doctorwayne Doughless For me? It would be fun to watch and run. Not to design and build. While I don't have the knowledge or expertise to build such a system, if I did, I'd think the design and build would be the fun and interesting part. Watching it run would be boring, like a loop in a department store window, at Christmas. Wayne
Doughless For me? It would be fun to watch and run. Not to design and build.
While I don't have the knowledge or expertise to build such a system, if I did, I'd think the design and build would be the fun and interesting part. Watching it run would be boring, like a loop in a department store window, at Christmas.
Wayne
Boring, like a loop in a department store window at Christmas, only if it were pre-programmed out of the box to only run one way. But, ideally, the end product would be designed in such a way that the user could program a myriad number of different scenarios to suit the modeler's operational requirements. Every programmed automation would be a different one. Now, that would be fun to sit back and watch.
Rich
Alton Junction
Great discussion. I can see that I'm going to get to like you guys.
I need to spend a little time to survey what really is available off-the-shelf. I have such a DIY mindset that I run off designing completely new solutions so they can be optimized for such a project. You know, if you want something done right...
:Bruce
Look up Miniatur Wunderland to see what can be done with automation. I think you'll be suitably impressed.
DCC I don't think is kludgy at all, it's a robust system for controlling the trains. It's kludgy when you stretch it to do things it wasn't designed to do. The full track voltage waveform is the signal as well, unlike older carrier control systems that superimposed a low amplitude control signal on a fixed track voltage - the signal could easily get lost in noise. That add-ons to DCC to send data back the other way are rather kludgy, low signal strength pulses during specific points in the signal waveform. The Digitrax system is proprietary to them, but doesn;t require anythign special in terms of modifying the signal. The Lenz system is part of the NMRA standard but requires that the booster have a 'quiet period' to allow the decoder to signal back to the system, also a relatively weak signal (in both cases, the reply message is generated from stored energy in the decoder). Frankly, I've seen no need for this capability. Sensing occupancy is enough to know where trains are, either to operate signals, which works very well, or to automate motion. The automation system knows what blocks connect to what, and it knows the direction of travel of the train. So if train 1 is in block 1, which connects to block 2 heading west, and train 1 is going west, when block 2 becomes occupied, it's quite reasonable to assume that it's train 1 now in block 2. This simple method has served for automation for years now.
If you need precise spot locations, you need some sort of IR sensors which get triggered exactly when the train hits a certain point.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
richhotrainhow to track individual locos and individual pieces of rolling stock
All of the above conversation is beyond me. However, reading an article in one of the other model railroad magazines there is an article about someone who has an HO railroad and installs an RFID device on the end of each car along with readers that communicate with a computer.
You can Google Seth Neumann Model Railroad Control Systems.
maxmaninstalls an RFID device on the end of each car along with readers that communicate with a computer.
that allows you to identify a car and it's sequence in a train as it passes a reader.
it doesn't help you precisely locate a car on an industrial spur or on yard track to drop-off a car or know when the end of a train is about to couple to pick-up a car.
greg - Philadelphia & Reading / Reading
gregc maxman installs an RFID device on the end of each car along with readers that communicate with a computer. that allows you to identify a car and it's sequence in a train as it passes a reader. it doesn't help you precisely locate a car on an industrial spur or on yard track to drop-off a car or know when the end of a train is about to couple to pick-up a car.
maxman installs an RFID device on the end of each car along with readers that communicate with a computer.
I agree. I was only addressing the part about identifying a car in a train.
Yeah I think the only way to get the precise location is to trianglate the tag on the layout. Probably need resolution down to just a few centimeters, um, like an inch or two. So the layout is one big reader.
RFID is not a bad idea since the tags are relatively inexpensive and easy to affix and hide on an engine or car. Any kind of active transmitter would be too expensive. Might be worth the expense for a one-off complex automation like we've discussed but it would generally not be accepted by you folk. I don't know. Maybe we can produce the necessary beacon inexpensively in quantity. Every car then would have to pick up energy from the track bus.
I bet the FCC just loves this DCC! It swings square wave sharp signals from +18V to -18V and the layout has to be an excellent radiator. Fast rise and fall times with sharp corners like that carry a ton of harmonic frequencies. Haha. I wonder?
To do a great freight yard simulation we would have to very precisely move the engines and have really good and reliable coupling and decoupling. You would have to be able to do that practically anywhere on the freight yard or sidings and not just over a decoupling track (my c1955 Lionel set has/had one). Cars would need brakes too. I think all doable for a one-off custom simulation but not for everyone.
All the commercial vendors of DCC have their systems approved (at least those being sold in the US) per FCC Part 15 rules. There are several DC power systems that use PWM, so the same thing, high amplitude square waves with varying pulse width applied to the rails. There are a milliooon and one things dooing the same thing in everyday life - in today's digital workld, it's the easiest way to have a digital device control an analog one, such as varying the speed of the fan in your computer.
The sharp edge issue can be easily mitigated while still allowing the wavefoorm to remain within spec - there's ALWAYS a rise and fall time even under the best of circuitnstances and while it's easiest to represent the waveform as nice perfect squares, the standards don't require the impossible. Some designs seem to do a better job than others, in that in some situations you can get ringing on long bus runs and so snubber circuits are recommended, but this is to keep the waveform clean, not to mitigate RF interference.
