Hidden staging yards, electrical control thereof
Bruce H. Stull
Hmmm. I think a mainframe should do the trick.
Bruce
--
"My life must be of some benefit to others".
David J. Starr <David...@analog.com> wrote in message
news:37229F...@analog.com...
> Our club is doing an underground, out of sight, staging yard.
The
> operator won't be able to see his train, or all the other trains
in the
> staging yard. It will have 6 staging tracks, with a turnout
ladder at
> both ends, and additional turnouts to let a train arriving or
departing
> the staging yard to come in from, or go out onto either the east
bound
> main or the west bound main. This feature makes the whole
staging yard
> a big reversing loop.
> Collisions and derailments underground require a lot of hands
and
> knees work to fix up, so the staging yard controller should do
> everything it can to prevent them.
> It ought to prevent an arriving train running onto a full
staging
> track. It should stop an arriving train before it runs thru the
> departure turnouts. On departure it should insure that only the
desired
> train moves, rather than all of them. It should inhibit
throwing
> turnouts while the train is running thru them and it should
inhibit
> accidental backing moves underground. Selecting a track for
arrival or
> departure should align all the turnouts and apply track power to
the
> selected track (and ONLY them selected track) .
> We could do a hardware controller out of gates and PALs and
5 volt
> logic, or we could add staging yard control duties to the PC
that runs
> the signalling system, or we could get a single board computer
to run
> just the staging yard.
> As a backup to the controller we plan to have closed circuit
TV
> coverage of key sections of the staging yard. However we feel
that
> running trains by watching TV is pretty tricky, so we want the
control
> system to actively inhibit operator mistakes that lead to
someone
> crawling under the table to rerail trains.
> Any advice, experience, suggestions or war stories about
hidden
> staging yards would be welcome.
>
> David Starr
David J. Starr
Charles A Davis
snip
> Any advice, experience, suggestions or war stories about hidden
> staging yards would be welcome.
>
> David Starr
Hi David:
Model Railroader Magazine, May, 1999, page 84. Should get you going.
Chuck D.
--
-----------------------------------------------------------
He, who will not reason, is a bigot; William Drumond,
he, who cannot, is a fool; Scottish writer
and he, who dares not, is a slave. (1585-1649)
While he that does, is a free man! Joseph P. 1955-
-----------------------------------------------------------
Chuck Davis / Sutherlin Industries FAX # (804) 799-0940
1973 Reeves Mill Road E-Mail -- c...@gamewood.net
Sutherlin, Virginia 24594 Voice # (804) 799-5803
Arthur Armstrong
big clip
> Any advice, experience, suggestions or war stories about hidden
> staging yards would be welcome.
>
I once had a much smaller and simpler hidden staging yard on a home
layout. I had five tracks doubled-ended with a capacity of one train
per track.
I put an IR optical detector at the fouling point at each end and
another detector in the middle of the yard track. The detectors lighted
LEDs on the control panel.
I pulled an unseen train in until the end lights were out and the center
light was on. I had to remember (or peek!) to recall which train was on
which track and in what direction it was pointing.
Turnout control was tortoise machines with panel switches - nothing
automatic but there were panel lights indicating the direction.
Power to the tracks was indicated only by toggle switch position on the
panel.
Good luck with yours. It sounds pretty imposing.
Art Armstrong
David J. Starr
turnouts and apply track power. But it does not allow optical train
detectors at the fouling point of each staging track, to make sure the
train gets completely clear of the yard throat turnouts before killing
track power to stop the train. And it does not protect against running
the new arriving train into an occupied staging track, which is the most
obvious operator error. We figure one rear end collision under ground
could derail about 50 cars, and block the entire staging yard. Even if
nothing is damaged, the necessary groveling around under the layout to
rerail everything, AND all the unprintable things that would be said,
put this mishap into the major catastrophy class.
David J. Starr
> I put an IR optical detector at the fouling point at each end and
> another detector in the middle of the yard track. The detectors lighted
> LEDs on the control panel.
OK, we are thinking of doing the same, as far as detectors go. We
just want to go one step further and automatically stop an arriving
train before it hits the departure turnouts, just in case a new operator
makes a mistake.
>
> I pulled an unseen train in until the end lights were out and the center
> light was on. I had to remember (or peek!) to recall which train was on
> which track and in what direction it was pointing.
We plan to go only one way thru the staging yard since we can depart
in either direction. So we don't have to worry about direction. And,
we just want to guard against derailments. If the operator brings out
the wrong train that's too bad, but we don't have to duck under the
layout and put cars back on the track.
>
> Turnout control was tortoise machines with panel switches - nothing
> automatic but there were panel lights indicating the direction.
All our yards have a "one button" turnout control feature. The one
button lines up all the turnouts to get to the selected track. Since
everyone is used to yards that work this way, we plan to do the staging
yard in the same manner just to make it easier to remember how
everything works.
