minimum size of a TRUCK counts. A longer size becomes necessary only when a truck is not really a truck but a single axle. In other words: the long wagon only has one power pickup at each end (never mind the number of wheels, the number of pickups is what counts). Theoretically, it should STILL work. But in practice, a short on a one-axle truck can be very, very easily missed as it would be just the wheelset closing the gap  in the rails on its own. See what happens in the figures above when the short in step 5 is not seen. After that, spanning the C section (fig. 6) DOES mean trouble, because the B<->C short has not occurred and the reversers now fight - exactly what should be prevented from happening.

One solution (yours) is to make the C section long enough to span the entire wagon, which is fine and works. However, if that is not convenient (and the layout might well dictate that it isn't) then the other possibility is to stagger the cuts in the rails. That increases the chances of the one-wheel short occurring by a lot. This is the only advantage to staggering that I have ever been able to see. But I repeat: there is ONLY a problem in case of one-axle trucks (if you can even call them trucks).

Wouter

On Sun, 23 May 2021 at 12:27, Blair <smithbr@...> wrote:
BillD
Think of any single loco as a pair of wires hovering, one over each rail, drooping to touch the rail at each point where a loco's wheels do.  In order for a short length between AR sections to do its job, it must be long enough to prevent the wire from bridging the two AR sections.  So before the wire enters the second AR it must leave the first AR.

Now generalize this.  The user must be aware of the longest electrically connected "thing" he runs.  For example if you run a dummy loco but wire its wheelsets to an adjacent engine, then that becomes the longest "loco".
That defines the shortest non reversing section between AR sections.
Hope this helps.
Blair

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932=yahoo.com@groups.io>
Date: 2021-05-22 22:30 (GMT-05:00)
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

Blair

I stand corrected, Wouter - sort of.

In the case you outline, you are correct, Wouter - if we presume that the 1st AR was preconditioned by a previous section that is in the same phase as the short fixed section between the two ARs, then there can be no series of shorts.  (However, why you chose to reverse the train in your diagram is beyond me; length or direction of train is irrelevant to this portion of the discussion).  The thing is, can we assume that the AR was entered from a segment of the 'correct' phase?  Certainly, in the convoluted topology of my former layout, that was not the case(a moibius strip, it was).  My train came from a section that was anti-phase, because it had done a traverse across the base of a peninsula, coming back up the far side.  Hence I had fixed trackage(+/-), AR, fixed trackage(-/+), AR, and fixed trackage(+/-).   In that case, I think the description I used is applicable.

So I think we can both be announced to be correct, depending upon the topology of the greater layout.

Blair

On 5/23/2021 3:16 PM, Blair wrote:

Wouter

Not so, because there is no technical reason why, when the second gap is bridged, that the first AR doesn't respond at the same time as the second, which is the same as having the two ARs adjacent with no fixed zone between.  As long as you create a condition wherein the two AR's see trip conditions from the same shorting action, the result will depend on the response times and sensitivities of the devices, resistance of the wiring internal to the loco, and the occasional interjection from Murphy.

Continue backing your train, repeating your diagram, until you place LLLLLLLL across all three - AR/Fixed/AR; additionally, ensure that you acknowledge in your scenario that LLLLL is one electrically contiguous locomotive, not a 4-diesel consist, for example, with independent pickups.  When your loco hits that second AR, you have a short, and yes, the first AR does respond as well as the second one, as that short propagates across the loco to the original AR.  That is the case I described (please revisit my description earlier in the thread), and the case I have encountered before on my own layout.  If you insist on describing the second short only in terms of a wired truck, passenger car, or wheelset that is independent of the locomotive wiring, then yes, you're correct, but not answering the OP question.

Blair

On 2021-05-23 12:27, whmvd wrote:
Blair,

What you describe is safe. No question. But it will also, usually, mean that you make a longer intermediate section than necessary. The 'bridging' you describe is indeed the tricky bit, but: only when the AR action HAS NOT YET TAKEN PLACE.

Consider a loco that pushes a lighted long wheelbase 4-axle wagon through the reversing sections, and all axles of the wagon have power pick-ups. Also assume that the polarity on entering is the worst it can be.

A = fixed polarity section
B = AR section 1
C = fixed polarity intermediate section
D = AR section 2
E = fixed polarity section
| = isolators
L = loco
W = wagon, with P being power pickup

Initially B has opposite polarity/phase to A, D and E (which are fixed), but the same as C.

