Insulated Rail Relay Based Sensor

I think that's a good idea. I don't recall seeing much (or anything) like this for 3 rail. I do think it would fill a need and would be worth a try. Of course, if you posted the gerber files and parts list I would certainly try this one myself at the inexpensive PCB maker you found (forgot the name).    It just might be too good of an idea to do that with though.

We are discussing the IR detectors in another thread (ITADs and Making Your Own) and the IR devices are, so far, having some problems, but more things to try are in the works. Your input is welcome, BTW, if you haven't been there already. 

John

If your looking for feedback I think this is a great idea and would buy these if you bring them to market.

Bill

Nice design, I would be interested in a few.

I would be interested.

Dear John

Nice simple design. What would the manufacture (you) suggest retail price for one of these?

John

Looks like another winner. I have been using the SPDT version based on input from you and Stan 2004  in the past with no problems. I have also been looking at A DPDT version using the same relay with terminal boards for ease of connection as shown above. I'm sure this will go over big wheather it be in kit form or completely assembled for those who are adverse to assembly and soldering.

Gerry

I have a couple of prototype PCB's coming, and I am trying to line up cheap components.  Note that I used all thru-hole stuff as that's typically easier to get cheaply and also makes it easy to build.  Maybe there'll be some low cost kits in the future.   It's a basic circuit, and one I've been using on our club layout hand building them.  I figured I'd like to have about 15 of these myself to replace the hodgepodge of stuff we use.  I did find some nice terminal blocks for a very good price and ordered a bunch of them.  I know I can use them somewhere if not here.

I'd like it to be cheap, in the $10 range, depending on the component costs.

Hey John,

You know more about the relay you are using than I do so I will just ask......

Do you think there is a need for a northbound diode across the coil?

Lou N

That thought crossed my mind, but I've been using this relay in a similar lashup for a long time with no ill effects.  It has a 15ma coil, so any kickback seems to get lost in the cap.  I've had them running several years running signals, still running.

In the off chance that it should prove to be an issue, I left provisions for an SMT diode at D2 on the left.

So will this trigger an MTH per signal bridge?

You get relay contact closures when something is on the insulated rail with bounce protection.  If you can trigger it with something like the MTH ITAD, you can trigger it with this board.

prrhorseshoecurve posted:

So will this trigger an MTH per signal bridge?

If it does count me in for a bunch!

Hi John - Great looking easy circuit - I'd be in for a couple.

I'd be interested in some too!

gunrunnerjohn posted:

That thought crossed my mind, but I've been using this relay in a similar lashup for a long time with no ill effects.  It has a 15ma coil, so any kickback seems to get lost in the cap.  I've had them running several years running signals, still running.

Thanks John. Perfect timing. I had a call today from a friend asking about IR devices and I said there were better ways. Will look forward to your new product introduction.

Lou N 

I would be a buyer too !!!!   come on John what are you waiting for !!! lol Let's go, they should be packaged and ready to go !!

Alex

Well, I have to get some parts in and actually assemble a few for final tests.  Since it's a circuit that I've been using, I really just want to see the boards and make sure the components like the terminal strips and the like actually fit with the current layout.  I've been burned too many times buying a bunch of boards, only to find out I made some spacing error!

Lou, see previous post, I added a SMT diode on the end in case for any reason I deem the diode should be there.  I didn't want to make the board any larger for the full sized diode.

I would definitely take some too. Kinda what I was getting at at my earlier post referring to the gerber files and trying it myself. Assembled, kits, boards with parts list, any way is fine with me. I do like the idea of the through hole components for the kits though, I think many other will too. But I do need to try that SMT stuff one of these days.

Also, I still haven't forgotten trying the Watchdog devices. I'm still missing some parts for the layout changes I want to make. If I don't get them pretty soon I will go ahead and try them and rework things later on.

John:

Am I reading the diagram right as you show the relay being powered by the “center rail”?  If so, that means it can be powered from an AC transformer thereby eliminating the need for a separate DC power supply, which is another selling point. 

But why connect it to the center rail?  Why not an accessory power terminal instead because in conventional layouts the power to the center rail is quite variable. 

Bill

There's no reason at all why the unit couldn't run from any aux power source, AC or DC for that matter.  I confess that I'm always thinking command operation and I always have 18 volts handy on the center rail.  It just needs a voltage that has a common with the outside rail.

gunrunnerjohn posted:

There's no reason at all why the unit couldn't run from any aux power source, AC or DC for that matter.  I confess that I'm always thinking command operation and I always have 18 volts handy on the center rail.  It just needs a voltage that has a common with the outside rail.

John:

Thanks for verifying that as it will greatly reduce the complexity.  Be prepared to sell out at York!

Bill

Unless I sell out before York!

rtr12 posted:

But I do need to try that SMT stuff one of these days.

SMT boards can be a lot smaller, but it is more trouble to assemble them.  Obviously, for stuff that I sent out to an assembly house I use SMT, things like the Super-Chuffer would be impossibly large with thru-hole parts.

Add me to the list of customers for these new boards.

The talent here never ceases to amaze me! Great job GRJ ! 