Roger L. Traviss
"David J. Starr" wrote:
> Looked at that. We need something fancier. It does line up the
> turnouts and apply track power. But it does not allow optical train
> detectors at the fouling point of each staging track, to make sure the
> train gets completely clear of the yard throat turnouts before killing
> track power to stop the train. And it does not protect against running
> the new arriving train into an occupied staging track, which is the most
> obvious operator error. We figure one rear end collision under ground
> could derail about 50 cars, and block the entire staging yard. Even if
> nothing is damaged, the necessary groveling around under the layout to
> rerail everything, AND all the unprintable things that would be said,
> put this mishap into the major catastrophy class.
Agreed. I asked this question a few months ago and mention a UK modeller in
the 60s, the Rev Peter Denny, who had a staging yard controlled by a device
called the "Automatic Crispen" (Named after one of his sons). This back the
days of relays and such, no fancy electronics.
I too was disappointed in the MR article. The question I put several months
ago was for someone who could design a fully automatic staging yard
control. One that would automatically pick up an inbound train from the
road engineer and drive it into an empty staging track and stop it in the
clear. At the correct time (With or without a random delay time to mimic
over the road delays on the unmodelled portions of the railway) the
automatic staging yard controller would dispatch a train onto the modelled
portion of the railway, aligning all switches enroute, and drive it in a
realistic fashion to a spot where the road engineer could take over for the
rest of the trip over the model railroad.
I envisage a staging yard that both receives and dispatches trains with no
human control at all and fully automates that unrealistic part of model
railroading. Surely that can't be too difficult a task for the right person
in this computer age?
Cheers,
Roger T.
David J. Starr
semi automatic staging yard, where trains are run by walk around
throttles, just as we do it on the rest of the layout. The control
system is a backup to the operator, to stop the train before it hits
something.
David Starr
David Starr
Robert C. Schwartz
may give you some ideas for controlling your yard. Mine is a one direction
yard so the below might not fit exactly, but I hope it helps. I have the
yard at the end of the line followed by a reversing loop so everything
coming off of the loop feeds right in to the yard.
1) Take a look at Paul Mallery's "Electrical Handbook for Model Railroads,
Volume 2". On page 128 Figure 13-14 he shows a stall switch machine
controller. I use an extension of this to control a five-track hidden
staging yard. One track is normally a through track so it's really a
four-track yard but additional stages can be integrated to do about
whatever you want. Now, that may take care of one item; here's some more
that I did:
2) All tracks have a stopping section about 20" ahead of the departure
turnouts; the power to the stopping section is controlled by the turnout -
I use Tortii, so the contacts provide the power routing.
3) All tracks have a photo detector in the stopping section; this detector
provides two functions: 1) Panel indication and 2) selection of inbound
track. If the detector is lighted, the corresponding input turnout is
thrown for that track; if dark, that turnout is thrown to the through
position. Since all input turnouts are controlled the same way, the
turnouts are lined up to fill the first available staging track.
4) The input yard throat and the reverse track turnout have an IR beam that
shows when something is on the turnouts - when occupied, I defeat the input
track selector in 3) above so a turnout won't be thrown under a car or
engine. I do this by disconnecting the power to the turnouts. The IR device
I'm using is an old Radio Shack door annunciator from my scrap box with a
couple of mirrors to direct and reflect the beam.
5) The output yard throat hasn't been protected yet, but I haven't had any
problems with derailments so far.
6) An interlocking pushbutton switch (converted to momentary action) is
used to select the outbound track and drives the Mallery circuit in 1)
above. I have two sets of selectors, one at the operator's position (with
occupancy lights) and one at the staging yard (another room) so I can
operate and test the yard.
If I want to bypass the staging for continuous running and don't have it
full of trains, I just put a paint jar over each unoccupied photodetector.
Then, the staging end just acts like a reversing loop.
Hope this gives you some useful ideas.
Good Luck,
Bob Schwartz
David J. Starr <David...@analog.com> wrote in article
<3723C1...@analog.com>...
> Looked at that. We need something fancier. It does line up the
><SNIP>
Steve Cizek
It used RailCommand for train control, and the computer took the
train from the town before staging (a crew change point) and into
staging, putting it safely away. Likewise, it would bring a train
out of staging to the first town (a crew change point) were the
road crew would pick it up. The interface was a track diagram of
staging that allowed you to click on a train that you wanted brought
out next. Very nice.
Roger L. Traviss wrote:
> I too was disappointed in the MR article. The question I put several months
> ago was for someone who could design a fully automatic staging yard
> control. One that would automatically pick up an inbound train from the
> road engineer and drive it into an empty staging track and stop it in the
> clear. At the correct time (With or without a random delay time to mimic
> over the road delays on the unmodelled portions of the railway) the
> automatic staging yard controller would dispatch a train onto the modelled
> portion of the railway, aligning all switches enroute, and drive it in a
> realistic fashion to a spot where the road engineer could take over for the
> rest of the trip over the model railroad.