1. Pushing loco moving to the right.
LLLLL-WPPWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

2. A<->B boundary bridged, AR B switches, so A, B, D and E now have the same polarity/phase.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

3. Wagon spans the A<->B boundary. No AR action. All further bridging of A<->B now irrelevant.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          +                -                     +                                +

4. B<->C bridged, AR B switches.
LLLLL-WPPWWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     +                                +

5. C<->D bridged, AR D switches, so B, C, D and E now have the same polarity.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

6. C<->D still bridged, No AR action.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

7. Nothing bridged, wagon spans entire section C, but, crucially: there is no problem, because there is no short!.
LLLLL-WPPWWWWWWWWWPPW  =>
=============A===========|==============B=============|==C==|==================D=====================|==========E===========
+                          -                -                     -                                +

and so on - the rest is trivial..

And this is why only the minimum size of a TRUCK counts. A longer size becomes necessary only when a truck is not really a truck but a single axle. In other words: the long wagon only has one power pickup at each end (never mind the number of wheels, the number of pickups is what counts). Theoretically, it should STILL work. But in practice, a short on a one-axle truck can be very, very easily missed as it would be just the wheelset closing the gap  in the rails on its own. See what happens in the figures above when the short in step 5 is not seen. After that, spanning the C section (fig. 6) DOES mean trouble, because the B<->C short has not occurred and the reversers now fight - exactly what should be prevented from happening.

One solution (yours) is to make the C section long enough to span the entire wagon, which is fine and works. However, if that is not convenient (and the layout might well dictate that it isn't) then the other possibility is to stagger the cuts in the rails. That increases the chances of the one-wheel short occurring by a lot. This is the only advantage to staggering that I have ever been able to see. But I repeat: there is ONLY a problem in case of one-axle trucks (if you can even call them trucks).

Wouter

On Sun, 23 May 2021 at 12:27, Blair <smithbr@...> wrote:
BillD
Think of any single loco as a pair of wires hovering, one over each rail, drooping to touch the rail at each point where a loco's wheels do.  In order for a short length between AR sections to do its job, it must be long enough to prevent the wire from bridging the two AR sections.  So before the wire enters the second AR it must leave the first AR.

Now generalize this.  The user must be aware of the longest electrically connected "thing" he runs.  For example if you run a dummy loco but wire its wheelsets to an adjacent engine, then that becomes the longest "loco".
That defines the shortest non reversing section between AR sections.
Hope this helps.
Blair

Sent from his Galaxy

-------- Original message --------
From: "mgj21932 via groups.io" <mgj21932=yahoo.com@groups.io>
Date: 2021-05-22 22:30 (GMT-05:00)
Subject: [w4dccqa] Minimum length on Non-reversing segment

I realize that we have addressed the question of Auto-Reverse (AR) segment wiring extensively in the past ( a discussion in which I even participated).  I understand that the track of an AR-wired segment cannot be electrically connected with another AR-wired segment, as the two will go to war with one another, switching back and forth until something blows.  And I also understand that for similar reasons, the minimum length of an AR-wired segment must be as long as the longest train that will run over it (or at least as long as the length of the train to the last car with either steel/metal wheels or an electrical pick-up that would cause a short if the "polarity" is not correct).

What I am unclear about is the minimum length of the fixed (non-reversing) "polarity" segment that separates AR-wired segments.   I believe I recall a suggestion that the non-reversing segment be at least the length of a locomotive, but not sure why that would be required.

In my case, my layout is a basic figure-8 pattern with two direction reversing segments, each wired with an AR circuit.   Problem is that one of those segments (about 6 feet long) was shorter than my nearly 7-foot long N&W passenger train (six 11+" MTH passenger cars plus a 13+" MTH N&W J loco).  In reconfiguring, I can lengthen the AR-wired segments to include portions of the figure 8 (assuring that both AR-wired segments are long enough), particularly if I do not have to be concerned that the length of the non-switching "polarity" segment have a substantial minimum length, e.g., not at least as long as my longest train.

I also assume that I can have more than one non-reversing segment separating AR-wired segments without any problems as long the "polarity" of all of the non-reversing segments is the same.

Apologies if we have already covered this sufficiently in previous discussions; but I do not recall the definitive answer.

Bill D
N&W Steam Only

Don Vollrath

Bill D... you said...

“In my case the fixed segments are several feet long, but not the length of my longest train; however the AR-switched segments are comfortably longer than the longest train. “

If this is true it should work as is. If this doesn’t work your layout is not isolated or wired as you describe.

DonV

D B

Blair,
👍
Bill D

 1 - 16 of 16