Okay here's the deal.
I posted this in General with no responses, so I figured the geek squad was a better fit.
I'm combining an oval helix with staging. I need about 2 1/2 turns and 2% grade. The helix will be 3 tracks with two being staging. I intend that there be room for eight trains, each with their own siding that starts on one straight and ends on the next. So basically each staging siding will be a semi circle.
Obviously, If staging is full, one train has to leave before the other can take the siding.
The whole thing is in another room from the layout and I'd like to be able to work it from the layout.
So what is the best way to manage the staging? Go as low tech or high tech as you want. Cost is always a factor, but value is as well. I have spare Pentium IIIs sitting around. I also have an interface with loconet, if that helps.
Chip
Building the Rock Ridge Railroad with the slowest construction crew west of the Pecos.
USB camera...
BDL168, block detection, led read out on a panel. You connect your led were the tester goes, run the wires to a panel.
SE8C and PanelPro. You can make the signal do what you require and you get 8 turnout control, block detection, manual turnout control (or computer via PanelPro).
UpNorth wrote: USB camera...
Cameras. I can't picture being able to see both sides of the loop from one location.
Okay, I get I need block detection and a readout. But that's it. I'm not familiar with these parts. Where can I find out more? I'm just starting this process.
I get that PanelPro is software, but what are we talking about here with the SE8C. Is this a DCC decoder?
One thing I should mention is that the layout is set in 1909 and a computer will look out of place. This is not to say that I won't use one, but there will be no space above the layout because the trees will go within inches of the ceiling. It will have to go low, so figure 36" to the top of the monitor.
Jim, Modeling the Kansas City Southern Lines in HO scale.
jamnest wrote:Sounds like you have a Digitrax system with a computer interface. For information on Digitrax block detection devices you can download the manuals free from the Digitrax web site. The block detection can communicate via loconet with panel pro or you can place indicator lights on the facia of your layout.
Jim,
I do have Digitrax, but is I understand it, I don't have to be married to it if there is something better. Right now, I have no plans for a control panel, although it could certainly be added. I'm planning to use caboose switches on the layout, but I figure powered switches in staging would be appropriate.
I'm really looking for a system of control, more than specifics, but I also understand it may be the specifics that allow a system to be created.
For instance, at this point, if I'm going to use LEDs on a panel, wouldn't reed switches be a cheaper alternative for block detection? Why do I need electronic block detection? What makes it superior?
I would want at least one video camera bringing me a feed from the staging area to the layout control area. Depending on the circumstances more than one camera angle or a remote controlled panning camera might be helpful. I think I would use a spy cam etc. rather than USB as UpNorth mentioned. A six inch LCD would be fine for a display I imagine. If you have to use a PC as a display then maybe USB is the way you would have to go.
PanelPro
SE8C stationary decoder
BDL168
SpaceMouse wrote:Cameras. I can't picture being able to see both sides of the loop from one location.One thing I should mention is that the layout is set in 1909 and a computer will look out of place. This is not to say that I won't use one, but there will be no space above the layout because the trees will go within inches of the ceiling. It will have to go low, so figure 36" to the top of the monitor.
I agree. That's why I mention using more than one camera or a remote panning mount for the camera to increase the view. I suggest the LCD display because they are compact and getting cheap. Using PanelPro does require a computer. If space is a real issue maybe it's not the answer?
SpaceMouse wrote:if I'm going to use LEDs on a panel, wouldn't reed switches be a cheaper alternative for block detection? Why do I need electronic block detection? What makes it superior?
Reeds will work fine too, IMo. Electronic detection employing infra red or photosensitive technology is superior in that you are actually detecting the objects passing the detection point. No magnet is required to trip a reed switch. There may be other reasons...?
SpaceMouse wrote: jamnest wrote:Sounds like you have a Digitrax system with a computer interface. For information on Digitrax block detection devices you can download the manuals free from the Digitrax web site. The block detection can communicate via loconet with panel pro or you can place indicator lights on the facia of your layout.Jim,I do have Digitrax, but is I understand it, I don't have to be married to it if there is something better. Right now, I have no plans for a control panel, although it could certainly be added. I'm planning to use caboose switches on the layout, but I figure powered switches in staging would be appropriate. I'm really looking for a system of control, more than specifics, but I also understand it may be the specifics that allow a system to be created. For instance, at this point, if I'm going to use LEDs on a panel, wouldn't reed switches be a cheaper alternative for block detection? Why do I need electronic block detection? What makes it superior?