GRJ, You beat me too the punch hahaha I have a similar circuit set up on a prototype bread board bench testing it to make sure it works before field testing. We too have a "hodgepodge" of detector circuitry for signaling at the GSME club. Hoping this will clean everything up! 

GRJ, now having a totally blank benchwork, I'm sure I could find places for half dozen of these.  Keep me in mind if you decide to produce.

John, I'd be interested in getting a number of these to replace the relays I'm using now for signals. Nice work!

Seems like a simple, straight forward approach.  I think for this particular product a fully assembled board would be the most marketable. 

JGL

Nice job John!  Is there any provision or perhaps add-on module that would add time delay to this. Thanks.

Bruce

I was just thinking about something like this.

i could use at least a couple or three

i would be interested in a kit, but all built would be fine as well

BruceT47 posted:

Nice job John!  Is there any provision or perhaps add-on module that would add time delay to this. Thanks.

Bruce

I'm surprised it took so long for someone to suggest this.   I confess that thought crossed my mind.  The attraction of this design is that it's simple in concept and execution.  If I started adding features, I don't know what would make the most sense.

Nice John, if you get around to selling them are they going to be as easy as these?

Hi John,

I need a gizmo like your circuit to trigger Rail King crossing gates, Rail King crossing flasher signals, a K-Line shanty signal man, and the signal man inside a Rail King boxcar. If your circuit is rugged enough to handle these loads individually, I will be a customer.

mike g. posted:

Nice John, if you get around to selling them are they going to be as easy as these?

Even easier, all screw terminals for the connections.

Bobby Ogage posted:

I need a gizmo like your circuit to trigger Rail King crossing gates, Rail King crossing flasher signals, a K-Line shanty signal man, and the signal man inside a Rail King boxcar. If your circuit is rugged enough to handle these loads individually, I will be a customer.

The relay contacts on these are rated at 2 amps.  It was a tradeoff on price/size/power handling.  Higher capacity relays are larger and, have more coil current, and are typically more expensive.  I suspect most of the stuff you mention shouldn't need 2 amps, but I'm not sure of everything.

Good morning John

I have never done isolated rail thing, how long does the isolated rail have to be and how far from the grade crossing protection can it be ? Also will this work with Z Stuff ?

thank you 

Clem k

I have been using these similar boards. Unfortunately, the company no longer exists. With the jumpers they have endless possibilities: timers, flashers, one on one off combos, etc. I have four and one for lighting also. They still need an isolated rail input, so that is why I was looking into the ITAD possibilities for track I didn't want to rip up. They make hooking up signals and gates much easier, as would John's. With diodes on the inputs, I can hook more than one on the same relays, like a signal bridge and crossing gates for two track operation. Great idea, John. Your little gray cells are always at work.

gunrunnerjohn posted:

rtr12 posted:

But I do need to try that SMT stuff one of these days.

SMT boards can be a lot smaller, but it is more trouble to assemble them.  Obviously, for stuff that I sent out to an assembly house I use SMT, things like the Super-Chuffer would be impossibly large with thru-hole parts.

Probably just a try would be sufficient for me, don't think I would want a steady diet of SMT to assemble. Sending those out sounds like a good plan to me. I am amazed at how small things can be made with SMT.

I think the time delay suggestion above might open the door for another module, but I think you are right to keep these simple and as they are. I think there would be many more users that way. You could add the time delay, larger relays, flashers or whatever in separate modules later if the demand is there. A modular system so to speak. 

I finally figured out how to set up individual relays and you have to go and make something simple.  For those using Command, while John is setting these up to draw track power I like to use AUX 14 volts to power the relay(s) and signals.  The reason being I sometimes forgot to shut down the layout, having all of the signals lite up is a visual que to me to shut things off before going to sleep.  Also, if I have an un-powered but occupied section of track (hidden return loop) I don't want any surprises when I power up that section.  A red signal is a sure sign that something is under there.  

Henry

clem k posted:

I have never done isolated rail thing, how long does the isolated rail have to be and how far from the grade crossing protection can it be ? Also will this work with Z Stuff?

The insulated rail can be any length, and it can also be in any position.  It really depends on what you're using the signal capability for.  We have some on the club modular layout that are 20-25 feet long or more, they protect a couple of long tunnels that otherwise we could have issues if someone stops or gets stuck in there.  There's a signal at the entrance, and right past the signal, the insulated rail starts until the tunnel exit  We have some insulated rails that are shorter and are for crossing signals.

rtr12 posted:

Probably just a try would be sufficient for me, don't think I would want a steady diet of SMT to assemble. Sending those out sounds like a good plan to me. I am amazed at how small things can be made with SMT.

 Yep, I only assemble what I need for prototypes or small low count projects.  I also have learned quite a bit about laying out a design for easier hand assembly.  Positioning the parts for easy soldering, and also picking sizes that are reasonably easy to handle are key.  I did one design early on and I specified a size 0201 resistor!  I don't don't know if you realize how small a 0201 resistor really is!  Check the table below!  I couldn't even see to position it using my desk magnifier, and forget about soldering it!  I've since set a self-imposed lower limit of 0603 parts with a preference for 0805 or larger.  I also steer clear of any multi-pin part at a spacing of less then about .095mm between pins.

rtr12 posted:

I think the time delay suggestion above might open the door for another module, but I think you are right to keep these simple and as they are. I think there would be many more users that way. You could add the time delay, larger relays, flashers or whatever in separate modules later if the demand is there. A modular system so to speak. 

