Part 1 DETECTION

My suggestion for detection is Bruce Chubb's optimized detector.  This was presented in an MR series in the late 1980's and in another series on MRC a year or so ago.  The basic idea is to pass current through back to back diodes and use the voltage drop across the diode when a train is operating in a block to activate the detector.  The detector has a number of desirable features - controlled sensitivity, time delays to simulate the prototype and eliminate problems from dirty track, and  an output trough a transistor in a open collector design.  This means that the transistor when turned on will provide a path to the circuit ground and turn on a LED or circuit connected with power to the collector   The detector was designed for DC -  a connection to the track was routed through the diodes on the detector.  This is the only connection to the track.   Besides the diodes, it uses 1 IC, 1 Transistor, 1 pot, 2 caps and a number of resistors. One detector is needed for each block that that you want to show "occupied" and where signals are to be installed.

When the articles were published they were part of a complete computer system and as I didn't have a spare computer for the railroad, I passed it over.  However, later I realized that just the detector could be used  and it would provide indication on a dispatcher panel and well as detection for a signal system. At that time I was using DC power and the detectors worked just as expected.  When I converted to DCC (a Lenz System)  it worked again as expected, with one problem.  The diodes that were in the original design were 3 amp capacity and my system was set up for 5 amps.  Joe Fugate to the rescue.  Using auto lamps (1156) in series with the detector limited the current to 2.1 amps and saved the diodes from destruction.  If they were built now, I would use 6 amp diodes.

Chubb also designed a detector for just DCC  (DCCOD) that uses a transformer instead of  diodes to detect the presence of a train on the track.  This circuit board as well as the transformer are available form his company.

 The DC detector requires a +12, -12 volt power supply.  These voltages, as well as 5 volts for all the signals can be obtained from an old XT computer power supply.  These are available on line and instructions on how to use them are on YouTube.  Be aware that it has to have a load  (can use a light bulb) to work if you want to test it  Otherwise, Chubb has a circuit to built a =12-12 supply.

I have not posted pictures on this site for a long time, so I need  to be refreshed on how to do it. .

Bill     More to follow

Pictures are coming Looked at video on how

I'll be back with some pictures and diagrams

Bill

Part 1 Detectors con't

Bruce Chubb was a web site (JLC Enterprises, Inc) that shows parts that are available , costs and downloads on how to populate the boards and directions on how to install them on the layout   https://www.jlcenterprises.net The OD (for DC) is listed for $5.00, the DCCOD,  $4.00 and the DCC transformer for $2.25.  The IC used is LM339N (Quad Comparator) and as a CMOS device does not require a regulated 5 volt supply.  The detector is designed for plugging into a mother board to aid in the installation, aslo available on the site.

The detector for DC part layout and circuit are shown 

.

The DCC detector is similar

On the Chubb wet site look for "products", "detectors" for the DCCOD the OD and mother boards for pictures and descriptions.  Under "Information station" look for downloads for Handbook Chapter2 Fundementals, Chap 3 OD and Chapter 4 DCCOD .

Next part will be on the signal circuits 

Bill 

DCCOD with DCC systems...

Readers might be interested in some key points...One of the benefits of the DCCOD rather than OD is the detection by use of the transformer (PT1). This way, the DCCOD is not actually part of the DCC bus circuit, but rather senses occupation by virtue of one side of the DCC bus wrapping around the transformer.

Part of the important significance of this that since some DCC systems require Opto-isolation with detectors, using a DCCOD eliminates this need even if the DCC system has Opto capability.

Other benefits include overall less voltage drop over the system and it keeps higher track power issues seperate from logic signals. DCCOD therefore are less susceptible to false occupation or false logic, and provide incredible sensitivity. Mine will sense a damp finger across the rails.

Cheers!

A couple thoughts

Bruce's design is up to rev K, do not know what the difference between J and K is though. You can see the design in the sample of his book, down a couple screens:  https://www.jlcenterprises.net/pages/chapter3

Remember that the 3 amp rating is continuous duty so although you have a system capable of supplying 5 amps as long as you are not drawing 5 amps continuous you would be fine. Even with a short where you can get the 5 amps the system could shut down before any diode failure would depend on how fast the circuit breaker reacts.

Although as you pointed you can take steps to prevent even a short from drawing the full five amps.