Using your Digitrax Loconet and panel pro with a Digitrax block detector may be less expensive in the long run. I have a Digitrax Chief and use Decoder Pro but I have not moved to block detection, yet. My understanding is the Digitrax block detector may be all that you need. It would communicate back to your computer via Loconet via panel pro. That is why I suggested you check out the block detector instructions.manual on the Digitrax web site.
BlueHillsCPR wrote: PanelPro SE8C stationary decoderBDL168
Holy crap! $400 to control my 8 sidings and that only counts 2-SE8Cs and a BDL168.
Breathe in-Breathe out--Breathe in--Breathe Out.
I'm okay now.
Best reason I can see for block detection circuit instead of simple switches, reed or optical, would be the variety in sizes of trains staged. detection circuits can tell you if the storage track is occupied and when it clears or fouls the switch ( two separate detection circuits or control points ) If I understand your description ( by browser at work will not let me see graphics) you will have 2 "sidings" on each curve of the loop making it 3 tracks wide in the turn of the helix, then it would be handy to know when the inside switch was not fouled so that a train on the outsid track could get by it, so detection circuits would be needed for each switch.
indicator lights (LED's) or computer screen color change per block signal is just simple wireing (granted verry repetitive, but do able) Great idea though!
Earl
I once caught a train in my pajama's. How it got in my pajama's I'll never know... (sorry, Groucho)
SpaceMouse wrote: BlueHillsCPR wrote: PanelPro SE8C stationary decoderBDL168Holy crap! $400 to control my 8 sidings and that only counts 2-SE8Cs and a BDL168.Breathe in-Breathe out--Breathe in--Breathe Out.I'm okay now.
Yeah, that's about what I thought too. Once the house is paid off and the kids are through college...then I could afford it, if I wasn't set to retire...
If you want track occupancy detection only, then you should only need the BDL168 for detection. The online discount DCC vendors list them for about $120. The BDL168 has AUX terminals for LEDS which can be placed on a facia or control panel to indicate if a staging track is free or occupied. I have not used the BDL168 but have read over the instructions. I think it would be worth your time to take a look at the instructions. It seems far more simple than relays, reed switches or IR detectors.
SpaceMouse wrote:Holy crap! $400 to control my 8 sidings and that only counts 2-SE8Cs and a BDL168.Breathe in-Breathe out--Breathe in--Breathe Out.I'm okay now.
Substitute cheap labor for electronics. Stick an operator in the helix room, feed him pizza and beverage of his/your choice. Now figure how many operating sessions it will take to equal the cost of electronics. And that's not counting your time figuring out how to install it, time to maintain it, and time to document it so you or somebody else can fix it at some later date. When labor is cheap, human beings are often the better choice. Besides, you can blame somebody else instead of electronics when things go wrong!
That said, reed switches require magnets passing near the contacts to actuate them. Which means magnets mounted on your 1909-era locomotives and caboose. Think about the implications of that. Then breathe in and breathe out again. Choose electronic detection or choose humans for the cleanest solution requiring least modification to your rolling stock.
my thoughts, your choices
Fred W
Fred,
My friend that uses the reed switches uses nothing more than the magnets in the loco motor.
SpaceMouse wrote:Fred,My friend that uses the reed switches uses nothing more than the magnets in the loco motor.
You are challenging the engineer in me.
It can certainly be done as your friend suggests. But there are trade-offs between sensitivity and false alarm (detection) rates. If your friend's system is sensitive enough to consistently pick up a small can motor magnetic field from below the track or beside the track, does it always ignore a locomotive on an adjacent track? Do any of your BB cars have steel weights that have become low strength magnets over the years?
If I were engineering a reed contact system for detection, I would go the other direction - low sensitivity contacts triggered by very small, powerful, strategically located rare earth magnets on locomotives and desired cars. The magnets would be located where they would be as close to the reed contact as possible while still being hidden from view.