 There's always room for a future product that does much more.  If I went that route, I'd probably consider a uP based project to give me a lot of flexibility.  A uP with a nice big jumper field for options would provide lots of flexibility, but it would also carry a higher price tag.  There are always tradeoffs...

Oh goody.  Now I don't have to use your LED lighting module to control a 99 cent eBay relay module for insulated rail sensing!   Now it's all on 1 board!     Video of GRJ's LED board drafted for insulated rail duty in this post.

Looking good GRJ!   BTW take a look at R1 for 14V Aux power operation.  If using the HK19 relay with 400 ohm coil that 620 ohm divider may not provide the minimum 9V spec to trip the 12V relay.

Not sure how many applications need the 2nd pole but just wondering about mixing the 6-position terminal screw strip connector with a 3-pin header.  For the average guy those 0.1" square-pin sockets are a nuisance if you don't have the crimp tool, soldering iron, whatever.

Also, in the creeping-features category, as mentioned earlier one sometimes useful feature of ITADs is the settable time-delay.  Like that optional SMT diode you tossed in for coil flyback (I concur it's not needed), I'm "imagining" a 50 cent optional SMT circuit to insert a settable time-delay.  Would require continuous power so would add another terminal to "ground" (outer-rail).

The three pin header was just to allow use of the second set of contacts.  My feeling was that a vast majority of the users would not need or use them.  The resistor is indeed sized for 18 volts, for a lower voltage, the resistor would also be changed in value.  I suppose I could leave that one unsoldered and include a couple of choices in the package for various input voltages.

Stan, I thought about an "enhanced" model, but I've found this to be pretty useful in our own setting, so I figured I'd float this one first.

I just realized I didn't address the last part of your comment Stan.  I'm using the HK19F-DC12-SH2C with the 720 ohm coil.

Holy mackerel, those 0201 items are only .001" x .002" according to the Google mm to inch calculator. That is TINY! I wouldn't want to try one of those, I will take your advice on that one and stick to the 0603 minimum. I think those little square chips with lots of little tiny fingers on all four sides would be pretty ugly as well. It's amazing they can get the robots to do that stuff. Probably have to pay them more, TINY duty pay. 

I guess on the good side those would certainly allow some very small PCBs.

rtr12 posted:

Holy mackerel, those 0201 items are only .001" x .002" according to the Google mm to inch calculator. That is TINY! I wouldn't want to try one of those, I will take your advice on that one and stick to the 0603 minimum. I think those little square chips with lots of little tiny fingers on all four sides would be pretty ugly as well. It's amazing they can get the robots to do that stuff. Probably have to pay them more, TINY duty pay. 

I guess on the good side those would certainly allow some very small PCBs.

You read that wrong. 0201 is .024 x .012. Inches!

Thats still big. We use 01005 in cochlear implants at work. Thats .016 x .008. There is now a size smaller!

We have to use solder stencils that are .003 etched stainless so that the solder paste isn't higher than the part. 

Lou N

A bit off topic of the thread, but he entire process of using solder paste and pick and place machines is completely fascinating to me.  Here's a good walk-through of the process used for machine-assembled boards: 

For anyone that is interested in electronics, btw, the EEVblog is a wonderful youtube channel with countless hours of content on all aspects of electronics, including basic components and theory, product tear-downs, in-depth theory, and BS-gadget-debunking.  Oh and in case anyone's interested, the circuit being built in the above video is a device that allows more accurate current reading at micro and milli amps when using a digital multi-meter.  

JGL

John,  great idea!  I built several relays using a design that was published in CTT several years ago,  the parts were available from Radio Shack,  your design would be a big upgrade and easier to use.  Since the Radio Shacks near me are no longer, sourcing parts are not easy either.  I like the idea of having kits available for those that need several.

Best wishes,  Dave

JohnGaltLine posted:

A bit off topic of the thread, but he entire process of using solder paste and pick and place machines is completely fascinating to me.

JGL

Those are big parts and really slow machines.  But it demonstrates whats happening. 

Let me know if you ever come to cleveland. I'll show you machines that put down 17000 parts an hour. 

I work here:  www.valtronic.com

Lou N

Hey John,

Great idea to provide this feature in a plug-n-play (of sorts) module. To keep cost down, could you use one of the many 2-relay modules from the bay, which already have LED's, isolation, CONNECTORS (!) and are priced at a little over a dollar ($1.10, $1.35 . . )?

Like this one - and just add your smaller board with a few components and a two-screw terminal block

Good luck and keep us posted.

Alex

Lou N posted:

rtr12 posted:

Holy mackerel, those 0201 items are only .001" x .002" according to the Google mm to inch calculator. That is TINY! I wouldn't want to try one of those, I will take your advice on that one and stick to the 0603 minimum. I think those little square chips with lots of little tiny fingers on all four sides would be pretty ugly as well. It's amazing they can get the robots to do that stuff. Probably have to pay them more, TINY duty pay. 