Good points Chris and Jim

If you have a DCC system, use the DCCOD!  The isolation with the transformer is the ideal way to go.  Several of my friends have gone this way and it works great.  The DC system using the OD has a common ground connection for the track, detectors and the signal system so is not completely isolated in an electrical sense. The new Rev K detector for DC has some additions to allow the detectors to be installed around the layout rather than on a mother board at one single location and has increased sensitivity.  If I were to start over, I would use the DCCOD and space them around the layout where signals were required and keep the wires shorter.

As Bruce points out in the handbook, the cost of the detectors with board and parts is $10.00-12.00 depending on quanitity discounts etc. Selling your old DC ODs is an option, if there is a market for them.  At this point, I think I'll stay with what I've got.

One other point, resistor R11 is part of the turnoff delay and is 2,2 meg ohms.  I changed it to 1 meg to slightly reduce the time with short blocks.

thanks for your comments.

Bill

Part 2 signal driver circuits

Years ago there was an article in MR about a "simple signal circuit"  Although it work, it did not always turn off the lights completely.  Rob Paisley on his model railroad electronics web site came up with a circuit "Not so simple 3 light signal circuit".  The circuit used one IC (LM339) with diodes and resistors and was designed for 5 volts, operates by grounding the inputs to provide red and yellow indications and can be set up for approach lighting. Thus it can be easily connected to the Chubb detectors through the open collector to provide the required ground.  LEDs are used for the signals and with 3 colors can be used for in-line, type G , color position, or PRR (all amber bulbs) signal types.  Searchlight signals using bi-color 2 lead or 3 lead LEDs needed another design, however the newer RGY smd leds could be used.  One circuit is needed for each signal head of three lights.  Signals with 2 heads as in the entrance to a passing siding would require two circuits.

As Rob designed the circuit, it has 2 input called detect#1 and detect#2.  Grounding either or both turned off the green (which is on with no inputs) and turned on the red.  I added two more detect inputs for a total of 4 to provide for Dispatcher or other inputs to give a red signal.  The circuit also sends a output to the next previous block to show a yellow signal there  If there is no red input, but a yellow in input is received, a yellow indication is shown, If neither red or yellow inputs, a green is shown..  

The circuit is connected to the detector through detect#1.  Detect#2 is used to connect the circuit to a switch machine so that a red will show when the turnout is against the direction of travel.     

The circuit can be built on a Radio Shack circuit board 276-170, which are still available.  Two circuits on each board

The board layout is as follows

Another circuit for searchlight signals using bicolor leds will be in next post.

Bill

Additional info on last post

The circuits shown in the last post are designed for three color signals (three leds) and with a common anode connection.  This means that the post on the signal is the positive connection and the leds are attached to the post by the plus lead.  The negative posts of the leds are soldered to wire wrap wire the goes through the post and to the R, Y, G connections on the board.  I will go into additional info on making the signals in a later post. As mentioned above , this circuit will work for seachlight signals if the RGY SMD leds are used.  In later circuits that Rob Paisley published, he dropped this design for a more universal 3 transistor design that can be used for all types of signals including searchlight with bicolor leds, as well as signals that are either common anode or common cathode. I have used the IC circuit and find that it works great and supplies the needed signals on my layout and the club layout.

On the above circuit shown using the Radio Shack, there should be a connection from the positive lead on the left side of the board to the post of the signal (common anode design) that was left off.  The photo of the two circuits shows this connection.  

Rob's three transistor circuit will be in the next post with info on how to use it for three color and bicolor leds.

Note : the last five days started with a operating session at Curt Larue"s excellent PRR layout (good) and then a leak on the utlility room floor from a clogged drain, three Doctors visits for my wife and I , heavy rain followed by a snow storm, a leak in the roof from an ice dam, and several other problems.  I hope I can get back to finishing this blog in the near future, WITHOUT ADDITIONAL PROBLEMS.

Bill  

Another Low Cost Option - Photocells and Picaxe Microprocessor

Hi BIll,

If you don't run in the dark, you can use photocells to detect the trains.  They are cheap if you purchase them by the 100($25 or less). 

The Picaxe($3) chip can handle the detection, LED signal driving, distant block signals and is programed in Basic.  It can also drive target signals instead of R-Y-G if needed.