>
> I envisage a staging yard that both receives and dispatches trains with no
> human control at all and fully automates that unrealistic part of model
> railroading. Surely that can't be too difficult a task for the right person
> in this computer age?
>
> Cheers,
>
> Roger T.
--
Steve Cizek
steve...@lucent.com
David J. Starr
>
> 2) All tracks have a stopping section about 20" ahead of the departure
> turnouts; the power to the stopping section is controlled by the turnout -
> I use Tortii, so the contacts provide the power routing.
This sounds really good, and goofproof.
>
> 3) All tracks have a photo detector in the stopping section; this detector
> provides two functions: 1) Panel indication and 2) selection of inbound
> track. If the detector is lighted, the corresponding input turnout is
> thrown for that track; if dark, that turnout is thrown to the through
> position. Since all input turnouts are controlled the same way, the
> turnouts are lined up to fill the first available staging track.
Very interesting idea. If all the staging tracks are full, the train
just runs thru the one open track and proceeds out to the main line.
Good fail safe idea.
I like it.
>
> 4) The input yard throat and the reverse track turnout have an IR beam that
> shows when something is on the turnouts - when occupied, I defeat the input
> track selector in 3) above so a turnout won't be thrown under a car or
> engine. I do this by disconnecting the power to the turnouts. The IR device
> I'm using is an old Radio Shack door annunciator from my scrap box with a
> couple of mirrors to direct and reflect the beam.
Good idea. This will guard against the "last minute change of mind"
that throws the turnouts AFTER the train has entered them. "Oh I wanted
track 4 instead of 6. Push the turnout selector REAL fast to beat the
train".
>
David Starr
Frank A. Rosenbaum
gapped sections on each staging yard track. This way the toggle closest to
the train is on as the train enters the track and the toggle furthest from
the train is off at the front of the track, well back from the fouling
point. You should also have some kind of optical detection device, such as a
Circuttron. This could be wired to a panel and light up indicators as the
train moves through the yard. It is bi-directional so you don't have to
worry about which way the train is going. If you have a detector at the
throat, and fouling points, you will know if there is a potential problem
and can take corrective action before there is a real problem.
I haven't tried this, because my club doesn't have any staging tracks, yet.
And my railroad at home is just a switching RR.
--
From Frank in White Plains, NY
New Email: f...@cyburban.com
David J. Starr wrote in message <372479D1...@analog.com>...
> I do remember the phase "Automatic Crispen". We were thinking of a
>semi automatic staging yard, where trains are run by walk around
>throttles, just as we do it on the rest of the layout. The control
>system is a backup to the operator, to stop the train before it hits
>something.
>
>David Starr
>
>
>
>David Starr
>
>
>"Roger L. Traviss" wrote:
>
>> Agreed. I asked this question a few months ago and mention a UK modeller
in
>> the 60s, the Rev Peter Denny, who had a staging yard controlled by a
device
>> called the "Automatic Crispen" (Named after one of his sons). This back
the
>> days of relays and such, no fancy electronics.
>>
David J. Starr
track. However I am afraid that a couple of toggles might be accidently
left ON simultaniously, and so when the engineer throttles up to depart
the yard, two or more trains start to move, and we have a collision or
derailment before the operator notices he is in trouble. However if we
also had a "stopper" block just before the departure turnouts, the
selected train would proceed thru the turnouts and the un selected train
would halt in the "stopper" block just before the turnouts 'cause the
turnouts would be set against it.
We are thinking about a rotary switch for staging track selection to
make it impossible to select two tracks at the same time.
In line with "KISS", we want to make the as much of the circuitry be
plain old electro mechanical stuff so the stagng yard can be used even
if the fancy electronics go belly up.
David Starr
Frank A. Rosenbaum wrote:
>
> In the manner of KISS, how about using two simple on off toggle switches and
> gapped sections on each staging yard track. This way the toggle closest to
> the train is on as the train enters the track and the toggle furthest from
> the train is off at the front of the track, well back from the fouling
> point.
[some snippage]
David J. Starr
sensor shows the train has cleared the fouling point. Of course we need
some backup just in case the train never clears the fouling point,
'cause it's too long for the staging track or the train uncouples by
accident leaving the turnouts fouled with the uncoupled cars.
David Starr
John Purbrick wrote:
>
> Well, optical detectors are easy enough to set up, though I'd suggest
> infrared emitters and detectors rather than the light bulbs and CdS sensors
> Larry is looking at. You could put a detector near each end of each track,
> a few inches short of the fouling point, then gate all the outputs for
> the "west' end through the same logic used to select the switches at the
> "east" end. That way, the active sensor would be the one corresponding to
> the track the train came in on. The train runs till it hits the sensor,
> then stops--automatically perhaps. If a track is already occupied, one
> or other of its two sensors will be covered and that should warn an
> approaching operator--again, automatic rejection of an attempt to enter
> an occupied track is possible. This isn't absolutely foolproof but something
> like it should work if the operators are willing to cooperate.