I guess on the good side those would certainly allow some very small PCBs.

You read that wrong. 0201 is .024 x .012. Inches!

Thats still big. We use 01005 in cochlear implants at work. Thats .016 x .008. There is now a size smaller!

We have to use solder stencils that are .003 etched stainless so that the solder paste isn't higher than the part. 

Lou N

Thanks. You are right, I entered .06 & .03 mm instead of 0.6 & 0.3, was looking right at GRJ's chart when I did it too...duh. I thought that was awfully small.  Those implant parts really are small! I saw some automated PCB manufacturing in the late '80s. Boards were much larger and so were the parts, but still impressive. I can only imagine what they have now. I am going to watch JGL's video to find out.

GRJ,

Would an LED across the relay coil be something that might be wanted, to monitor status? Would probably take up more room and may not be worth the extra space, cost, ruining the design, etc.  Just a thought.

GRJ,

If and when, please count me in for three or four.  Another great idea.

Tons of stuff you could add to it, but I have to stop somewhere.  

I know that there are smaller parts, but when you're trying to put something together with tweezers and a soldering iron, the 0201 is 99% impossible, and smaller stuff gets to 100% impossible.   I find that 0603 and larger are fairly easy to deal by hand.  I'm getting to be an old fart as well, so maybe someone with 22 year old eyes and rock steady hands might be able to use the smaller parts, I'll leave it to such folks to try.

Alex, I've done things that depend on eBay products, but it seems I frequently get burned as they fade out and aren't available anymore.  I'm depending on such modules less and less and sticking to parts that I should be able to source for the foreseeable future.

John,

I agree. I have bought a number of 'modules' of different types from the bay, and when I have gone back for more, the same one I bought may no be available, but a similar one is. I understand how this may be a problem for any sort of production runs that me be done at different times vs just one single run.

Alex

John, I'm pretty ignorant as to exactly what your board can or can't do.   Presently I am using  one outside rail for a common and the other is isolated for signal detection.  If I make both outside rails common, can I use this board to detect train presence?

Art

Ingeniero No1 posted:

I agree. I have bought a number of 'modules' of different types from the bay, and when I have gone back for more, the same one I bought may no be available, but a similar one is. I understand how this may be a problem for any sort of production runs that me be done at different times vs just one single run.

Exactly.  If you're just creating a one-time solution, then the eBay modules are certainly the way to go as they're dirt cheap, far cheaper than I could ever do similar functionality.

Chugman posted:

John, I'm pretty ignorant as to exactly what your board can or can't do.   Presently I am using  one outside rail for a common and the other is isolated for signal detection.  If I make both outside rails common, can I use this board to detect train presence?

Nope, it uses the insulated rail.  The board is very simple, it's just there to eliminate any chatter and also provide totally isolated and sets of contacts to operate signals and accessories.

I did some "thinking" and a brief bench test of my design and came up with a couple of improvements.  I decided to add a low value resistor and the cap, then the larger resistor to drop the voltage for the relay coil.  This allows the capacitor to charge much quicker, and the 100 ohm resistor limits the inrush current to avoid arcing at the wheels.  The capacitor also has a higher voltage so that it has a larger delay before allowing the relay to drop out.  I also used more reasonable sizes for the discrete components and changed the relay flyback diode into a thru-hole part since there was room to do so.

Looking good. I'm in for a few.

Sounds like some good improvements. Where do I place my order? 

Still waiting for my parts to come in so I can check for fit and get a quantity of boards.

Great John! This is way better than what I rigged up.  I will swap mine out for yours.

George

If using the aux power supply option, what is the lowest voltage that would work? 

I'm giving thought to leaving out the one resistor and including a couple different values in the package, so you can "tune" it for the voltage you have.  Being a thru-hole design, it's simple for most anyone to solder the correct value resistor into place for the specific voltage you're running the relay coil at.  Obviously, I don't care what voltage you use on the relay contacts.  The values in the schematic are good for 16-22 volts AC on the track, but if you had 12 VAC, you'd probably want a lower value resistor.

I'm hoping the Chinese get my terminal strips and relays here so I can put a couple together.

I like it John. I'm always playing with relays so there is bound to be a use come up for one sooner or later.  Put me on your e-mail list and I will probably buy a few from you, as long as the price is reasonable of course.

RRmarioE@comcast.net

I'm determined to keep it under $100 Mario.

Count me in for a couple.

gunrunnerjohn posted:

I'm giving thought to leaving out the one resistor and including a couple different values in the package, so you can "tune" it for the voltage you have.  Being a thru-hole design, it's simple for most anyone to solder the correct value resistor into place for the specific voltage you're running the relay coil at.  Obviously, I don't care what voltage you use on the relay contacts.  The values in the schematic are good for 16-22 volts AC on the track, but if you had 12 VAC, you'd probably want a lower value resistor.

I'm hoping the Chinese get my terminal strips and relays here so I can put a couple together.