You'd only need a few resistors, an adjustable resistor for the light level and one picaxe chip for each block.

No matter what system you decide to use, get one of the professional groups to run off your printed circuit boards, it will save you a lot of time and headaches.

Rob Paisley's newer 3 transistor circuit

In case you  are not aware, Rob Paisley's web site is back up.    http://www.circuitous.ca/CircuitIndex.html

As I indicated in my last post, Rob has a new 3 transistor circuit that will work for most all situations. Printed circuit boards are available as the bare board, assembled and as board and parts.  The completed board is shown:

The circuit is:

The circuit is designed for either common anode or common cathode signals  Also it is designed for searchlight signals using bicolor leds, again either common cathode or anode.  The one change I wanted to make was to operate my signals on a 5 volt supply.  As a result, R1,2,3 is 33K, R5,7,9  3.3K and R4,6,8 between 330 and 1K depending on the brightness of the leds.  Most high intensity leds would be fine on 1K. The circuit is slightly modified for bicolor leds by removing diode D2 and replacing R6 with a jumper. The Biclor leds are wired as shown:

The resistors shown between the "Y" terminal and the led should be the same as R4 and R6. For much more information you should check out Rob's web site.  ( I have several searchlight signals that use an earlier version that removed D2 and run a jumper between "Y" and "R" and connected the led to 'G" and "R" .  I have not used the version shown above)  The other change for my signals was to add additional OD inputs as on the "not as simple 3 light signal"  circuit so that turnout position and dispatcher data could be added.

In my opinion, photocells and phototransistors will provide information that shows a specific location such as if a train is clear of a turnout for example going into stagging or for a grade crossing signal. I prefer to use current detection that cover a whole block so that I know the block is clear for the signals to change  I believe this is more realistic, but others may have other ideas.  For the circuits shown, PCBs are already available.  I just decided to build them on Radio Shack boards that were on hand.  The PCBs for the Detectors are also available as are the transformers for the DCCOD.  Both Bruce Chubb and Rob Paisley have the both items, however Rob's circuit is slightly different than the Chubb design.  I can only speak for the Chubb's  OD because that is the design I used and it has worked very well.     However, the others should work just as well.

Next time, signals

Bill

Making the Signals

An example of homemade signals - upper signal for going into a siding, lower for exiting a siding.  For the signal going into a siding there is a three light head over a two light head.  (GYR YR)  The heads were from IHC and or no longer available , but other sources are available.  The signal bridge is a Bachmann  plastic kit and the signals are both 3 light (GYR).  In both cases the signals are set for the main.  The LEDs used are T1 size and fit into the signal heads after the are drilled out to 1/8 inch.  Each signal head requires a signal circuit as described before. Having signals at sidings takes it a step above a simple ABS systems

The photo shows parts needed to make the signals.  At left bottom are the IHC heads for three color in line lamps, and at top are Type G heads, stamped out of shim brass and using 5/32" OD brass tubing as lamp holder and shields.  The mast is 1/8' tubing about 4" long so it can extend through subroadbed and plywood and about 2- 21/2 " above.  The lower part is 5/32" tubing long enough to extend through the subroadbed., but not as far as th 1/8"  tube.  This way the common lead can be soldered onto the tube and allow it to be inserted  into a 5/32 hole in the subroadbed.  A groove is cut into the tube just below the position of the lights for the wires.  My preference  for wire is 30 gauge wire wrap wire as it comes in colors and is not as likely be stripped when inserting it into the brass as would magnet wire. The top of the tube is closed with a small screw or bolt soldered in place. The base is a grommet that is also soldered on.  The ladder is brass signal ladder that I have had for years and may no longer be available.  An alternative is brass "N" gauge ladder with every other rung cut out.  The "G" type signal has 1/16" tubing soldered to the bottom of the head and a bracket is made of strip brass to attach it to the mast.  The common wire is soldered to the bottom of the mast so that it can be inserted into subroadbed. The LEDs are inserted into position in the head and can be held with a small amount of white glue (will dry clear) so that the positive lead of the LEDs are close to the mast. The leads are bent to touch the mast and are soldered in place. (Note: Be sure to determine if the circuit you are using requires positive or negative common as that lead of the LEDs are soldered to the mast.)  Clean up the solder joints and paint black (or silver).  It is a good idea to test the LEDs as you go to be sure that are still working after soldering.and when you are finished.  While the circuits have the required resistors, the test circuit will need resistors depending on the voltage used ( 1K will be OK for a 9 volt battery)

