Right of Way Crossing Signals

See this post from an earlier OGR thread.

Here's the back of the Right of Way 4060 carded-package from an eBay listing.

Are the stripes on resistor left-to-right brown-black-RED-gold?  If it is indeed RED (not Orange, hard to tell), then that is a 1000 Ohm resistor and would a suitable value to dim the LEDs given the comment from linked thread that each signal base already has 2 resistors of value 560 Ohms.  Why not simply apply 14-16V Accessory AC to the Red and Black wires and decide for yourself if the extra resistor is needed.

Blue wires explained in Step 7.

Hello Stan,

Thank you for your reply and photo of the wiring diagram.

The stripe in question has the following color sequence brown, black, red, gold.

As noted I saw two of them also on the board itself.

What I don't get is why add another resistor?

The lights themselves aren't that bright really as I found them to be acceptable.

When I tested the output end of the resistor I got 0 volts.

As I noted earlier this isn't my best suit.

@Allegheny posted:

..

As noted I saw two of them also on the board itself.

Q1. By "board itself" do you mean there's a board on the bottom of each signal pole itself with 2 resistors? 

Q2. Since it appears the RoW 4060 has 2 signal poles and 1 controller, is this all you got?  In other words with 2 controller boards it looks like you're missing a 2nd pair of poles.

Q3. Does the controller board on the right also have the brown-black-red-gold resistor in the dangling white wire too?

Q4.  Can you post a photo of the bottom of the controller board?

Q5. Can you either photo or read the markings/lettering on the components shown?  The markings will be cryptic combination of letter and numbers.  BTW, in reference to Q1, the controller board has 4 resistors on it as indicated below...just want to confirm you were referring to 2 resistors on a board in the base of the signal pole.

Q6. Were the 3 wires connected between the controller board and signal pole when you measured 0 Volts on the output of the resistor?  I'm assuming not - all will be explained if you address my previous questions. 

1) Why are there two control boards?  One board powers a set of two crossbucks.

2) Ditch the extra, external 1k resistor.  Someone may have been using that board for another use.

3) "The signals work just great if I apply between 5V - 8VDC."  I am assuming this is directly to the signals themselves. There is a 15VDC regulator after a half-wave rectified supply, on the control board. That's why the instructions show a minimum of 14VAC input to the board. (The rectifier supplies a higher voltage to the regulator)

4) "When I tested the output end of the resistor I got 0 volts."  Tested to what? The WHT wire is the output from the 15V regulator. The GRN wires sink the LED current through NPN and PNP transistors (when active), alternately driven by a 555 timer, as the crossbuck flashes.

As Stan said, just hook it up to a 14VAC accessory post, as shown.

Dave

Dave, thanks for chiming in.  I get nervous when I see modification like that dangling resistor!  That's why I want Kazar to confirm that his signal poles indeed have the 2 resistors (560 Ohms per your previous post) in the base.  For all I know the seller modified the signal poles to eliminate the resistors so as to operate with some other controller board!

I see you identified the regulator as a 7815 type.  Do you happen to have the part number of the two transistors...and did you happen to sketch out a schematic?  I believe this board can drive incandescent bulb flashers.  Point being, if the seller had modified his signal poles too, then directly driving them without the dangling resistor would be unsatisfactory.

@stan2004 posted:

Stan, 

I'll try my best to answer your questions: 

 

Q1. By "board itself" do you mean there's a board on the bottom of each signal pole itself with 2 resistors? 

No a board does not exist on the bottom of the signal.  I meant only the control board.

Q2. Since it appears the RoW 4060 has 2 signal poles and 1 controller, is this all you got?  In other words with 2 controller boards it looks like you're missing a 2nd pair of poles.

I have one other crossing signal.  I purchased two control boards and three signals.

Q3. Does the controller board on the right also have the brown-black-red-gold resistor in the dangling white wire too?

Only one control board has the resistor attached the other does not.

Q4.  Can you post a photo of the bottom of the controller board?

Not a problem be happy too.  Both control boards are identical.  Pictured below is the control board with the extra resistor.

Control Board without the extra resistor (shown below)

Q5. Can you either photo or read the markings/lettering on the components shown?  The markings will be cryptic combination of letter and numbers.  BTW, in reference to Q1, the controller board has 4 resistors on it as indicated below...just want to confirm you were referring to 2 resistors on a board in the base of the signal pole.

I'll try my best to capture your request fully:

These photos are from the board without the extra resistor attached to the white wire.

 

 

 

 

 

Below is the control board with the extra resistor.  It looks the same as the one above, but I'll post a picture of the resistors attached to the control board to help clarify any lingering concerns.