Well, let's see, 12 volts divided by 720 ohms is about 17 mA.

Why don't you just feed the coil with a CL2 at 20mA and it will adjust itself!

PLCProf probably has a much better idea (I didn't know where the CL2 would go?), but I was thinking of maybe having both resistors with a jumper to switch between them for the different voltages desired, like high or low (resistors are a little easier than the CL2, more my speed).

After looking at the maximized use of space on the board, I imagine that would probably make the board bigger and wouldn't be so desirable? Anyway, just a thought, but I defer to you and PLCProf (and Stan if he stops by).

As a card-carrying member of the Peanut Gallery, I concur with PLCProf's idea of the board taking care of the voltage difference.  I think the CL2 is a sole-sourced component so I'd go with a voltage regulator rather than a current regulator using, say, the widely available 78L12 12V regulator (10 cents) which is half the cost of the CL2.  I don't know what percentage of customers would run this with 18V AC command voltage vs. 14V AC aux/acc voltage vs. 12V DC...but I think shipping multiple resistors is a logistical rabbit-hole.  But it's GRJ's party and it's rude to insult the host! 

Actually, the peanut gallery is curious about the Ground terminal that presumably connects (optionally) to track outer-rail Ground...but then goes nowhere (N/C) on the board. 

I like the CL2 or regulator idea, it simply never occurred to me!  This is why I post, someone always comes up with a great idea at the last minute  Since I only ordered a couple of prototype boards, I think I'll make the change.  Thinking about it, Stan's idea of the regulator is probably a bit cheaper, something to be said for that.  FWIW, since I'll have to wait for boards anyway, the 78L12 is only a little over 3 cents shipped from Hong Kong.  The regulator should allow operation down to 9-10 volts AC, the low current draw of the relay will allow the cap to charge close to peak voltage.  Ding...ding...ding... we have a winner!   I'll still leave a small resistor in front to limit the inrush current, maybe cut it down a bit, 47 ohms or so.

The ground terminal is really just to allow you  to have a junction point.  I had given some thought to a .1" jumper that would connect the ground to the common of one set of relay contacts, but I didn't do it.  That would probably make it more useful.  Let me mull that one over...

Here's what the "new" mods look like.  I added the regulator and the jumper to ground the common of the relay as an option.

Just waiting on the terminal strips to come from China so I can do a trial fit and insure the layout is good.  When that happens, I'll order a batch of boards.

I did add a couple of options, might as well while it's still only on paper.

I expanded the ground center common to ground or power to the common of the #1 set of relay contacts.  If you want to power something like an LED signal directly, you don't need any aux power, just use track power with a diode and resistor.

I also added an off-board expansion capability for the capacitor.  This will allow you to extend the time the relay stays closed after the train leaves the insulated rail.  I got to thinking that maybe someone would want a longer delay...

I'm still following along here. Sounds like some more good additions to the project and not too difficult for folks to use. 

I am glad you say that RTR, I will be bothering you when the time comes! LOL

Certainly in concept it's pretty simple.  It was too simple, so I added a few options to obfuscate it.

LOL way to go John, I think your just making it so I keep reading. But then again I would any ways cause its the only way I am going to learn.

GRJ, if the price is in the $10+/- ball park, put me down for 2 or 3, plaease.

Mike, I think in this case I would bypass the middleman (me) and go directly to the source (GRJ). You'd probably get much better (and correct) answers that way.   Nothing beats talking directly to the design engineer!

That's the goal.  I have all the parts except for the terminal strip, I'm using the Euro style terminal blocks.  In order to keep the price of parts reasonable, I ordered a bunch of these from China, they're not here yet.  All the other parts are in.  I want to get the terminal blocks and verify fit for sure before I order 100 or more boards, it would really bum me out to have another fiasco like the WD Generator issue where I had to scrap the first batch of boards because I was in a hurry to get them.

John,

I think that I can actually use this without messing up to badly.  Please put me down for four or five if you have enough.

John:

Great product!  I may buy one to use a spare for my 40-year 12-volt AC DPDT relays now in use for the last 20 years on their 3^rd layout.

But following up on my comment from a few weeks ago, assuming that you're going to provide a copy of the schematic with each sale, I think it would be helpful to change the nomenclature on the J1 #5 terminal to not say “center rail” for those that run in conventional mode.

I run both command and conventional but some of the conventional-only runners may not understand that it should not be connected to such a variable-voltage source as the center rail of a conventional layout.

Bill

Great thread.....always up for new stuff ...keep me posted for a couple units..thanks

Looks great John. I'd like some of them -- maybe 10.

Gerry 

WftTrains posted:

I run both command and conventional but some of the conventional-only runners may not understand that it should not be connected to such a variable-voltage source as the center rail of a conventional layout.

Valid point, I can certainly cover that in the instruction sheet as well.

John:

gunrunnerjohn posted:

That's the goal.  I have all the parts except for the terminal strip, I'm using the Euro style terminal blocks.  In order to keep the price of parts reasonable, I ordered a bunch of these from China, they're not here yet.  All the other parts are in.  I want to get the terminal blocks and verify fit for sure before I order 100 or more boards,it would really bum me out to have another fiasco like the WD Generator issue where I had to scrap the first batch of boards because I was in a hurry to get them.