The dwarf is made from 1/4" x 1/4" brass stock 1/2" long, drilled for 5/32" tubing and rounded on the top. The hoods are 5/32" tubing to hold T1 LEDs.  The bottom of the dwarf is drilled 3/32" so that a length of that size tube can be inserted to hold the signal in place. Again the positive leads of the LEDs are soldered to the brass body of the signals and wires soldered to the negative leads to be fed through the subroadbed.  The rear of the signal can be sealed with caulk to cover the wires and prevent light leaks.  An alternative would be to make the entire signal out of styrene.  Again 1/4 x1/4" stock, drilled for 1/8" to hold the LEDs and using 1/8"  ID straws for the shields.  In this case, the positive leads woild be soldered together with a single wire and then separate wires for the negative leads.

The signal on the right has been made for a specific location and shows what can be made out of brass and soldered together.  The same idea can be used to make cantilever signal bridges.

The signals can also be made with 1.8mm LEDs and that would reduce the size of the tubing used.  Also instead of brass, styrene can be used for many of the parts.  With LEDs, we need not be concerned about melting he plastic as we were with grain of wheat lamps.

Next time searchlight signals.

Bill

Making searchlight signals

The signal is made using 3/32 round tubing for the mast about 4" long so that it will extend below the subroad bed and  2 to 2.25"  above.   The lower mast is 1/8 tubing 2"  long. soldered 11/2" below the top.  Square tubing 3/16" ID is soldered to the 1/8" tubing.  Most of the 3 lead LEDs I have used were common cathode so a wire is soldered to the bottom of the mast to connect to the negative connector of the circuit. The ladder stock (either HO brass signal ladder or N gauge ladder with every other rung removed ) is soldered to the mast at the top and with the brass strips as shown.  A grove is cut in the top of the mast with a cutoff disk and the center led of the LED is solder into the groove.  The provides the connection for the LED and also the support for the lsignal head.  Cut off the excess length of the lead.  Shorten the other two leads and using a wire wrap tool and wire wrap wire (red and green) connect the wire and solder, as close to the led as possible. Using wire wrap and then soldering provide a good mechanical connection and soldering can be quick. Test the LED for operation after soldering to be sure it still is working. (can use a 9 volt battery and a 1K resistor)  Cut the LED leads short and thread the wire through the mast tube. A disk of styrene 1/2" in diameter is punched out with a 1/8" hole in the center.  This is glued to the LED with white glue . A visor is made from a 1/8" coffee stirrer  Paint either black or silver as required by your railroad.  The signal is installed through a 1/8" hole after pullimg the wires through.

If a 3 color SMD LED is used, the procedure is slightly different.  The basic mast is made the same. However the LED has 4 wires soldered to it.  All 4 are pulled through the tube and no wire is needed attached to the bottom of the mast.  The LED is glued to the top of the mast perpendicular to the mast  There are lenses available from Mouser Electronics.that can be used, cemented to the LED,  Again cut out a styrene disk 1/2" diameter with a 1/8" hole.  Also cut out a stirrer for a visor.  Soldering the 4 leads to the LED is done with all 4 leads going in the same direction.  Different color wires will help the installation. .

       .       

For this signal, I used a cast base which adds to the look of the signal.

I will have to search for the info on the lens used on the signal.. Next time.

Bill

Computer interface

I have these new boards and like them but like to take it on step further and use them with JMRI. I have figured out how to get the detection part but how do I get the inputs to the signal driver boards. I have a NCE power pro system. Any suggestions.

Bob

Computer Interface???

Bob,

What boards are you using to get detection?  Are you planning to use Panel Pro and want to connect this to your layout to operate signals?  The system I used does not use a computer bur rather separate circuits to operate the signals.  

Bruce Chubb has published a number of articles and several manuals on how to connect his detectors and other inputs into a computer for operating signals and controlling trains.  Although I have not done this myself, and do not plan to, I have a friend that has and is currently operating his layout using Chubb's DCCOD detectors, Chubb's computer interface cards and Panel Pro and it works very well.  If there is more interest in this, I can get information on how he did it and what circuits and boards are needed.

Bill