Full function signal system

Started by DACS, January 14, 2016, 09:03:48 AM

Previous topic - Next topic

deemery

RMC has Part 3 of a series on signalling by Dr Bruce Chubb in the latest (Feb 16) issue.  There's also a favorable review of an interesting book being self-published through Amazon.com, "DIY Advanced Model Railroad Signaling Electronics:  Sensors, Interactivity, Track Control"


dave
Modeling the Northeast in the 1890s - because the little voices told me to

DACS

#16
Ok everybody...here are three blocks on a breadboard to show you what I built for a signal system many moons ago.  Back when the twin-t was king.  The only difference, instead of my input being from a relay, I used opto-isolators to activate the signals.  TTL was still a youngster.
Now, I have the DCCBOD type detectors.  Which I built also.   None shown here though.  Just using a piece of wire to drive the Re input.
What you see here are as I stated, no driving transistors, no dropping resistors.  Very inexpensive!  Each signal head, uses a 74LS00 quad two input NAND chip.
All the circuitry is inside the chip, with the exception of course, to interconnect the inputs and outputs between the pins and in and out from the board.
The Re is low, or ground when the signal is red.  This drives the signal behind it to yellow.  East bound and westbound heads all have their own board.
Here in these photos, you can see the progression.  Granted, it is only three blocks, but you should get the idea.
Each board costs around $1.50.  Even a little less if you set up an assembly line and not try to only cut one board at a time. 
It is a building block system with four components.  Detector, signal head board, turnout control board and crossing board.  The latter is not really used that much, but just in case you ever need one!
I will be putting together a couple of the turnout control boards and integrating them for you and show them working here in the forum.
This system is not more difficult that putting together snap track...one section at a time.

I do apologize for the brighness of the LED's and the little blur.  I do not have a very professional camera.  My tripod broke so it's handheld.  Anyhow...
I used this system for 9 years and it operated flawlessly.  Not a single LED burned out in all that time.  The boards were never in need of repair or replacement!

This is a pic the chips used and the led's.  The wiring of course could be put on those little pro boards and hardwired just like you see here, but I make my own printed circuit boards with a laser printer.
The bent orange wire represents the detector being activated.  It is just being used directly to drive the signals.  As this thread progresses, I will show the detectors and all the interconnections that make this system work.



Here are three of the blocks showing none occupied...Green Board!!



Here the second block is occupied.  These signals represent Eastbound.  Westbound would be wired up the same.  One chip missing, but just pretend.  It would be green.



Here the top signal is showing occupied.



Dave

Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

DACS

Oops, this pic should have been the second above.  It shows the first block occupied and the two blocks in front clear.



Dave

Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

DACS

Small note here.  I removed the drawing showing the relay contacts for changing the interlocking for the turnouts.  I drew a D.P.D.T relay showing Ground and -5V.
I will redraw.  That was incorrect.  It is actually, supposed to be a S.P.D.T relay.  Center feed going to ground only.  Then the contacts going to NOR or REV.
I will draw it again and put it in the thread.

Dave

Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

jbvb

I'm curious, as I have applications for a bipolar LED driver which costs less than Circuitron's SD-3.
James

DACS

#20
Is that a complete question James.  Curious about?  Maybe I am just getting old and it's going over my head.   :D :D
I don't know much about Circuitron products.  I do know a driver for two color LED's can be built rather inexpensively, (inexpensive being a relative term determined by who you purchase from), using a 555 timer.  I have done that also. 
This system will not drive bi-polar LED's.  Just color position signals.
A detector can be built using a 555 timer with the pin 3 output driving bi-polar LED's.  Pin 7 is open collector.  It gives you two ways to use it.   I believe Mr. Paisley has one on his web site.  It is DCC only Transformer type Block occupancy detector (555 timer) It is rather compact and would not be that expensive to build.  All the parts for everything he uses and for that matter, what I use for this, can be had at Nebraska Surplus for a really good price.
I pay .40 for each chip on the signal drivers.  Not including the board or signal itself of course.  As I stated, the cost of each signal board is about 1.50 per.  That includes the board.
Three color position light is actually, my personal preference.