Might be interesting to test it out, but a spectrum analyzer is one piece of test gear I don't have. Don't really have a need fooor one, and since the electronics is really just anoother hobby of mine, I can't really justify it (although 3 bench meters, 2 power supplies, and 4 handheld multimeters (not counting the unknown number of cheap HF ones I have) may argue against that.
rrinker The sharp edge issue can be easily mitigated while still allowing the wavefoorm to remain within spec
when i worked on 4G, they were allowed to operate on spectrum that allowed a max power > adjacent spectrum. They needed to limit the power when operating on the edge of the spectrum to get the interference down below that allowed in the adjacent low power spec.
considering that the fastest DCC changes polarity is 50 usec, a period of 100 usec, a freq of 10 kHz and harmonics above that, not sure what the FCC interference spec is below 100 kHz.
AM radio starts around ~143kHz. Lowest cordless phone is 43 MHz. some 5G will operate above 28 gHz but must be line of site. Lots of labs need new higher freq spectrum analyzers.
There may be some opportunities - most of the equipment I have is second hand so as it is. Only my scope, logic analyzer, and 2 of my handheld meters are brand new. I wouldn;t say no if a decently priced used spectrum analyzer showed up for sale. Over on the EEVBlog Forums, they call it TEA - Test Equipment Aquisition syndrome. I'm not actively looking for anything else, I have more than enough to do what I ever need to do. I barely use my scope so as it is, the el-cheapo USB logic analyzer, which is plenty fast enough for the stuff I'm doing, gets more use monitoring and decoding the data between devices I am connecting.
Only radio prooblem I've ever seen with DCC is at a club show ehen someone set up the microwave with the side opposite the electronics (the side that radiates the most RF energy as basically everything generated by the magnatron that doesn;t get reflected down at the top of the chamber shoots out that side) pointed right at the radio receiver for the wireless DCC throttles. I mentioned to someone that this was not a goood idea. But it was left there. Everything worked great - until someone went to heat up a cup of coffee they had and suddenly everyone using the 2.4GHz throttles lost control, The old 916MHz ones were ok. They turned the microwave sideways and no more problems.
I used to listen to AM talk radio while working on my old layout, only got interference fromt he DCC system if I held the radio right over the tracks, or a running loco. With the radio on the workbench on the opposite wall as the layout(probably 6-10 feet from the closest bit of track), no problems. At least not with the fairly local stations I listened to, I never tried to pull in weaker ones - nearly pointless as I was almost completely undergroound in the basement, the back wall had about 5 feet exposed but the side and front only had a foot of wall above the grade outside. I'd say despite the POTENTIAL for problems, in reality the manufacturers have this well in control. There seems to be a large difference between the two biggest brands when it comes to what happens when you hook a whole bunch of neatly parallel conductors to the systems. Based on scope data though, both tend to end up rounding off the sharp edges even more, thus reducing the harmonic effect from the sharp transition.
I am also very interested in the concept of a fully automatic and functioning layout. Let us know how that goes for you in the future...
I'm beginning to realize that Windows 10 and sound decoders have a lot in common. There are so many things you have to change in order to get them to work the way you want.
BNSF UP and others modeler I am also very interested in the concept of a fully automatic and functioning layout. Let us know how that goes for you in the future...
Well, I wish Bruce well on his projects, but total automation does not do much for me.
For display operation I'm very happy with some carefully planned dedicated loops. Otherwise I like for each train to actually have a "human" at the controls.
But everyone should do what works for them.
Sheldon
I've been "playing with trains" for well over 65 years.... A huge part of the attractiveness is that I am controlling/running the trains.
Ha, I would welcome "automated" trains as much as I would welcome "automated" autos.
My hobby is model railroading - not electronics or robotics or automation.
ENJOY !
Mobilman44
Living in southeast Texas, formerly modeling the "postwar" Santa Fe and Illinois Central
SOME automation can add to the fun. Friend's layout, the actual (steam, turn of the 20th century) railroad was not automated. At all. However, in his city, he had a group of trolleys that would automatically circulate and stop. Made for some nice animation without tying up 4 operators doing a very boring task of running a trolley around a loop, stopping at designated stops, and making sure not to run into the one ahead of them.
All this fancy talk of complex detection - not needed. All that was used here was the loop was broken into 16 blocks, with a 16 channel block detector, and RR&Co software reading the blocl detectors. That's all that was needed to start up and run the 4 trolleys, have them stop at certain points, and never run in to one another. As I mentioned earlier, if the starting position is known, as well as the possible path, the system knows which loco (trolley) is in which block the entire time, there is no need for some sort of feedback or anything like that. Yes, you can mess it up by picking up one of them and putting it back down out of order. But if the point is a hands off automated operation, why would you do that?
Randy, you are spot on!
Having a trolley run a loop or having crossing signals go off when a train approaches, etc., is my idea of automation for a train layout. You said it well.
mobilman44 Randy, you are spot on! Having a trolley run a loop or having crossing signals go off when a train approaches, etc., is my idea of automation for a train layout. You said it well.
Same here, I'm planning an automated trolley system, crossing gates, etc.
And in my view, any good signal system for a model railroad should be semi automated to simplify some of the dispatching tasks. I know some operations purists will not agree, but our compressed distances make full scale CTC to cumbersome in my view.
I can set my layout into a simple display mode with four dedicated loops. And the signals will then require no human input, but will continue to show proper indications as the trains travel around their respective loops.
And my whole signal system works automatically in CTC mode with a dispatcher or tower operation mode with no dispatcher on duty.
But with a dispatcher, signals do not clear until the dispatcher fully clears and assigns the route.
That's enought automation for me.
I don't even bother with working locomotive headlights. Automation is not for me.
.
-Kevin
Living the dream.