 

 

 

Q6. Were the 3 wires connected between the controller board and signal pole when you measured 0 Volts on the output of the resistor?  I'm assuming not - all will be explained if you address my previous questions. 

Ok, so I ran an experiment and found the following:

1. With a 15V AC pwr supply, the control board with the additional resistor did allow the  crossbuck signal flash back and forth.   The bulb was lit at a reasonable level of brightness.  Using a Voltmeter this is what I learned: It had a reading of 14.5 -15.0 A DC with the dial set to 2A.  Of course it was not steady and would go to zero with each pulse of the flash.

2. With a 15V AC pwr supply the control board without the additional resistor did function correctly flashing back and forth, but the bulb brightness was significantly brighter.  Using the very same setting on the voltmeter I was getting a reading of 0.346 to 0.455 A DC BUT the flashing would stop when the leads of the voltmeter were attached.?

In both cases one lead was on the white wire and the other on the green.

I hope my responses are enough to solve the puzzle that I put forth.   

I hadn't counted on it being as time consuming on the forum members part to assist me.  

I'm grateful to each of you who have responded thus far.

 

I think we are at the 2 yard line and it's first down.  In other words we're almost there!

OK.  So this might not be photo-graphable but could you peer at the hidden face of the component shown at the red arrow and read the markings?  The markings should be something like "B744" white marking on the greenish component shown in photo below but likely something different - a letter followed by 3 numbers.  You may have to gently bend one or both components away from each other to read the markings.  Since we've come this far, I can draw out a schematic of the circuit for-the-OGR-record.  This crossing signal has come up in previous discussions so stands to reason it will come up again!

Also, could you take a photo of your meter or tell me its model number?  That is, if you're measuring the VOLTAGE between a green wire and white wire, you should be using a mode-setting on the meter that measures "V" or Volts...rather than "A" which is Amps.

So to be clear, with no meter attached, both controller boards work albeit the one with the added resistor being dimmer? 

Stan:

LM340 15V regulator

MJE182 NPN transistor

MJE172 PNP transistor

Transistors were quite the overkill for a couple of LEDs.

Dave

@stan2004 posted:

I think we are at the 2 yard line and it's first down.  In other words we're almost there!

OK.  So this might not be photo-graphable but could you peer at the hidden face of the component shown at the red arrow and read the markings?  The markings should be something like "B744" white marking on the greenish component shown in photo below but likely something different - a letter followed by 3 numbers.  You may have to gently bend one or both components away from each other to read the markings.  Since we've come this far, I can draw out a schematic of the circuit for-the-OGR-record.  This crossing signal has come up in previous discussions so stands to reason it will come up again!

Also, could you take a photo of your meter or tell me its model number?  That is, if you're measuring the VOLTAGE between a green wire and white wire, you should be using a mode-setting on the meter that measures "V" or Volts...rather than "A" which is Amps.

So to be clear, with no meter attached, both controller boards work albeit the one with the added resistor being dimmer? 

 

Ok Stan,

Let's see if we can get it into the End Zone on the first down. 

The component behind the B744 is:  NEC,  D794 P,  D  2Y

The photo of the attached meter is depicting the setting I was using to make the measurements.  I'm certain its not correct, but like I said I'm not very good at electronics as I do not dabble in it often enough to really learn it well. 

Also the control board with the extra resistor does indeed dim the crossing buck lights to a very acceptable level while flashing normally.

The control board without the extra resistor does flash normally, but the lights are much brighter.

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

Also for those with sharp eyes, that is indeed a Texas Instruments SR-50 calculator on my desk.  I bought it new back when they first came out in 1973 during my 2nd year in college.   I've used that same calculator everyday since then for school and my job.  No need to get anything else. 

When the lockdown was about to be implemented, that was the first thing I packed to take home with me in order to work from home.   Everything still works and charges like a champ.   

Dave, perfect!  BTW, where exactly are the 2 resistors in the signal pole?  Does the bottom plate (Made in China) pop out, or are the resistors up near the LEDs?

As for the transistors, I'm convinced this board was designed to drive either LED or incandescent bulbs.  The MJE172/182 pair are 3 Amp transistors (as are the B744/D794 transistors on Kazar's version). 

For inquiring minds, the give-away is the 680 Ohm base resistor driving the two transistors.  That's about 20 mA of base current.  For a practical transistor switching circuit, that supports an output current of, say, 1/2 Amp.  You wouldn't need 1/2 Amp to drive two LED strings!  Additionally, if only meant to drive ~20mA LEDs, they would not have used "expensive" 3 Amp transistors.  