John:

Take your time and avoid the stress of "fiascos".  All of us are currently operating without these boards so we can all wait a little longer.

That said, when all the parts are finalized, if these units are still in the ballpark of $10 -- and they help to avoid chattering -- you can sign me up for 5 to start.

If you need help writing an installation manual that can be followed by most users, I'd be happy to volunteer to proofread whatever you come up with.  I wouldn't know a "CL2" if it walked up and handed me a grilled cheese sandwich, but I do know how to write procedures that anyone can follow...

Thanks for taking the initiative on this.

Steven J. Serenska

This looks great John. Count me in for six.

I'm currently using AC ice cube relays and bases and would gladly move to your solution.

Mike

So if I understand the device correctly, will it operate trackside accessories like a crossing gate, etc. ?  I run strictly conventional.  I use an insulated outside rail to operate such accessories.  

Dan Padova posted:

So if I understand the device correctly, will it operate trackside accessories like a crossing gate, etc. ?  I run strictly conventional.  I use an insulated outside rail to operate such accessories.  

Great minds, Dan, as I just saw this thread and your post anticipated my question.

Same for me. PW style layout with gateman, switch tower, & the usual signals (block, semaphore, etc.) all activated by insulated track sections. They work OK but the action can be balky at times. Also lots of rail sparking.

So wondering if these devices would be suitable for that purpose.

johnstrains posted:

Dan Padova posted:

So if I understand the device correctly, will it operate trackside accessories like a crossing gate, etc. ?  I run strictly conventional.  I use an insulated outside rail to operate such accessories.  

Great minds, Dan, as I just saw this thread and your post anticipated my question.

Same for me. PW style layout with gateman, switch tower, & the usual signals (block, semaphore, etc.) all activated by insulated track sections. They work OK but the action can be balky at times. Also lots of rail sparking.

So wondering if these devices would be suitable for that purpose.

Judging from the first post in this thread that is the designer's (GRJ) intended purpose for this device.  I made a similar one from an article in CTT, parts from now mostly defunct Radio Shack.  Works well but it's kinda bulky and not as sophisticated as John's version (no choke, no inrush protection).  I'm hoping a kit will be made available, I could use a few more.

Pete

Dan Padova posted:

So if I understand the device correctly, will it operate trackside accessories like a crossing gate, etc. ?  I run strictly conventional.  I use an insulated outside rail to operate such accessories.  

Yep, that's the intended audience.  It will provide bounce-free activation of anything triggered by the insulated rail.  It will be powered by the track or by an aux power supply, your choice.  For command, it's obviously easy to power it with track power, for conventional, you will likely want to use an aux fixed supply.

A key design target is to minimize the arcing at the wheels when activating the attached accessories.

John, I assume it can take 22 VAC input.   Will it be spark suppressed to handle old Lionel solenoid-operated accessories?

If so,I'll take 6.

Spark suppressed?  The relay handles the load, what kind of suppression are you looking for?  You won't see any sparks.

gunrunnerjohn posted:

Dan Padova posted:

So if I understand the device correctly, will it operate trackside accessories like a crossing gate, etc. ?  I run strictly conventional.  I use an insulated outside rail to operate such accessories.  

Yep, that's the intended audience.  It will provide bounce-free activation of anything triggered by the insulated rail.  It will be powered by the track or by an aux power supply, your choice.  For command, it's obviously easy to power it with track power, for conventional, you will likely want to use an aux fixed supply.

A key design target is to minimize the arcing at the wheels when activating the attached accessories.

It looks like I've misunderstood the purpose.  

I know yours works with an insulated rail, but to put this in Lionel contactor terms, it seems like your device would be a substitute for a 145C contactor where the passing train completes a circuit (e.g., to activate a gateman).  This would be versus a 153C contactor where the passing train either interrupts a circuit and/or diverts the current from one circuit to another (e.g., from the green bulb to the red bulb on a 153 signal).

Or am I completely misunderstanding and your device can do both?

Please advise.

Thanks.

Steven J. Serenska

gunrunnerjohn posted:

WftTrains posted:

I run both command and conventional but some of the conventional-only runners may not understand that it should not be connected to such a variable-voltage source as the center rail of a conventional layout.

Valid point, I can certainly cover that in the instruction sheet as well.

Well, as long as you didn't choke on the regulator idea.....

I didn't work the math all the way through, but if you were to use a 6 volt relay and 6 volt regulator, or a 5 volt pair, you would have a truly universal design that would work on 6 - 24 vac, both conventional and command. You would need to go bigger than a TO-92 regulator to handle the dissipation at the higher end of the voltage range. I would expect a TO-220 package would handle it without a heatsink, you need about 1 W rating for continuous operation of a 6 volt 180 ohm relay with 24 VAC in.

Just sayin'

Well, one of the reasons I stuck with this design was that I could get the relays really cheap, I couldn't fine a similar 5V relay at anything like the same price.  However, the point is well taken, perhaps I can change the regulator package and then see if I can round up the 5V relay at a later time.

