Back to DC

[Sokramiketes]

Is all this added slow-down complexity just so that you can somehow automatically switch the staging tracks, have a train pull in, and another one come out, all without manual intervention?

Yes, that's the gist of the idea. 

What's the simplest way to cycle trains through reverse loop staging, allowing a parade of trains, without operator input.

Then, what can be done to this simplest approach to ease the rough edges (on/off toggle switch power delivery being rough).

I feel like there would be a diode matrix solution to corresponding track voltage to what tracks are active in staging.  Letting the 9v power be tweaked through the matrix.  And maybe the staging tracks are just stepped down (voltage wise) compared to the mainline, to ease the power delivery without needing a full mapped ramp of applied voltage.

For example, 6V on the mainline, 4V into the staging ladder, and 2-3V on the staging track.  Then power is harshly cut from 2-3V instead of 6V, all with speed zones set by simple diode bridges between the transformer and the track feeder.

By the way, keep throwing it at me from the DCC side too.  I'm not unconvinced about DCC and it's controllability.  But I want to talk it through with a full picture in mind.  Again, DCC two wires is great, until you need to block the layout for signals, and control/animation points, etc.

[DKS]

I stayed silent on this topic until now, but I can't hold back any longer You gave me a good chuckle! You are injecting a processing device and an advanced speed control into the basic pure-DC system.  That is what DCC is (and it is also highly configurable and customizable). The only difference is that your speed control microprocessor and PWM throttle are build-in into each locomotive (for maximum flexibility).  We have come the full circle. But if you want to reinvent the wheel. . .

EDIT: Looks like DKS already addressed this subject and presented his reasoning why it make sense. I still don't see it. Once you start introducing computing devices into a basic DC system you have lost me on how it is a simple old-school solution.  Sorry, I won't buy it.

DCC complicated and overkill?  My "layout" consists of few pieces of Kato Unitrak I set up for testing locos and a $150 NCE power cab all-in-one DCC system.  It plugs into a 120V AC outlet on one side and 2 wires to the track.  Over-complicated?  Really?  Cobbling up a complex relay system with a home-brewed microprocessor speed control and PWM throttle sure seems much more complicated that my ready-to-use NCE PowerCAB.  Call me lazy.  If you are driven by a desire to tinker by yourself then your solution is valid. But don't say that DCC is over-complicated.

Peteski, really now, I'd have thought you were more knowledgeable about basic hardware. You can build what I'm taking about for less than $20: $5 tops for a PWM circuit, and $15 for a programmable hobby module. Now, tell me how much it would cost to accomplish the same thing using off-the-shelf DCC stuff.

So, I'll do what you suggested: I'll call you lazy. 

And I'll also say just what you told me not to: DCC is over-complicated. I don't claim that it's bad or wrong, and in fact I think it may be one of the best things to happen to model railroading since N Scale; but for certain applications, such as the one under discussion, it's like using a PC to turn your desk lamp on and off.
 

[DKS]

I've tried hard to stay out of this...

I stayed silent on this topic until now, but I can't hold back any longer.

I like how the high-tech crowd can only stand just so much DIY stuff before they blow a gasket. We're not all geared the same. I have nothing against DCC, but as I've said, in some cases it's too much technology used to solve relatively simple problems. If DCC has done any harm, it's put a big dent in creativity (not to mention our wallets!). DIY folk thrive on the satisfaction of creating our own solutions, rather than just throw canned technology at a problem. Again, not saying it's wrong; it's just not the Universal Solution some of its more ardent fans make it out to be. And what happens when some of that tech becomes obsolete? Toggles and relays won't be disappearing anytime soon. Just sayin'... 

 

[Cajonpassfan]

I too have tried hard to stay out if this, but the last few exchanges made me realize just how much I could contribute and shine a bright light on this topic, I just can't hold back.
Well, actually, I can
Otto K.,
(heading back to the trainroom)

[DKS]

I too have tried hard to stay out if this, but the last few exchanges made me realize just how much I could contribute and shine a bright light on this topic, I just can't hold back.
Well, actually, I can
Otto K.,
(heading back to the trainroom)

Why hold back? I promise this old unenlightened dinosaur will just bow out of the conversation and let y'all have at it.

[Greg Elmassian]

The "simple problem" seems to get more and more complex.

Run JMRI and use all that free software to do what you want.

Sure, you can buy a $15 computer module and spend years learning programming... so I don't get the "low tech" solution involving people learning programming.

Somewhere "low cost" got swapped in for "low tech" or "low complexity".