>
> John Purbrick
John Purbrick
Robert C. Schwartz
experiences with ambient (sometimes supplemented) and CdS detectors. If one
picks the solid state route, an insurance policy would be to modulate the
emitters and filter the detector output (like is done with IR remote
controls for TV) to minimize interference.\
Bob Schwartz
John Purbrick <jpur...@nyx10.nyx.net.nyx.net> wrote in article
<92534787...@iris.nyx.net>...
> Well, optical detectors are easy enough to set up, though I'd suggest
> infrared emitters and detectors rather than the light bulbs and CdS
sensors
<Snip>
daForce
<37242DFF...@islandnet.com>:
>"David J. Starr" wrote:
>
>> Looked at that. We need something fancier. It does line up the
>> turnouts and apply track power. But it does not allow optical
>> train
[]
>I too was disappointed in the MR article. The question I put
>several months ago was for someone who could design a fully
>automatic staging yard control. One that would automatically pick
>up an inbound train from the road engineer and drive it into an
>empty staging track and stop it in the clear. At the correct time
>(With or without a random delay time to mimic over the road delays
>on the unmodelled portions of the railway) the automatic staging
>yard controller would dispatch a train onto the modelled portion
>of the railway, aligning all switches enroute, and drive it in a
>realistic fashion to a spot where the road engineer could take
>over for the rest of the trip over the model railroad.
the software for this is trivial Roger, given sufficient reliable
sensors so the software knows where the train is and can control
all the various turnouts and power, the rest is pretty much easy.
the major thing is for the puta to know that the train is moving
when it expects it to be moving. if it applies a certain voltage
to make the train move, and the loco has stalled, hit a dead spot
etc, then the puta needs to know this state. the problem is how to
detect train 'movement' instead of 'presence', and what to do about
failures.
>I envisage a staging yard that both receives and dispatches trains
>with no human control at all and fully automates that unrealistic
>part of model railroading. Surely that can't be too difficult a
>task for the right person in this computer age?
I envisage doing this someday, prob in N, prob with an automated
transfer table instead of turnouts, using an old 286/386 PC to
control it. just made a bit of a drawing and one needs quite a few
optical sensors etc, but it's not rocket science and can by built
using offtheshelf track detectors etc. the software can be written
to autoaccept trains into the nearest open track, and deliver
trains out of either end by selecting the transfertrack number and
direction at the keyboard, or extended to produce trains out of
either end at scheduled times (any takers on a system with an
airhorn to announce the train? :-) one error condition that must
also be handled intelligently is a train that is too long for the
transfer table. can use DC or DCC, does change some aspects
though, but if one uses DCC decoders that will autoswitch to DC,
the transfer can be DC only on reception. sending DCC stuff out may
be more complicated - simple solution is to simply have outgoing
stuff put directly under operator control as requested. I would use a
CoolerCrawler driven by an 8bit DAC.
an all DCC system hinges on the PC being able to either be told, or
detect, the loco number on each track (and MU's!), so it can select
and deselect as needed.
one nice thing about software systems is that they can easily be
changed to encompass your new ideas.
but as I said earlier, a critical thing is to detect motion, not
just presence. secondary is speed. if the train is travelling
fast, the controller must know to slow it down, but also know it is
still moving, so that it doesn't hit the end detector going so fast
it cant stop before fouling the exit.
dang, I knew I shouldn't come back here, I knew I'd get some bright
idea and get all enthusiastic about it (-:
steam and wind
--
David Forsyth DaForce A-T Iwr.Ru.Ac.Za
Keeper of the listserver for South African Railways fans _|_
His Part time gricer, kiter, photographer, father etc etc
| Way http://www.ru.ac.za/departments/iwr/staff/daf/welcome.html
| Up
asper...@mrmag.com
"David J. Starr" <David...@analog.com> wrote:
> Looked at that {Mike Dodd's SYC2 staging yard controller in the May MODEL
RAILROADER]. We need something fancier. It does line up the
> turnouts and apply track power. But it does not allow optical train
> train gets completely clear of the yard throat turnouts before killing
> track power to stop the train. And it does not protect against running
> the new arriving train into an occupied staging track, which is the most
> obvious operator error.
See the June issue of MR, pages 86-90, for Mike Dodd's laser occupancy
detector, which does indeed make sure the train gets completely clear of the
yard throat turnouts before killing track power to stop the train. It's true
that Mike is relying on a procedural method (car card packs present or not in
file boxes) to determine when staging tracks are occupied, but it would be
very easy to add occupancy detection to the staging tracks for a positive
indication. See Workin' on the Railroad in the December issue, pages
133-136, for a simple approach using Irdot detectors.
So long,
Andy Sperandeo
MODEL RAILROADER Maga
-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
Roger L. Traviss
[Good stuff snipped]
> but as I said earlier, a critical thing is to detect motion, not
> just presence. secondary is speed. if the train is travelling
> fast, the controller must know to slow it down, but also know it is
> still moving, so that it doesn't hit the end detector going so fast
> it cant stop before fouling the exit.
