I am determined to build a “N” scale historical SP freight yard known as the “College Park”, freight yard. In it, there were 81 switches weaving in and out of 14 lines for freight cars. There was one line for a cattle yard. There were two lines for fueling with oil and water for the steam engines. There were two lines leading into the round house. Being this yard was in San Jose, California it was the hub between San Fransico, Sacramento, and Los Angeles with all the ranches and farms inbetween. They use to load up to 60,000 freight cars per year or about 100 to 160 railroad cars per day, with six yard goats, a round table and 81 switches DCC is the only way to build this layout. I have most of the know-how to do it except knowing how to sense the rolling stock anywhere on the layout. Anyone out there working the same issue? Sincerely, Yardboss
I’ve read the question several times. What do you mean by “sense the rolling stock anywhere on the layout?”
If the need is only to know that there is rolling stock present on a certain piece of track, then what is needed is called a block occupancy detector. These come in several types that sense either current flowing to the track (like a Twin-T circuit), or they sense a shadow over the track (photo-optical). There would have to be one block occupancy detector for each section of track that needs unique detection. For the current sensing type each freight car has to be fitted with metal wheels and a resistor so that it draws the desired amount of current. Be careful here in the calculations because even if each car draws a tiny amount of electrical current, a yard of 100s of cars can draw a lot. When I calculated this for our club’s main yard, which appears to be much smaller than the one you are talking about, the cars just sitting in the yard would have drawn 3.5 Amps. We decided to “sense” only 1/10 of the fleet rathern than each car. That means if only five cars are sitting on a track we have only a 50% chance of detecting them.
On the other hand if you need to know exactly where an exact car is then there are several more options available. I know one club who put a bar code on the bottom of each car. Key points on the layout have a bar code reader imbedded in the track that reports its position to the computer. With one of these on the entrance to each yard track it would indicate which track the car was on but not where on the track the car was. The more finely a location needs to be isoloated the more readers would be needed.
Another option I’ve heard people talking about but don’t know anyone who has implemented is the electromagnetic (RFID?) tags like are used for thieft prevention in the stores. Each car gets its unique passive transponder and the readers are set up according to how fine a position is needed
Sorry to take so long to get back to you. After getting your answer I decided to research this issue more to learn more terminology.
What I mean by “sense” is to know where a specific freight car is located in the freight yard. My layout is the freight yard. What I have researched out since I posed the question is that there are solutions such as resistor wheels, transponder decoders, bar code readers, 2D digital code and RFID systems. I have also found two computer systems that have been used in the past, Winlock by Dr. Hans Tanner of Digitoy Systems and rnrdigital.com that seems to be newer, can both used to control freight yards.
I am looking to “sense”, the type of car via ID number. Whether it is a box, stock, gondola, flat, tank, or specialty car, via on board sensor via computer or optical mark sensed by an under track reader via computer, so I can send an engine (one of six yard goats) to pick up a specific car from sixteen yard tracks all via computer.
I believe that I can do this with a transponder decoder. However, I have found that each car would need wheel wipers. They are no longer made. I called four model train manufactures and not one makes these parts. I thought that I could get both six and four wheel truck wheel wipers and I was unsuccessful. I tried to find used wheel wipers at the model train shows but no one would sell them and most did not have any. I am now looking to manufacture them from shim stock via a progressive die or chemical machining with a bending die.
That is where I am at presently.
Sincerely,
Milton Chris Carris (yard boss)
Sorry to take so long to get back to you. After getting your answer I decided to research this issue more to learn more terminology.
What I mean by “sense” is to know where a specific freight car is located in the freight yard. My layout is the freight yard. What I have researched out since I posed the question is that there are solutions such as resistor wheels (you mentioned), transponder decoders, bar code readers (you mentioned) , 2D digital code and RFID systems (readers have not been standardized) . I have also found two computer systems that have been used in the past, Winlock by Dr. Hans Tanner of Digitoy Systems and rnrdigital.com that seems to be newer, can both used to control freight yards.