And I think the analogy of using a computer to turn on a light is great, but you should use the analogy of using an Alexa to turn on a light, even a 5 year old can do this, because of all the software already developed to make his life easy.

Every time I see a hardwired approach to a specific problem, it's quite often the next thing is "Well, I just want this feature, or to tweak this a bit"... that's when I use a tool designed to solve the problem.  On another forum, there is a guy who suggests doing EVERYTHING with relay logic, and I mean pretty complex stuff... no one has ever successfully even understood his schematics. (Turns out the guy bought a huge amount of 4 pole double throw relays ages ago), he does indeed have a completely automated layout, and when something goes wrong, he is the only person on the planet who can fix it.

Anyway, I'm watching to see what the solution is, and how much effort it will take.

Greg

[peteski]

By the way, keep throwing it at me from the DCC side too.  I'm not unconvinced about DCC and it's controllability.  But I want to talk it through with a full picture in mind.  Again, DCC two wires is great, until you need to block the layout for signals, and control/animation points, etc.

Whoa!  And you don't need to do all those complicated things when you want to add signaling, control/animation points, etc. to a DC setup

DKS, yes, it might be laziness, but to be honest, if a ready-made solution exists I rather spend my creative time on other tasks (instead of designing a Rube Goldberg circuitry, just for the sake of seeing if it can be done).

[alhoop]

  On another forum, there is a guy who suggests doing EVERYTHING with relay logic, and I mean pretty complex stuff... no one has ever successfully even understood his schematics. (Turns out the guy bought a huge amount of 4 pole double throw relays ages ago), he does indeed have a completely automated layout, and when something goes wrong, he is the only person on the planet who can fix it.

Greg

Do you have a link to that forum? Brings back memories of my work on crossbar switches and the relay logic used in the F106 Fire Control System.
All extremely reliable.
Al

[nickelplate759]

I think the working definition of "simple" technology here is closely related to what we individually understand and feel comfortable with. 
So fundamentally, there's no point trying to convince each other that DC wiring is simpler or more complex than DCC - because what's simple for me might not be for you.

[mmagliaro]

I think the working definition of "simple" technology here is closely related to what we individually understand and feel comfortable with. 
So fundamentally, there's no point trying to convince each other that DC wiring is simpler or more complex than DCC - because what's simple for me might not be for you.

Jackpot.  You put your finger on it.  In fact, for people who grew up throwing together home-brew circuits with transistors, relays, and diodes, that stuff seems so intuitive that even if there are 30 parts on a board, it seems "simple" to them.  And I say, what's wrong with that?  As long as the end product works and is reliable (and that's a key point).

[nkalanaga]

You only have one train running at a time, right?  So you actually only need one throttle for the staging tracks.  Get a basic transistor throttle of your choice, replace the rotary potentiometer with a linear one, and connect that to a motor-drive switch machine.  Run THAT at a voltage that gives you the acceleration rate you want.

Going into staging, a relay drives the machine one way, reducing track voltage to zero.  Once the switches have been thrown, and the next train is ready to leave, the relay reverses, driving the motor back, and increasing the voltage.

All mechanical, just relays, a motor, and a simple linkage.

If you want to use the same throttle to run the trains on the visible track, it should be possible to design a circuit that would use the track power as input to a secondary throttle, and use that to control the staging tracks.

[peteski]

Jackpot.  You put your finger on it.  In fact, for people who grew up throwing together home-brew circuits with transistors, relays, and diodes, that stuff seems so intuitive that even if there are 30 parts on a board, it seems "simple" to them.  And I say, what's wrong with that?  As long as the end product works and is reliable (and that's a key point).

Well . . . I am perfectly comfortable with designing home-brewed circuits from scratch. But if an affordable technology already exists which I'm familiar with, and which will produce similar results, I'll chose to go that route.

This thread started as a simple single-track main DC layout with return loops which now has evolved into much more complex design with multiple stage tracks and automatic speed control. To me that is no longer a basic and simple DC layout. Although Arduino was also mentioned right from the get-go.  I don't know anymore.

You know, many DCC decoders (not Digitrax) have automatic brake control which is implemented with just few diodes. 

[mmagliaro]

Well . . . I am perfectly comfortable with designing home-brewed circuits from scratch. But if an affordable technology already exists which I'm familiar with, and which will produce similar results, I'll chose to go that route.

This thread started as a simple single-track main DC layout with return loops which now has evolved into much more complex design with multiple stage tracks and automatic speed control. To me that is no longer a basic and simple DC layout. Although Arduino was also mentioned right from the get-go.  I don't know anymore.