I wonder how they detect motion and speed on such automated transit
systems and Vancouver's "Sky Train"?
After all, this is a real world application of everything I envisage in a
fully automated staging yard.
> dang, I knew I shouldn't come back here, I knew I'd get some bright
> idea and get all enthusiastic about it (-:
If they price is right, there might be a small market for this kind of
product.
Cheers,
Roger T.
Chris Webster
> See the June issue of MR, pages 86-90, for Mike Dodd's laser occupancy
> detector, which does indeed make sure the train gets completely clear of the
> yard throat turnouts before killing track power to stop the train. It's true
Knowing that Mike doesn't read rec.models.railroad, I sent him the
earliest messages in this thread. Here's what he wrote back:
> > Thanks for copying me on this, Chris. You may pass along the following if
> > you wish:
> >
> > David J. Starr is correct that the SYC2 does not detect individual track
> > occupancy. The Mark I (first) version did this, but had a capacity of only
> > eight tracks, not 12. Furthermore, the first version had an LCD display
> > and keypad - which made it really slick to use, but it was too complicated
> > for the typical operator to use on a moment's notice. Conversations with
> > several operating gurus showed me that nearly all of them do not use
> > individual track detection; instead they use car cards in pockets or
> > accurate yard scheduling to determine which tracks are occupied or vacant.
> > The bottom line is that it was a design decision: Simple and 12 tracks, or
> > complex and eight tracks. I chose the former.
> >
> > You are correct that the laser occupancy detector (now in the June MR)
> > allows full coverage of the throat area - which may be extended into the
> > yard tracks themselves to include fouling points. Since it's optical,
> > there's no concern for where the tracks are electrically gapped.
> > ==========
> >
> > On the newgroup, Roger T. writes:
> > >I too was disappointed in the MR article. The question I put several months
> > >ago was for someone who could design a fully automatic staging yard
> > >control. One that would automatically pick up an inbound train from the
> > >road engineer and drive it into an empty staging track and stop it in the
> > >clear. At the correct time (With or without a random delay time to mimic
> > >over the road delays on the unmodelled portions of the railway) the
> > >automatic staging yard controller would dispatch a train onto the modelled
> > >portion of the railway, aligning all switches enroute, and drive it in a
> > >realistic fashion to a spot where the road engineer could take over for the
> > >rest of the trip over the model railroad.
> > >I envisage a staging yard that both receives and dispatches trains with no
> > >human control at all and fully automates that unrealistic part of model
> > >railroading. Surely that can't be too difficult a task for the right person
> > >in this computer age?
> >
> > Roger is correct - it's not too difficult a task at all, especially with
> > command control. Alas, it's probably also a project that Model Railroader
> > isn't interested in publishing for several reasons: Cost, complexity, and
> > the fact that many modelers simply don't want that sort of automation. I
> > designed my staging yard controller for my four staging yards; Andy
> > Sperandeo thought it would make a good article, so I wrote one.
> > ==========
> >
> > Finally, the SYC2 will handle double-end and reverse-loop yards fairly
> > well, as long as there are no more than six tracks. Here's a 75K Acrobat
> > file that explains how this can be done:
> > http://www.mindspring.com/~mdodd/syc2_tips.pdf
> >
> > Regards, Mike
> >
Hope this helps!
--Chris Webster Work: cweb...@aashto.org
www.aashto.org
Personal: cweb...@eng.buffalo.edu
www.acsu.buffalo.edu/~cwebster
----------------------------------------------------------------------
David J. Starr
soon as it cleared the fouling point. Can you minimize on detectors?
We want fouling point detectors to stop the next arriving train until
the yard ladder turnouts are clear of the current train.
Can we kill two birds with one stone (one detector) by using the fouling
point detector to stop the next arrival AND stop the train on the
staging track? After all, once the fouling point is clear, you don't
have to pull the train any farther down the staging track, you can stop
it as soon as the arrival turnouts are clear. On our yard, there is a
very long exit track, so it doesn't really matter just where the train
stops on the staging track, just so long as it clears the turnouts at
both ends.
However, just a fouling point detector fails if the train never
clears the fouling point. If the train is too long for the staging
track (some folk like really long trains) or the train uncouples halfway
into the staging track, the fouling point will never clear, leaving
power on the staging track until the locomotive runs into the departure
turnouts. Murphy's law says the departure turnouts won't be set to
accept the train, which gives us an underground derailment.
So, in addition to the fouling point detector, we probably want
positive derailment protection at the departure end, presumably a short
block just before the departure turnouts that only gets power when the
departure turnouts are aligned properly.
David Starr
John Purbrick wrote:
>
> Note that my suggestion, made as part of a different branch in this topic,
> uses a detector to STOP a train, not to protect the throat area of the yard.
