Interesting thread, guys, especially for a DCC guy who only started with the basics in DC before going to DCC.
Rich
if there are separate turnouts to the reverse section, i think it’s common to re-enter a reverse loop with the turnout already aligned, forgetting that the reverse loop polarity needs to be flipped.
i don’t thnk the east/west concept would always make sense on smaller layouts. I don’t think ARs would be so popular if it did.
i proposed that a signal would indicate to STOP if the reverse section polarity needed to be toggled. (a green signal could be used with DC to indicate it’s OK when there was voltage on the rails).
the position of a reversing switch wouldn’t indicate polarity if there were 2 at different locations wired in series
Linn Westcott explained it pretty clearly in An HO Railroad that Grows (1972) before Atlas components were available
1 Like
Thank goodness for DCC.
Rich
i don’t think it’s DCC, it’s the same technology that made DCC affordable that could be used on DC layouts.
I’m just grateful for DCC auto-reversers. So simple.
Rich
1 Like
Even if the reversing section is an effectively straight section with different entrance control turnouts, it can be visually defined as “entering from the left, west” and “entering from the right, east”, and the switch labeled and oriented accordingly.
On the layout at the Severna Park club, one of the two reverse sections was simply a block in a long run. It had no exit/entrance turnouts. You just knew you were entering from east or west and set the switch accordingly.
Sheldon
1 Like
That seems to make perfect sense.
Rich
1 Like
Yep, and still do. “NOTE: When using controller, do not use reversing switches on power packs…”:Atlas #220 Controller Wiring Instructions
1 Like
With respect to having the direction control on the control panel and not the throttle: Use toggle switches for the direction control where they are placed on the track diagram with the switches oriented to move in the intended direction of travel. In the case of a reversing loop, the engineer would make sure that the toggle was in the right direction before entering the loop and once the train was in the loop would then make sure the toggle on the main line would be in the right direction before exiting the loop.
I remember a discussion from ~55 years ago about automating a reversing loop. The other person mentioned that he tried coming up with a truth table and seeing the was no way to make it work with combinatorial logic. I now realize that an automated reversing loop would require some sort of state machine that remembers which direction the train is headed.
doesn’t it just depend on turnout alignment and presence of loco approaching a particular turnout if there are multiple entrances to a reversing section?
The presence of the locomotive is part of what I was referring to by a state in a state machine as you don’t want to be changing polarity on an occupied block with a DC locomotive.
i don’t think a state machine is the appropriate tool.
it seems similar to tracking a loco position by recognizing that a block adjacent to the current block occupied by the loco became occupied.
a DC auto-reverser needs to control the polarity of both the mainline and reverse section and needs to know the polarity to maintain the direction of the loco.
one approach could use IR detectors on either side of gaps delimiting the reverse section and knowing the turnout position. the cases below are for when the mainline and rev-sect polarities match thru the non-diverging turnout route
IR-pts TO-norm set rev-sect same as mainline
IR-pts TO-rev set rev-sect opposite of mainline
IR-Ndiv TO-norm set mainline same as rev-sect
IR-Ndiv TO-rev turnout mis-aligned - STOP
IR-div TO-norm turnout mis-aligned - STOP
IR-div TO-rev set mainline opposite of rev-sect
there would be 4 IR sensors if there are separate entrance/exits to the reverse section (not thru a common turnout) and no need to know turnout position
IR-r1 set mainline normal
IR-11 set rev-sect normal
IR-r2 set mainline reversed
IR-11 set rev-sect reversed
i beleive the following is correct if implemented as a state machine
each state is represented by the combination of the mainline and rev-sect polarities, either normal/reversed and the turnout position, normal/reversed
the stimuli are the 3 IR sensors for the single turnout rev-loop: sensor at the points, the diverging and non-diverging tracks.
Stimuli that require not change in state are not shown, as well as changes in turnout position.
What happens when you simply reverse the direction of the train?
How do you reverse the direction of the train inside the reverse loop?
Are we assuming the IR sensors simply respond to based on the one triggered first?
If the mainline polarity is reversed automatically, how will the operator keep track of which direction a train that is out of sight is traveling? His direction switch will no longer indicate that.
How does automatically reversing the main get reconciled with other operators in other blocks on the mainline?
What kind of throttles are we assuming are in use? power packs?
Sheldon
use the direction switch on the throttle to move forward/reverse
how else?
whatever provide power to the track
OK, but I suspect this will have practical operational issues on multi cab/multi block layouts. And this is not one size fits all, since there are lots of different methods by which DC layouts are wired.
Historically, as DC layouts get more complex, and/or larger physically, some sort of “east - west” convention is established for direction of travel identification.
DC trains don’t recognize the idea of “forward - reverse”.
If I’m standing in front of a piece of track with a locomotive. If the rail farthest from me is positive the locomotive will move to my left, with no regard for what is the front or back of the locomotive.
When I change the polarity so the closest rail is positive, the locomotive will move to my right.
Two F7’s back to back will move in the same direction, one forward and the other in reverse relative to their cabs.
If you only change the polarity of the mainline block attached to the reverse loop? What will happen when you try to enter the next block?
At this point I have not analyzed this, but I suspect there will be polarity conflicts throughout any multi block/multi cab system.
I do not know for certain without plugging this idea into some standard DC wiring schemes.
Sheldon
just trying to figure out what it would take
For DC equipment, when the “right-hand rail” is positive, the powered equipment shall move forward.
power to the mainline, not just the adjacent block
may need separate circuits for each throttle.
ssems that there needs to be a single circuit controlling for reversing sections since they all control the mainline polarity
If I have a piece of track in front running left to right and the nearest rail is positive, a DC locomotive placed on that track will run left to right no matter if the locomotive is facing left or right. In the first case it will be running backwards. Hence the concept of east-west.
Correct, “forward” and “right hand rail” are defined as sitting in the cab of the locomotive looking down the track.
So back to my illustration. You the operator are looking at a piece of track the runs left and right in front of you.
You sit a locomotive on that track facing to your left. That makes the rail farthest from you the right hand rail of that loco. It will move to the left if that rail is positive.
Without changing the polarity of the track, you pick the loco and turn it around. It is now facing to your right. That makes the rail closest to you the right hand rail of the locomotive. The locomotive will still move the left, only now it will be moving in reverse.
Sheldon
yes. that’s what the NMRA standard says