I would take at least 4 of these! Kit or assembled, either way is fine. Just in case I haven't said that already, I don't want to miss the boat with all the requests coming on so quickly. 

Also, no hurry on my account either as other have said, I can wait until you are ready at your leisure. I know you have a lot of other things to do too.

Serenska posted:

It looks like I've misunderstood the purpose.  

I know yours works with an insulated rail, but to put this in Lionel contactor terms, it seems like your device would be a substitute for a 145C contactor where the passing train completes a circuit (e.g., to activate a gateman).  This would be versus a 153C contactor where the passing train either interrupts a circuit and/or diverts the current from one circuit to another (e.g., from the green bulb to the red bulb on a 153 signal).

Or am I completely misunderstanding and your device can do both?

 

It will certainly replace the 153C contactor, it has both a normally-open and normally-closed set of contacts.  In point of fact, there is actually another set of contact on the board to a second header so you get even more functionality.  Also, there is a jumper option to allow you to apply either power or ground to the center common contact, or you can isolate it and connect your own.  It more than replaces the 153C.

PLCProf posted:

Well, as long as you didn't choke on the regulator idea.....

I didn't work the math all the way through, but if you were to use a 6 volt relay and 6 volt regulator, or a 5 volt pair, you would have a truly universal design that would work on 6 - 24 vac, both conventional and command. You would need to go bigger than a TO-92 regulator to handle the dissipation at the higher end of the voltage range. I would expect a TO-220 package would handle it without a heatsink, you need about 1 W rating for continuous operation of a 6 volt 180 ohm relay with 24 VAC in.

How does this look?  I took it a step farther.  Since I already have the components for the first run, I left the TO-92 regulator in place, but I put a TO220 regulator in parallel with it.  I can use either one, so if I find the 5V relays, I can drop them in, all other functionality remains the same.  The only thing that changes is the regulator and the rating on the relay coil.

gunrunnerjohn posted:

Dan Padova posted:

So if I understand the device correctly, will it operate trackside accessories like a crossing gate, etc. ?  I run strictly conventional.  I use an insulated outside rail to operate such accessories.  

Yep, that's the intended audience.  It will provide bounce-free activation of anything triggered by the insulated rail.  It will be powered by the track or by an aux power supply, your choice.  For command, it's obviously easy to power it with track power, for conventional, you will likely want to use an aux fixed supply.

A key design target is to minimize the arcing at the wheels when activating the attached accessories.

Thanks John.  It sounds like something I can definitely put to work.  By the way, your signature mentions Hennings.  Will you be selling these through them or directly ?

It will certainly replace the 153C contactor, it has both a normally-open and normally-closed set of contacts.  In point of fact, there is actually another set of contact on the board to a second header so you get even more functionality.  Also, there is a jumper option to allow you to apply either power or ground to the center common contact, or you can isolate it and connect your own.  It more than replaces the 153C.

John:

Thanks.  All understood. The way people were talking about using it to run gatemen, semaphores, and gates, I got confused.

Using this for block control (like a 153C or an insulated rail connected to a chatter-y old fashioned relay) is what I'm after.  I'll take 5 whenever you finalize the design and production.  Like RTR12, there's no rush on my part.  Please take the time to get it right at a pace that's enjoyable for you.

Thanks for doing this.

Steven J. Serenska

gunrunnerjohn posted:

Serenska posted:

It looks like I've misunderstood the purpose.  

I know yours works with an insulated rail, but to put this in Lionel contactor terms, it seems like your device would be a substitute for a 145C contactor where the passing train completes a circuit (e.g., to activate a gateman).  This would be versus a 153C contactor where the passing train either interrupts a circuit and/or diverts the current from one circuit to another (e.g., from the green bulb to the red bulb on a 153 signal).

Or am I completely misunderstanding and your device can do both?

 

It will certainly replace the 153C contactor, it has both a normally-open and normally-closed set of contacts.  In point of fact, there is actually another set of contact on the board to a second header so you get even more functionality.  Also, there is a jumper option to allow you to apply either power or ground to the center common contact, or you can isolate it and connect your own.  It more than replaces the 153C.

PLCProf posted:

Well, as long as you didn't choke on the regulator idea.....

I didn't work the math all the way through, but if you were to use a 6 volt relay and 6 volt regulator, or a 5 volt pair, you would have a truly universal design that would work on 6 - 24 vac, both conventional and command. You would need to go bigger than a TO-92 regulator to handle the dissipation at the higher end of the voltage range. I would expect a TO-220 package would handle it without a heatsink, you need about 1 W rating for continuous operation of a 6 volt 180 ohm relay with 24 VAC in.

How does this look?  I took it a step farther.  Since I already have the components for the first run, I left the TO-92 regulator in place, but I put a TO220 regulator in parallel with it.  I can use either one, so if I find the 5V relays, I can drop them in, all other functionality remains the same.  The only thing that changes is the regulator and the rating on the relay coil.

Looks good, but I found an envelope with a clean back, so I did some more figuring.

Because the load current on the regulator is fixed, we can cheat a little. Keep the 6 volt TO-92, but shunt it with a 910 ohm 1 watt resistor. 