> Therefore every occupied track has a blocked detector all the time and hence
> the accident you describe shouldn't happen. If a car gets uncoupled though,
> that's a problem.
>
David J. Starr
staging yard problem was akin to the block signaling problem, ie, just
keep a train from entering a block until the block is empty. As long as
the block is occupied, the system should stop a following train from
entering the occupied block. Does it matter if the train occupying the
block is moving or stopped?
A likely mishap is to try and run the arriving train into an occupied
staging track. We think the control system needs occupancy detectors
on each staging track and the capability of stopping the arriving train
before the yard ladder until the ladder turnouts are aligned to an EMPTY
staging track. This means the control system needs enough inputs to
handle the fouling point detectors and the occupancy detectors. Andy
Sperandeo mentioned a clever optical fouling point detector is coming in
the June MR, which I have not seen yet.
We planned to use diodes to limit the max track voltage to something
reasonable to keep underground speeds down. It may take some tinkering
to find a track voltage high enough to move all our locomotives and low
enough to keep the quicker ones down below Mach 1.
David Starr
Christian Tucker
SNIP
> Andy Sperandeo mentioned a clever optical fouling point detector is coming
in
> the June MR, which I have not seen yet.
It's nifty, uses a pocket laser.
Christian
Brockport
Marshall Abrams
>Arthur Armstrong wrote:
>
>> I put an IR optical detector at the fouling point at each end and
>> another detector in the middle of the yard track. The detectors lighted
>> LEDs on the control panel.
>
> OK, we are thinking of doing the same, as far as detectors go. We
>just want to go one step further and automatically stop an arriving
>train before it hits the departure turnouts, just in case a new operator
>makes a mistake.
>
May I suggest a stopping block instead of, or in addition to, the
optical detector. Isolate a length of track in advance of the fouling
point and wire a resistor-protected LED in series with the power feed
to this section. When a locomotive enters the section the LED will go
on and the locomotive will stop. Also wire a normally-open push button
in parallel with the LED and resistor. Push the button when you want
to move the locomotive.
You can automate this manual set-up by using relays.
Robert C. Schwartz
section's power eliminates the need for an optical detector in my staging
yard design. Thanks; this news group is great!
Also note - the stopping section-LED approach is applicable to DC or DCC
systems.
Bob
Marshall Abrams <marshall...@computer.org> wrote in article
<373020d8...@news.mitre.org>...
Patrick LaTorres
>section's power eliminates the need for an optical detector in my staging
>yard design. Thanks; this news group is great!
>Also note - the stopping section-LED approach is applicable to DC or DCC
>systems.
Another simple option is gap one rail and bridge the gap with a
diode, when the wheels cross the gap they loose power, when you revers
the direction control power can flow through the diode and the train
can back out of the track. You loose the detection option, but I don't
you'd find a simpler circuit.
Enjoy,
Pat LaTorres
Clark Martin
I thought about this and have discussed this with a friend who was trying
to design a staging yard. First of, obviously this only works with a
single ended yard. Second, it requires you to always drive trains in the
same way (may not be a problem). Finally it doesn't work well with
multiple unit trains.
My friend and I did work out a number of variations of trying to operate
double ended yards which used (IIRC) a rotary switch at each end to select
the inbound and outbound tracks.
I've been working out how to handle a single ended staging yard I just
built. I'm planning on using a Microprocessor to control it. The
computer will eventually be integrated with the rest of the layout (when I
get further along with it's computer control). I plan on using one IR
detector at the far end and one at the fouling point. Power will be shut
off when either the fouling point is cleared or when the train hits the
end detector. If the end detector trips and the fouling detector isn't
clear then it will gripe.
--
Clark Martin
Redwood City, CA, USA
Macintosh / Internet Consulting
cma...@pacbell.net
"I'm a designated driver on the Information Super Highway"
Anthony New
automatic staging yard, which accepts trains and runs them into empty
sidings, was described in 1939? Anyone with a access to an old book
collection might like to look up "Auto-electric model railways", by A
Duncan, published in (I think) Toronto in 1939. I read this as a boy
many years ago and based my own automatic layout on it, though at the
time I didn't have the space for the full setup. The book also described
how to make barious kinds of train detectors, interlock systems to
control crossings, and lots more. And all with post-office relays and
rotary selectors!
I have used various train detectors including those which detect wheels
passing (axle counters), track occupation (track-circuiting), or optical
train detection. Nowadays it's reasonably simple to build a
computer-operated train detection system with photodiodes; some
university departments in the UK use it to teach undergrads.
Anthony New, England, UK.
Anthony New
>
> This might work. The original idea was to stop the current train as
> soon as it cleared the fouling point. Can you minimize on detectors?
> We want fouling point detectors to stop the next arriving train until
> the yard ladder turnouts are clear of the current train.