I see the date code on the 555 IC timer chip on Kazar's board is 1992!  This doesn't tell me when the board itself was designed but assuming RoW was cranking out these puppies and using "fresh" factory parts, one must note that LEDs were still somewhat of a novelty in O-gauge 30 years ago!   Today, hobbyists buy red LEDs for a penny or two whilst incandescent bulbs cost at least ten times that.  I wouldn't be surprised if in 1990 the pricing was exactly opposite!  Hence, it was prudent for RoW to support either style of illumination.  In my opinion of course!

Wow, that's amazing the batteries in the SR-50 are still holding a charge after all these years.  I seem to recall having to change the batteries in my HP LED scientific calculator from that era.

Anyway, you nailed it on the transistor.  The D794 is a match for the B744.  And thanks to Dave, we have the design of the controller board for posterity.

So you can decide for yourself which controller board (with or without dangling resistor) to use based on your brightness preference.

As for meter, when you are trying to measure a Voltage (AC or DC), the red probe needs to be plugged into the meter jack labeled V/F.  So I think we're in the end-zone but if you want to Monday-morning-quarterback, I suppose you could go back and measure the DC voltage (use the DC 20V range) between the white wire and one of the green wires.  You'll measure near 15V and then near 0V as the flashers alternate.  And for the controller with the dangling resistor, you'll measure maybe 10V and 0V between the white-wire "after" the resistor and one of the green wires.  In other words, the resistor reduces the voltage that the signal pole receives and hence the LEDs are dimmer.

Ta da!  I think we just made it under-the-wire... case opened and closed in the same day (24 hours)!

… under-the-wire  Not quite, I was up late watching an old horror movie and decided to check the thread.

Yes, the resistors are in the base on a small board, held in by a glued on base plate. I would not try to open it if they work.

Stan, your theory about incandescent compatibility makes sense, given your historical hypothesis. 

Dave

Thank you so very much both Stan and Dave for your assistance in solving the mystery of the "extra" resistor.   So I guess the original owner must have needed to add it to tone down the intensity of the light.  Possibly to preserve its longevity given it was a bulb?

Now that the circuitry has been fully determined and analysed, could you give a brief tutorial on exactly what is going on in it?  I'd like to learn how this thing works as its now a curiosity on my part.  

Given what these control boards do, how different would they be today for new cross bucks?   For example there are many items for sale that flash lights but they do not have anything like the size of this control board packed in them.

Thanks!!

I think the extra resistor was just to reduce brightness of an LED version.  The particular resistor used would not have been suitable to lower brightness of a bulb version; the bulbs would never turn on though I suppose one could argue doing so preserves their longevity.

As for how it works. As you observe, there are dozens of designs out there.  There a three key components: resistor, capacitor, and switch (aka transistor).  A resistor limits the amount of current.  Feeding current into a capacitor charges its voltage upward, extracting current from a capacitor discharges its voltage downward.   A switch passes current or blocks current. 

So the idea is a charging switch turns on allowing the current to charge the capacitor.  When the voltage reaches some defined level the charging switch turns off and a discharging switch turns on to discharge the capacitor.  When the voltage drops to some defined level the process reverses.  Lather rinse repeat.  Adjusting the resistor value controls the rate at which the capacitor charges and discharges and hence the frequency.  The LEDs are attached to the circuit so that one LED is on when the circuit is charging and the other is one when the circuit is discharging.

The RoW board is large in part because it is capable of driving incandescent bulbs which consume an order of magnitude more power than LEDs.  In general, handling larger power requires larger components.  If done today, I'd think any cross buck manufacturer would use LEDs and simply not support bulbs.  I'd think it would be routine to fit an LED flasher into the base of an O-gauge signal pole.  Here's a hand-built LED flasher circuit to alternate two ditch-light LEDs which coincidentally cycles at about the same rate as a crossing flasher.  AC track voltage comes in on the left driving a pair of LEDs on the right.

Wow, that was a really exciting analysis.  I enjoyed the read.  The product was made in the early 90's.  The boards were made at Techno-Mind and the signals came from Lesear.  Techno-mind had quite an operation including wire bonding.  They also made the first Williams Tru-blast II sound systems and the RoW 3 color signal internals. Quite an operation in Hong Kong.  Lesear was known as the best die caster over there; they made Hot Wheels, Matchbox, and many other automobile models.  Lots of neat stuff to see in their showroom.

So the key to the board is the 8 pin component marked LM555.  It is a timer IC wired in the oscillator mode.  The capacitor and resistors connected to it set the timing at 1Hz.  The transistors are overkill but people still used 154 incandescent bulb cross bucks.