I am looking to “sense”, the type of car via ID number. Whether it is a box, stock, gondola, flat, tank, or specialty car, via on board sensor via computer or optical mark sensed by an under track reader via computer, so I can send an engine (one of six yard goats) to pick up a specific car from sixteen yard tracks all via computer.
I believe that I can do this with a transponder decoder. However, I have found that each car would need wheel wipers. They are no longer made. I called four model train manufactures and not one makes these parts. I thought that I could get both six and four wheel truck wheel wipers and I was unsuccessful. I tried to find used wheel wipers at the model train shows but no one would sell them, if they had them and most did not have any. I am now looking to manufacture them from shim stock via a progressive die or chemical machining with a bending die.
That is where I am at presently.
Sincerely,
Milton Chris Carris (yard boss)
Umm - in addition to having sensors which can identify cars as they go into a track, you also would need a computer which would keep track of which cars have been put on which tracks and in what order.
I realize that you seem to want to automatically control the whole yard by computer, and automatically do some fairly un-prototypical “cherry picking” - picking out single cars from anywhere in the yard, instead of pulling a whole track to classify cars.
But have you considered the obvious, fairly low tech and totally prototypical (depending on the era you model) option of just making a handwritten list of what goes into the track as it gets put into the track ?
Using chalk and a blackboard would also make that list fairly easy to modify without leaving tons of paper around.
If you need to get an overview of what is in a track, you could do the totally prototypical thing of stationing your yard clerk at the end of the track and have a yard goat pull the cut of cars slowly past, as you note down the cars on your sheet or black board.
Two factors to consider:
-
What prototypical mechanism are you trying to simulate? Part of that question is era, location, theme and what subset of reality you are trying to simulate.
-
What modeling mechanism are you using to model that effect ? The same effect can often be modeled in many different ways - it just takes a little imagination.
Smile,
Stein
You really don’t need to go through all the trouble to set up a PICL (Perpetual Inventory Car Location) system. In the yard, the big guys do it all in the computer system. We issue a computer switchlist, and complete it as the crew completes the work. The computer keeps track of where the cars are and in what order. There is off the shelf MR software to accomplish this.
Nick
Actually RFID chips are very tiny - not much larger than one or two of the letters you are reading right now. If you ever get your pet microchipped, the RFID chip is no larger than a grain of rice. The large tags on the store items are oversized to make them easily seen and placed. And they can be thinner than a credit card; if you live in a city that uses a contactless “Smart Card” system for public transit, those use RFID technology. So RFID would be very practical even in Z scale.
Thing is, the components are going to be pricey, at least at this time.It would be interesting to tie RFID technology into DCC…I’m sure that time will come.
You do realize this isn’t prototypical. A real railroad yard can’t scan all the tracks in the yard to read what’s in each track, even though each prototype car is equipped with an RFID chip.
Yes, technology has advanced quite a bit in the 3 years since the original question was posted. Since that time a lot of work has been done on the NMRA Recommended Practice RP-9.3.1 and RP-9.3.2. The problem with those solutions is a decoder in every single piece of rolling stock. For a layout the size you are talking about, a solution of the caliber you describe along the decoder concept is going to be very costly.
Or axle wipers. All you need is for one wheel on each axle to conduct electricity to the axle. Then you put a wiper on the axle to pick up the power to the decoder. These can be made very very simply from brass spring wire. I’m going down to see if I can find a unit so equipped and take a picture of it for you.
Basically two pieces of some fine brass spring wire is connected through the truck. I often just drill holes perpendicular to the axle and put the wire through. The holes should positioned so the wires would then run directly through the center of the axle. Since they can’t they have to bend to one side. This applies pressure and keeps them in contact to conduct electricity. I solder a connector piece of brass between the two axle wipers. This gives me one point to collect the power and prevents the wipers from twisting. Put two wheel sets insulated on the same side in each truck and make each truck opposite the other so one truck picks up from one rai
Yeah, today they are. But back when I originally wrote that post the smallest I knew of was about a one inch square - most were larger. Today’s are amazingly small. Could very easily fit in an N or Z scale box car.