You know, many DCC decoders (not Digitrax) have automatic brake control which is implemented with just few diodes. 

I'm with you on the point that this project has grown to be a lot more complicated than first presented.  Multiple staging tracks, if there were just manually selected by the operator, would be no big deal.  But cycling through them automatically and slowing/stopping then starting the next train automatically do present some tricky things to work out.

[mmagliaro]

Okay, so here's a way to get your track selection working.  Yes, the complexity and component counts are going to start climbing.

I have two proposals.  One based on electronics and which still has some issues, and another totally
Rube Goldberg approach that is purely electromechanical.

Electronics idea:

Use a decade counter chip, driven by a 555 timer in one-shot mode.
Before anybody freaks out, I found links that have ready-made schematics for these things using very few components. 

This DOES assume that you understand the basic of transistors, current amplification, and basic electronic
assembly.   But I think we already established that the OP is in that camp and can do this.

The decade counter chip has 10 outputs that can drive a small load like an LED.  Everytime a clock pulse is applied to the counter's input, it switches to the next output and turns that on.  So by applying pulse-pulse-pulse, the 10 outputs will switch on in sequence (1, then 2, then 3, and so on and then back to 1 after it hits 10).

A circuit for this is here:
https://www.allaboutcircuits.com/textbook/experiments/chpt-7/led-sequencer/
(forget the 10-segment bar graph thingy they show for all the LEDs.. just use those 10 outputs to select your
tracks... more on that in a minute).

But as shown, that thing is going to just clock along automatically.  We want it to only advance to the next output on some "detection" from your staging loop.  Sooo.... instead of wiring the 555 timer shown in the above circuit
as a timer, use it in "one shot" mode.
Like this: http://www.dummies.com/programming/electronics/components/electronics-components-555-timer-chip-in-monostable-one-shot-mode/

SW1, the switch shown in the above diagram, is the key.  That will be your optical (or whatever) detector in the reverse loop/staging yard.  The train comes into the yard, pulls into the already-selected open track (where it originally came OUT of), and when it reaches the end of the staging yard, it triggers the detector.
That will one-shot the 555, which will advance the counter to the next output in the sequence.

Up to this point, the whole thing is only two chips and 2 or 3 other components (resistor, capacitor)

Now what? 
Here's what.  Each of those 10 outputs needs to be connect to a small transistor driver so it can handle a bigger load.  And from the transistor driver, it goes to your diode ladder that you wired up to allow selection of the turnouts in the staging yard.
So now you have a bunch of diodes for the ladder, 10 little transistors, and some other resistors and caps just to
make the transistor current drivers actually work.  Yep, it's a bunch of parts.

This will handle up to 10 tracks.  If you are only doing, say, 5, the component count would be a lot less.

Issues:
1.
Since this thing has no "memory", when you shut of the power, it will start up at track 1 every time, and if a train is already out on the main and has pulled out of, say, track 6, this is going to cause trouble.  When it comes into the yard, track 1 will be selected and the train is going to try to pull in there.  The only way to avoid this is to make sure the train is parked back in the yard before you shut off the power.
...or... add yet more circuitry to "remember" what track was selected and go back to that one on power up.


2.
This doesn't address the automatical slow-down/speed-up.

==========================================================

The second idea?  An "electromechanical" approach.

Okay... try not to laugh.  You know how a timer on a dishwasher works?  It's a motor with a bronze wiper finger set that slowly rotates around, with the fingers contacting various traces on a piece of PC board to complete circuits and turn on features.

Well....  Hook up a slow-speed motor with a wiper arm like that, and make a printer circuit board with the traces arranged so that as it rotates, it powers your diode ladder to select a track. 
The PC board will take a bit of work to draw out and etch, and then it has to be mounted to the motor.
But this does solve the problem of memory.  The state of the system will be wherever the motor is, so when you turn things off and on, it will just continue on from where it left off.

1. Train enters yard, reaches end of its staging track, activates detector
2. Detector starts motor rotation.  Motor rotates, connects two new traces on the board, that powers a different input to the your diode ladder, turnouts throw, a new staging track is active.

There is still some magic to be worked out, like stopping the motor and gradually starting and stopping trains.
But honestly, if you want to go all DC with this, the motor/wiper system is sounding really good to me.

[Sokramiketes]

You could call it scope creep.  I just want to plan for further phases past one-train two reverse loop operation.  I want to get most of the feeder wires in place once.

My comments about complex wiring for signals on DCC was not to say it isn't complex on DC as well, just that I don't think the DCC wiring should be classified as "easier" or "simpler".  At least not by the factor of 10 usually portrayed.