> Can we kill two birds with one stone (one detector) by using the fouling
> point detector to stop the next arrival AND stop the train on the
> staging track? After all, once the fouling point is clear, you don't
> have to pull the train any farther down the staging track, you can stop
> it as soon as the arrival turnouts are clear. On our yard, there is a
> very long exit track, so it doesn't really matter just where the train
> stops on the staging track, just so long as it clears the turnouts at
> both ends.
> However, just a fouling point detector fails if the train never
> clears the fouling point. If the train is too long for the staging
> track (some folk like really long trains) or the train uncouples halfway
> into the staging track, the fouling point will never clear, leaving
> power on the staging track until the locomotive runs into the departure
> turnouts. Murphy's law says the departure turnouts won't be set to
> accept the train, which gives us an underground derailment.
> So, in addition to the fouling point detector, we probably want
> positive derailment protection at the departure end, presumably a short
> block just before the departure turnouts that only gets power when the
> departure turnouts are aligned properly.
>
A method I have seen used on fully automatic layouts is to make sure the
last vehicle (caboose?) is either track-circuited or (alternatively) has
a magnet fitted in a suitable position, and this is detected to lock
out the previous section until it has cleared the points. This copes
with split and derailed trains, and a time delay can sound an audible
alarm for operator attention. In the magnet version, "reed switches" are
used as detectors. (Please excuse me if the UK terminology is not clear)
Anthony New.
Anthony New
>
> Another simple option is gap one rail and bridge the gap with a
> diode, when the wheels cross the gap they loose power, when you revers
> the direction control power can flow through the diode and the train
> can back out of the track. You loose the detection option, but I don't
> you'd find a simpler circuit.
> Pat LaTorres
There are several snags with this technique, one of the greatest being
that when multiple-locos are used the front one stops and the rest still
push. A relatively simple solution is to replace the diode with a relay
which cuts the power to the whole track occupied by the train. A more
sophisticated version interfaces with the throttle to drop the speed
reliastically first. A PC isn't necessary for this though it can
simplify things a bit (sometimes!) if you have one connected already.
Anthony New
Jon Miller
yard throat detection using a laser. While this all sounds high tech and
expensive it's actually quite cheap. The author sells a kit than will do
two locations (all the electronics and board) for around $40. You have to
buy the lasers which are the cheap pointer type. I picked up mine for under
$8 each. You then use the lasers and mirrors to cover the yard throat for
detection. Really quite neat, look at the article!
Roger L. Traviss
Anthony New wrote:
> Interesting thread, this. Does anyone realise that a completely
> automatic staging yard, which accepts trains and runs them into empty
> sidings, was described in 1939? Anyone with a access to an old book
> collection might like to look up "Auto-electric model railways", by A
> Duncan, published in (I think) Toronto in 1939. I read this as a boy
> many years ago and based my own automatic layout on it, though at the
> time I didn't have the space for the full setup. The book also described
> how to make barious kinds of train detectors, interlock systems to
> control crossings, and lots more. And all with post-office relays and
> rotary selectors!
Sounds like the "Automatic Crispen" that the Rev Peter Deny used on his
"Buckingham Branch" in the UK back in the '60s.
While I don't personally like automatic control (driving) of model railways,
I've always though that the unrealistic "operation" of staging yards
should/could be made fully automatic. I envisage, as I've written before, a
staging yard dispatcher that completely takes care of the staging yard.
Automatically dispatching and receiving trains with any intervention from a
human operator. With this system, the human operator just drives the train
over the sceniced part of the model railroad and never has to worry about
the unrealistic staging yard.
The automatic staging yard dispatcher could dispatch trains from "beyond the
basement" at the appropriate time and, if required, automatically delay
trains to simulate late running etc..
I really don't see why an operator should -have- to drive trains into and
out of the staging yard as some people have advocated. Wouldn't it be much
more realistic if the operator waited down by the depot at the crew change
point for his/her train to arrive, then took over control from that point
for the trip over the line? Much better than peering under or around
scenery and throwing toggle switches to bring your train out or into the
staging yard yourself.
I'd build one myself except I'm electronically challenged.
Personally, I feel this is far more important to realist operation than
perhaps the use of DCC is.
Cheers
Roger T.
Robert C. Schwartz
two engines and the stopping section works... perhaps the momentum + the
power on the second engine carries the units into the section.
Bob
Clark Martin <cla...@pacbell.net> wrote in article
<clarkm-0505...@banshee.martin.home>...
> In article <7gpopi$16n$2...@nntp8.atl.mindspring.net>, duh...@pipeline.com
wrote:
>
> >>Why didn't I think of that??? Putting a LED in series with a stopping
> >>section's power eliminates the need for an optical detector in my
staging
> >>yard design. Thanks; this news group is great!
> >
> >>Also note - the stopping section-LED approach is applicable to DC or
DCC
> >>systems.
> >
> > Another simple option is gap one rail and bridge the gap with a
> >diode, when the wheels cross the gap they loose power, when you revers
> >the direction control power can flow through the diode and the train
> >can back out of the track. You loose the detection option, but I don't
> >you'd find a simpler circuit.