I have a question for the OP (and anyone else).
Why?
By which I mean… If you go to all the trouble of building what sounds like a massive layout, getting all the stock and locos… then why run it all on a computer screen? If you don’t want to play Engineer, yard Master and all the rest while actually using a prototypical eyeball… then wouldn’t it be easier to just get some digiotal/virtual simulator/game and do it all on screen?
I’m not being nasty. I’m just intrigued.
That’s not the way real flat-switched yards work. Real railroads don’t cherry-pick a single car out of a yard track, except very rarely in a line-stop situation.
Although I am not familiar with the entire history of College Park Yard, I don’t know if there was ever a time that there were six yard engines simultaneously active there, either before or after the 1930s line relocation project.
What era are you modeling? In many eras, I believe local and through trains did much of their own work at College Park.
Seems like an amazing amount of effort to create something that doesn’t operate like the real thing did, but perhaps you have different goals for the project than realism and recreating the prototype.
… which makes my head spin. No wonder I don’t like yard switching.
I’m curious as to how many people in the world are capable of designing computer software designed for efficient switching on a flat yard. Has it been done?
Mark
Back in the late 50’s, while going to college, I worked for Santa Fe summers in Oklahoma as a yard clerk/mudhop at the yard in Enid Ok. The yard had 16 thru tracks and a rip track.
I walked the tracks with a switch list in one hand, and on the night shift with a lantern recording car numbers on a switch list, or verifying car numbers after the track had been switched. The lists were turned over to the yardmaster who “marked them up” as far as car destinations, gave them to the footboard yardmaster, who took the switch engine crew and went out to do their thing.
That procedure was the common way for many years, even after computers came on the scene. I can testify to that since I was by then working in the Topeka HQ programming the computers. One thing we learned, no computer at that time could be relied on as much as the switch list taken manually, even with bad hand writing.
But, to each his own.
Bob
Real railroads have, some are on there 3rd or 4th generation software.
Classification yards don’t work that way but industrial support yards do. All over the Gulf Coast there are SIT yards (storage in transit), warehouses for carloads of plastics and chemicals. A typical yard might have 500-1000 cars in it and every day 50-200 cars are billed out and 50-200 new cars show up to be held. The outbound cars are more or less ordered “randomly” from the railroad’s perspective.
Similarly any industry support yard where spot cars with different commodities or characteristics can have specific cars ordered in or chosen to be spotted, so the cars have to be 'cherry picked".
But those are not classification yards that build freight trains.
College Park switchers were responsible for a vast terrority, Passenger depot and coach yard, 4th street yard-Santa Clara, Newhall yard, the Los Gatos branch and westside industrial districts. On average a dozen switchers comprising of 0-6-0’s, 4-6-0’s and 2-8-0’s were maintained at College Park.
All major classification was performed at Bayshore-thus single car moves would not be uncommon at these locations.
Dave
While the chips in the cars may be that small, the real problem may be going to be reading them down to the small area you may be looking for. If you are just wanting to know whether they are in the yard or not and the yard is of sufficient size, then that will likely work. But if the desire is to know which track the RFID chip is on or to break up the area into small blocks and know which block it is in, then there may be a problem with the number of wireless units, their coverage area and such to get to that level of precision.
My point exactly. Six engines working simultaneously doing classification switching within College Park (as the original poster proposes – he calls them “yard goats”) seems unlikely. Even when a number of crews were on duty at the same time, they would be spread out over a significant geography, not all doing classification switching in the yard.
After talking with a number of real-life conductors and engineers who worked at various times over the last forty years, I’ll stick by my point that it is very rare to cherry-pick out a single car from a flat-switched classification yard track. To build a whole train by cherry-picking single cars out of a dozen or so flat-switched classification yard tracks is extremely unlikely – certainly not routine.