Dave

Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

jbvb

I'm modeling the B&M's Eastern Route in an era when all mast and dwarf signals were searchlights.  Oregon Rail Supply is the only supplier offering a style of ladders and platforms I can use to make relatively accurate mast searchlight signals.  I've built my own dwarf signals: http://www.railroad-line.com/forum/topic.asp?TOPIC_ID=39866

I could use Mr. Paisley's circuit board and build it without the approach lighting parts, but my dwarf signals are built with red/yellow bipolar LEDs for two aspects.  Something simpler would do the job, which is what I'm curious about.
James

DACS

#22
James
Well, just before I started texting this, I went to your link and saw what you were doing.  It blew my suggestion out the window!  You seem to have the patience of Job!  I see your dilemma.  Very nice looking signal you got there sir!
Am I reading that your problem lies in the size of the LED?  Not the electronics.  As red/yellow led's are available in 3mm, it seems you want 1.8mm.  That would put you closer to scale. 
You have texted you can use R.Paisley's circuit, just leaving out the approach lighting aspect.  Am I understanding correctily?
If so, have you considered piping the light into the signal with fibre optics?  Not sure about the simpler part.  Then you could use the 5mm Led for more light, driving it for just red/yellow.  You could use a 3 lead red/green.  Then it is just a matter of red on one lead.  Yellow if you drive both leads.
If I am still not understanding, exercise a little more patience and forgive me.  My hard drive is full.  Too many cookies up there.  ;D ;D  :o

Dave
Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

jbvb

Thanks, but the styrene dwarfs actually go together fairly fast.  I haven't yet found the patience to work with surface mount LEDs.  And the 1.8 mm and 1.6 mm LEDs are too easy to damage when I try to reduce the size of their rectangular bodies to fit in commercial head castings.  Years ago I lit one of the Century cast metal dwarf signals with fiber optics, but the result was not bright enough.  So it's 3MM diameter bi-polar for the time being.

I have 8 locations where a Red/Yellow bi-polar LED gives the two aspects I need.  Right now I'm driving them with a transfer contact on the turnout relay.

I would prefer to drive them electronically.  The ideal circuit would have one input.  Float is normal, ground reverses the polarity to the LED.  It would be cheaper than Mr. Paisley's 556 driver.  If it was more compact, even better.
James

DACS

#24
Now that I have shown you pics of three blocks working.  I will continue on with schematics.  As was said, some of you may not understand it, but there are many who do.
I will try not get geeky on you, so ask questions.  That's the only way to knowledge!

The chips that are used in this system, are the 7400 series TTL.  When I first built this system back in the late 70's,  TTl was still not that old.  For those of you who do not know what TTL stands for...Transistor, Transistor, Logic.

Today, we have at our disposal, 74LS00 chips.  Same circuitry, just greatly improved as a product.  I pay .40 cents apiece for these little wonders of electronics wizardry.

Each signal head has one chip mounted on a small PCB.  The signal driver board has on it, the 74LS00 Quad two input NAND

The internal circuitry is not important to build this system, it is too complicated.  So we use shorthand, or symbols.
This shows the 4 individual NAND gate symbols, inside the chip.



Pin 1 is the orientation pin.  Pin 7 is the gnd and pin 14 is +5VDC.

You need to have a 5 volt regulated power supply.  I have a 5 volt, 4 amp dual regulated power supply.  Though you do not use the negative voltage.

This system uses reverse logic for it to work.  Normal usage uses positive logic.  High input means true or 1.  Low input means false or 0.  I use Low input means true or 1.  High input means false or 0.

This is how I use the NAND circuitry, to make a signal.