Are there any other questions that didn't get answered?  I know I'm late to the party.  Thanks to Steve at Atlas for tipping me off.

Lou N

Lou and Stan,

Wow you guys are simply great!

Where were you when I was in school taking my first electronic course?  If I had known you then, I wouldn't have had a lifetime of confusion how this stuff works.  It was an easy read and quite logical.  Make sense in terms of how it works.

The size of the present day flasher is incredible.  I have no idea how you can not only assemble something that small but also solder.  You must have some very specialized tool sets in order to do this type of work.

Out of curiosity, is that flasher unit available for sale?

Again thank you everyone very much for the information provided on this thread.

I've noted everything, printed out the circuit diagram and installation instructions and put them away with the crossing signals for future reference.

With many regards,

@Allegheny posted:

Lou and Stan,

Wow you guys are simply great!

Where were you when I was in school taking my first electronic course? 

Well I'm here now.  After 45 years in the electronics industry I'm a professor teaching MEMS (Micro Electro Mechanical Systems).  We teach the process of miniature assembly, soldering, and wire bonding.  And tomorrow we're finally allowed back in our clean room lab to catch up!

Lou N

Interesting thread. Lots of information.

Stan, have you got a schematic for that ditch light flasher board by chance? It's nice and compact.

Here is a video clip of one of a pair of ROWI #4062 crossing flashers on my layout. They are using the same flasher module board as in the OP. I trigger them with a pair of 153IR's set several feet to the left and to the right of the crossings. They operate in parallel so that all flashers trigger whenever either 153IR is tripped. Been working great for many years. To some the brightness of the LED's might be a bit much, but I kind of like them. Visitors always notice them right away.

Not sure why the horizontal scale is so huge. Maybe because I took this using my I phone I guess.

Rod

@Rod Stewart posted:

...

Stan, have you got a schematic for that ditch light flasher board by chance? It's nice and compact.

I posted the above schematic, how-it-works, and video on the JC Studio O-gauge archive. 

The alternating flasher portion is the classic relaxation oscillator (who comes up with these names? ) that has been floating around for years:

By using LEDs instead of incandescent bulbs, lower power and smaller components can be used.  The flasher circuit is controlled on/off by a DC offset detector so the ditch lights flash when the conventional-control engine horn is active.  Less than $1 in easy-to-find components.

Rod, time to warm up DipTrace and make that into a PCB.

@gunrunnerjohn posted:

Rod, time to warm up DipTrace and make that into a PCB.

I am thinking the same thing. It would be a handy little board for ditch light retrofits. 

Rod

Hmm... You have to have a horn input to work the ditch lights.

@gunrunnerjohn posted:

Hmm... You have to have a horn input to work the ditch lights.

Yes, very good point. I don't imagine there is any way to easily extract the horn signal out of a modern engine sound burst. And I can't think of any other convenient way to activate it.

Rod

@stan2004 posted:

Wow, that's amazing the batteries in the SR-50 are still holding a charge after all these years.  I seem to recall having to change the batteries in my HP LED scientific calculator from that era.

Anyway, you nailed it on the transistor.  The D794 is a match for the B744.  And thanks to Dave, we have the design of the controller board for posterity.

So you can decide for yourself which controller board (with or without dangling resistor) to use based on your brightness preference.

As for meter, when you are trying to measure a Voltage (AC or DC), the red probe needs to be plugged into the meter jack labeled V/F.  So I think we're in the end-zone but if you want to Monday-morning-quarterback, I suppose you could go back and measure the DC voltage (use the DC 20V range) between the white wire and one of the green wires.  You'll measure near 15V and then near 0V as the flashers alternate.  And for the controller with the dangling resistor, you'll measure maybe 10V and 0V between the white-wire "after" the resistor and one of the green wires.  In other words, the resistor reduces the voltage that the signal pole receives and hence the LEDs are dimmer.

Ta da!  I think we just made it under-the-wire... case opened and closed in the same day (24 hours)!

If I'm reading this correctly there are three flashers and two controllers.  One controller with no modifications connected to two flashers, all works well.  Has the other controller board been modified to operate with only one flasher and still maintain the correct illumination?  This may make sense if the the bulbs were driving incandescent bulbs.  Just a thought.

It's on my "to-do" list, I want to build a board that reads serial data and has outputs for stuff like whistle, bell, and perhaps blowdown.  That would give me the tools to add stuff like whistle steam, swinging bell, and blowdown steam to TMCC locomotives.