>
Anthony New
>........
> I really don't see why an operator should -have- to drive trains into and
> out of the staging yard as some people have advocated. Wouldn't it be much
> more realistic if the operator waited down by the depot at the crew change
> point for his/her train to arrive, then took over control from that point
> for the trip over the line? Much better than peering under or around
> scenery and throwing toggle switches to bring your train out or into the
> staging yard yourself.
>
I agree completely. The layout I'm building has automated hidden storage
loops which can accept trains and send them to empty loops, and send
them out again in random order when the human operator in the next
section has accepted them.
A very simple version of this can be made by putting a trip switch or
detector at the far end of each loop which changes the entry and exit
points to the next loop (in a fixed relationship). You can select the
exit points/route yourself manually if you prefer, but it's dead simple
to build!
Roger Amos (in the UK) has written several excellent books on model
railway electronics covering this and other topics - they may be worth a
look even if you do fear the thought of electrons wizzing round your
head.
Anthony.
Roger L. Traviss
Anthony New wrote:
> Roger Amos (in the UK) has written several excellent books on model
> railway electronics covering this and other topics - they may be worth a
> look even if you do fear the thought of electrons wizzing round your
> head.
Thanks Anthony I'll see if I can track some of his books down.
Cheers
Roger T.
daForce
turning balloons should be automated too, esp hidden ones, or
nonobvious balloons.
a simple set of sensors and controls and a uP or PC based bit of
software would make it possible to seamlessly drive into the balloon
and drive straight on out without stopping or flipping any switches.
It does require a cerain amount of computer integration though,
without integration the train would be required to stop, OR
the operator would be required to flip his direction switch at the
right time (when the loco is in a specified block on the loop).
esp doable in DCC systems, with no up, just 2 opto sensors,s om
logic, and a relay. details left to the reader
David J. Starr
loops). With a four transistor circuit I can tell if the reverse loop
polarity matches the mainline and light a red/green LED. I have one
such LED at each entrance to the reverse loop. If the LED shows green
you can enter the loop, if it shows red you have to flip the polarity of
the reverse loop before the train can enter it. I find these two LED's
by themselves make my reverse loop operation easier.
For the next step in automation, you could sense turnout position, so
the circuitry could tell which end of the reverse loop the train was
about to enter. Then the polarity detector circuit's output could be
used to drive a polarity reversing relay so the train could successfully
enter the reverse loop automatically.
Once in the reverse loop you now have to get out again. You can either
reverse the main line polarity or stop the train in the reverse loop and
change reverse loop polarity. You can have a main line polarity relay
that toggles to match the reverse loop when you have a train in the
reverse loop. I don't know of a commercial ready-to-run source for
something like this.
David Starr
daForce wrote:
>
> as a further idea in this direction, I've long held the idea that
> turning balloons should be automated too, esp hidden ones, or
> nonobvious balloons.
>
Robert C. Schwartz
One ground rule I made was that any entry to a loop would be in the same
direction. Worked fine.
Now, I have two loops operating under control of CTI's software (TCL) and
am converting to Railroad & Co. software. Still using the same consistent
entry rule. The other two loops are still automated as above and support DC
analog and DCC.
The main change I foresee, when I go to 100% DCC, would be simplification
of the power distribution. Now, I switch the polarity on the main lines to
keep up with the loop's polarity. With DCC, I should be able to switch
polarity only within the reverse loop and leave the mains alone.
73, Bob
daForce <daf...@iwr.ru.ac.za> wrote in article
<8DC665931...@news.ru.ac.za>...
> as a further idea in this direction, I've long held the idea that
> turning balloons should be automated too, esp hidden ones, or
> nonobvious balloons.
>
><SNIP>
David Holliday
David
"David J. Starr" wrote:
>
> I have one part of automatic "turning balloons" (we call 'em reverse
> loops). With a four transistor circuit I can tell if the reverse loop
> polarity matches the mainline and light a red/green LED. I have one
> such LED at each entrance to the reverse loop. If the LED shows green
> you can enter the loop, if it shows red you have to flip the polarity of
> the reverse loop before the train can enter it. I find these two LED's
> by themselves make my reverse loop operation easier.
> For the next step in automation, you could sense turnout position, so
> the circuitry could tell which end of the reverse loop the train was
> about to enter. Then the polarity detector circuit's output could be
> used to drive a polarity reversing relay so the train could successfully
> enter the reverse loop automatically.
> Once in the reverse loop you now have to get out again. You can either
> reverse the main line polarity or stop the train in the reverse loop and
> change reverse loop polarity. You can have a main line polarity relay
> that toggles to match the reverse loop when you have a train in the
> reverse loop. I don't know of a commercial ready-to-run source for
> something like this.
>
>
> David Starr
>
> daForce wrote:
> >
> > as a further idea in this direction, I've long held the idea that
> > turning balloons should be automated too, esp hidden ones, or
> > nonobvious balloons.
> >