Reed contact sensitivity is determined by the strength of the bias magnetic field. The stronger the bias magnet, the more sensitive the contact. A low sensitivity system triggered by strong impulses is going to be more reliable (less false positives and negatives) than others.
In my first post, I'd forgotten about today's readily available and inexpensive rare earth magnets. I was envisioning installing Kadee uncoupler magnets on your rolling stock to trip the contacts. With rare earth magnets, a magnetic reed contact detection system becomes a viable, inexpensive alternative, relatively easy to design, implement, and reconfigure when changes come along. And the good news is that the magnetic reed detection would be totally independent of your control system.
just my thoughts with my engineer cap on...
I know, it's not cheap. But what fun it is explaining the benefits reaped to your SO, just imagine...
You can run myltyple webcams using a 50$ software package called CamPanel. Probably something similar is availlable for less. http://www.trackercam.com/TCamWeb/download.htm
The SE8C can handle 8 turnouts (stall type only), can be used with DB-4 (Digitrax) block detection (4 blocks) and gives you signaling via your computer and PanelPRo. If that computer is quick enought, it probaby can handle the 2 or 3 webcams and the PanelPro. The web cams don't need high frame rate or high resolution so you can probably get away with less OMPHH from the computer.
Or you can go with the Team Digital's SIC24/DBD22 combo and just have block detection and signals. No need for the computer here. You don't have remote turnout control yet thow. You could also simply use the DB4 or DBD22 alone to detect and turn on some led indicator.
http://www.teamdigital1.com/prod_catalogue/dbd22_product/assets/dbd22_doc.pdf
The electronic detection only requires and engine present and the last car to have resistor wheel sets, caboose. This way you know for sure the last car is out. Reed and magnet is fiddling.
And if you use magnetic uncouplers under track you'll know were they are for sure, the floating car is a give away.
Are you talking me out of using reed switches or talking me into it?
UpN,
My computer is around 400 Mh, but I still run 98 on it. For a plugged nickel I'd convert it to Linux if JMRI would run on it.
SpaceMouse wrote:Fred,Are you talking me out of using reed switches or talking me into it?
The more I think about it, the more I like reed switches for a non-massive home layout, especially for a non-engineer. The remaining drawback to reed switches is that they only indicate a triggering locomotive or caboose passing by a point. They do not indicate track occupancy between the sensor points - not necessarily a meaningless distinction. The Free-mo boys who are big into signaling (not me, my layout is and will be dark territory) emphasize both track and passage detection - neither is sufficient by itself for their scenarios. They can't use reed switches because they don't control the rolling stock to be used on the layout - infrared becomes the preferred solution to point passage detection.
The other point in favor of reed switches is that you already understand them. The logic of simple contacts is pretty easy to hard wire, rather than having to program after installing programmable devices. It is important that you understand your detection and control system because it would be really difficult for somebody else to come in and understand what you did. So you get to design and maintain whatever system you eventually settle upon.
I get too many of these headaches at work (trying to figure out why somebody else did what they did so it can be modified properly, and of course the originator is nowhere to be found), so I like things very simple at home.
Option one:
Lets talk a little about the Digitrax SE8C. (I'm currently designing the signaling for our club layout.) One SE8C will control 32 signal heads. They can be used to control individual LED's on a panel if you want, instead of the signals. Also, the same SE8C can detect 8 blocks by adding two BD4's to it. The same SE8C will also control 8 Tortoise machines with no modification. JMRI panel pro is free, but has a learning curve. Download the manual and check it out.
Option two:
There is a detection ckt available from an eBay seller that costs about $7 each. It is self contained and has an LED output, and you can add a remote LED to it. (It requires a power supply for power.) Detection is done by reflecting an IR signal from someting (usually a train) and this is received by the IR sensor. This is a pre-built, one piece unit that you would just have to mount somewhere close to the track or under it. It doesn't reflect from a black surface all that well. (I am using one myself.) So you would be talking about eight of these units minimum, one for each train. (This would be the simplest one.)