The pin layouts, are very important to adhere to.  Do not deviate from this pin diagram schematic.  You might notice I do not have pin 7 or 14 on the drawing.  This will show up in a later posting.
Also note, that the LED's are common anode.  These are connected to the mast of the signal itself.  This is driven by pin 8 of the chip.
Coming off the board, there are only two inputs (Re and Ye) and four (kg, ky, kr)outputs.  Then it has the voltage and gnd.
This is one complete signal driver board.  There is one for every signal.  Eastbound and Westbound. 

The Re is Red enable input is actually, driven from gnd through the detector, when the block is occupied.    Yo is Yellow out to the previous board.  Ye is coming from the board in front and drives the yellow LED when the signal in front is Red, or occupied.
You may also notice, that pins 9 and 10 are tied together and grounded.  This is for NORMAL lighting.  Or the signals are on all the time.
Notice the dashed line going to Ge.  This is for making the signals APPROACH lighted.  Or, the signals come on only when the train enters the block behind it.  Then it will either light up Green, Yellow, or Red.
I will show how to connect this up later, if you prefer APPROACH over NORMAL.

Here is one signal head circuitry hardwired on my breadboard.



Dave
Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

DACS

#25
While I am waiting for the chips, led's, pcb material, etc., I thought I would go ahead and start up building some signals.
Today, I worked on a head.  Now these signals I am constructing are in 'O' scale.  I know most of you work in 'HO', but there could be some ideas here that you can glean.

The base for the head is a PCB that the led's are soldered into.  I made the traces by hand with rub on's for the leads.  The PCB is 13/16" x 13/64.  This signal system is common anode.  So the small strip of copper running from the top to the bottom, is where the anodes are soldered.  The opposite side, is done with small, individual pads with a small tab just below the led lead.  This small tab is where the wire will be soldered.  It should make for a very clean and neat little package when finished.  Shouldn't have big globs of solder.  At least that's the plan.



This makes the actual support for all the LED's, the wiring and the signal target.
The target is cut from .20mm brass shim stock.  The holes are drilled using a 5/32" bit, then cleaned up with a jewelers file.  The target is 61/64" x 15/32"
The prototype target is 57" x 34".



The LED's are 3mm.





The hoods can be made from brass or aluminum 5/32" tubing and are 5/16" in depth or length.



The LED's are soldered into the board, then the hoods are placed with ACC.  I have not done this yet as you can see in the pics.



Once that is done, the target is placed over the LED's and attached with ACC.
I don't have the wire yet.  Like the check, it's in the mail!







When I get the wire, I will attach everything permanently, then I will make an enclosure from plastic channel stock.  The wire will come out the bottom, of course.
The mast will have a hole for the wires to go into and down.
I will show all this once I get the materials I have ordered.
Well, it's a start!!

Dave
Seattle




I am never having another birthday.  The candles for the cake are starting to cost too much!

jbvb

The signals look nice.  I have been using (US) #30 Kynar wire off the rack at Radio Shack.  If they aren't still in business around Seattle, there should be at least a few electronics hobby/experimenter stores on the lines of You Do It in Needham, MA.
James

DACS

Thank you James!   I am using the Kynar wire also.  Yes, Radio Shack is still in business.  But, they are not what they used to be by far.

I get the Kynar at a local electronics supply house, but, they are very expensive.  20.00 a roll!  I am looking for a cheaper outlet!
Red, Yellow, Green, Black.  All the appropriate colors.

Dave
Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

jbvb

Your supply house probably considers M$ and Boeing their main audience.  Radio Shack lists 50' of 30 ga. Kynar for $5.49.
James

DACS

Thanks James for the info.

I have been in Radio Shack here,  but they sell it in a 3 pack with red, green and black.  That leaves yellow.  After I texted to you, I went to Jameco, and they have 100' rolls of #30, in all the right colors  They are charging 12.50 per roll.  Not bad for the amount of wire you get.  I am ordering it today.

Dave
Seattle
I am never having another birthday.  The candles for the cake are starting to cost too much!

Powered by EzPortal