Option three:
CTI makes a Train Brain system that provides automatic train control. Depending on what you purchase, you could automate your entire layout, or just a few blocks. (I am using that system also.) The starter set costs about $100 for reading 4 sensors and switching 4 relays. A computer is used, and their program can interface with most major DCC systems. I think Digitrax is one of them. You would need at least one other standard TB board to get 8 sensor inputs. Then you would need to setup the sensors and wire them to the boards. I use CSD cells on the track, which requires a light above them. You will have to write some code, but it is fairly simple.
Option four:
Cameras. Train cameras are not that expensive these days, and no layout or track connections are needed. (They don't need to be mounted to a train.) Plus, you would be able to actually "see" what is happening. You should be able to get some sort of video switch that will allow you to use one TV and switch between cameras.
OK, I think I vote for cameras for your situation.
Elmer.
The above is my opinion, from an active and experienced Model Railroader in N scale and HO since 1961.
(Modeling Freelance, Eastern US, HO scale, in 1962, with NCE DCC for locomotive control and a stand alone LocoNet for block detection and signals.) http://waynes-trains.com/ at home, and N scale at the Club.
I had to make this anyway so here it is for clarification.
This is the plan view. The Inside track believe it or not has a 22" radius and the center track has a 20" radius.
SpaceMouse wrote:UpN,My computer is around 400 Mh, but I still run 98 on it. For a plugged nickel I'd convert it to Linux if JMRI would run on it.
JMRI = java based, AND comes in a linux flavor ready for your use from here. Enjoy
-Dan
Builder of Bowser steam! Railimages Site
NeO6874 wrote: SpaceMouse wrote: UpN,My computer is around 400 Mh, but I still run 98 on it. For a plugged nickel I'd convert it to Linux if JMRI would run on it. JMRI = java based, AND comes in a linux flavor ready for your use from here. Enjoy
SpaceMouse wrote: UpN,My computer is around 400 Mh, but I still run 98 on it. For a plugged nickel I'd convert it to Linux if JMRI would run on it.
I thought I remember reading that.
Here's an idea that could simplify wiring and picking the right route in the helix. Asign each staging track a direction. (ie: Inside track East Staging, Middle track West Staging) Then set up the curved turnouts leading to the staging tracks as spring switches automatically lining the train into the appropriate track. Then instead of having to select which staging track to use, you only have to pick which pair of tracks. Selecting a track pair could be automated by using reed switches to sense occupied tracks to automatically throw the mainline turnout to the first available track. It should work but I don't know what type of special wiring it would take, so I can't help you there.
You know, I started this thread with a dream.
In this dream I imagined that I could schedule 8 trains to run automatically. That when one train left the layout, the next train in sequence would start up and travel into my ops session. The train arriving in staging would occupy the track that the departing train just left. The at the correct moment, the next would depart. And I would just switch the incoming trains as needed.
But it dawned on me that although there are circuits to detect a train or loco and tell if it is occupying, a particular block, there is nothing that can detect which loco is in the block.
And it's not like I need to know. But the computer needs to know so it can open the right turnouts and send the train on it's way at the right scheduled time.
Without that, I might as well use reed switches and toggle switches and LEDs and keep track of the locos with a set of car cards hung on the right hook.
Well then you need Railraod &Co. TrainController. it will only set you back 335$.
Have you ever viewed these You Tube before.
http://model-railway.me.uk/Manufacturers/Digitrax/Default.aspx
Go direct to part 3.
gandydancer19 wrote: Option one:... JMRI panel pro is free, but has a learning curve. Download the manual and check it out.
... JMRI panel pro is free, but has a learning curve. Download the manual and check it out.
Only problem with this is their is no manual to download for PanelPro. DecoderPro yes. But don't let that stop you. Their is a link to a clinic by Dick Bronson of RR CIRKITS. I used the clinic and got my panel up and running, block detection, turnout control, simple signal logic control.
http://www.rr-cirkits.com/Clinics/Clinics.html
SpaceMouse wrote:...But it dawned on me that although there are circuits to detect a train or loco and tell if it is occupying, a particular block, there is nothing that can detect which loco is in the block. ...