6 volt 180 ohm coil draws 33 mA. Because the voltage is fixed the current is fixed.

Let's say the DC supply is worst case 36 VDC, the voltage across the regulator and resistor will be 30 volts, but 30/910 = about 32 mA, so the resistor will handle almost all the current, the regulator will do nothing, hence, no dissipation. Resistor will dissipate about 1 watt

At say 24 VDC, the drop across the regulator and resistor will be 18 VDC, so the resistor will pass about 20 ma, the regulator will make up the 13 mA difference and dissipate about 230 mW.

Worst case will be at 21 VDC, the regulator dissipation will be about 250 mW at that point.

This all assumes that resistors are less expensive than regulators.

Or so the henscratch sez.....

What I was thinking of GRJ, was the point-eating large sparks that are generated when breaking a circuit involving old '40s & 50s Lionel solenoid devices

Not to hijack the thread but it is relevant to some of the discussion above.

Anybody know how to properly activate a 153 block signal using an insulated track section? I have it set up on my layout (Lionel PW tin track style) and it works -- sort of. I have green light on and when train crosses into insulated track section the red light comes on but green stays on. I know only one light is supposed to be on at a time.

I refuse to use those pesky contactors and would like it to work with insulated track.

Of course, GRJ's product could make it all moot!

Johnstrains.  You have to have a SPDT relay, to open the green light circuit and close the red.  An insulated track will operate a 151 semaphore without a relay, but it might affect a command signal since the solenoid coil is tied into the outside rail.

Thanks, RJR. I had a feeling I might need a relay. Played around with different connections to the three posts on the 153 but could never get it to work properly.

johnstrains posted:

Thanks, RJR. I had a feeling I might need a relay. Played around with different connections to the three posts on the 153 but could never get it to work properly.

GRJ's product is the perfect solution for your application.

Bill

PLCProf posted:

Looks good, but I found an envelope with a clean back, so I did some more figuring.

Because the load current on the regulator is fixed, we can cheat a little. Keep the 6 volt TO-92, but shunt it with a 910 ohm 1 watt resistor. 

6 volt 180 ohm coil draws 33 mA. Because the voltage is fixed the current is fixed.

Let's say the DC supply is worst case 36 VDC, the voltage across the regulator and resistor will be 30 volts, but 30/910 = about 32 mA, so the resistor will handle almost all the current, the regulator will do nothing, hence, no dissipation. Resistor will dissipate about 1 watt

At say 24 VDC, the drop across the regulator and resistor will be 18 VDC, so the resistor will pass about 20 ma, the regulator will make up the 13 mA difference and dissipate about 230 mW.

Worst case will be at 21 VDC, the regulator dissipation will be about 250 mW at that point.

This all assumes that resistors are less expensive than regulators.

Or so the henscratch sez.....

If the resistor dissipation is 1 watt, I'd want at least a 2 watt resistor.  They're kinda large.  Since I can probably find the TO220 part in the 15 cent range, it makes more sense to me to just do this the "clean" way and not have to balance the currents.

It's is an interesting thought, shunting some of the power does kinda' make sense. However, it appears that the TO220 package should be able to handle the 40ma at 30 volts without going into thermal shutdown, but it might be a bit "toasty".   Perhaps going with a higher value resistor across it would make more sense, at lower voltage drops, I don't need tune it as fine as almost all the current.  I'm thinking maybe provisioning for a 1W resistor and then see what makes sense at expected operating voltages.  I'm thinking more along the lines of maybe 1500 ohms of shunt resistance, I don't have to dissipate all the current, just some of it.

Another point to consider is the resistor has to be large enough where if the input voltage is maybe 15 VDC from a low AC voltage of perhaps 10 volts, I don't want the regulator be totally out of the picture.

You know John, I don't understand a thing you just posted, but if it is going to make a simple guys life like mine easier, I want in for some also!

gunrunnerjohn posted:

PLCProf posted:

Looks good, but I found an envelope with a clean back, so I did some more figuring.

Because the load current on the regulator is fixed, we can cheat a little. Keep the 6 volt TO-92, but shunt it with a 910 ohm 1 watt resistor. 

6 volt 180 ohm coil draws 33 mA. Because the voltage is fixed the current is fixed.

Let's say the DC supply is worst case 36 VDC, the voltage across the regulator and resistor will be 30 volts, but 30/910 = about 32 mA, so the resistor will handle almost all the current, the regulator will do nothing, hence, no dissipation. Resistor will dissipate about 1 watt

At say 24 VDC, the drop across the regulator and resistor will be 18 VDC, so the resistor will pass about 20 ma, the regulator will make up the 13 mA difference and dissipate about 230 mW.

Worst case will be at 21 VDC, the regulator dissipation will be about 250 mW at that point.

This all assumes that resistors are less expensive than regulators.

Or so the henscratch sez.....

If the resistor dissipation is 1 watt, I'd want at least a 2 watt resistor.  They're kinda large.  Since I can probably find the TO220 part in the 15 cent range, it makes more sense to me to just do this the "clean" way and not have to balance the currents.

It's is an interesting thought, shunting some of the power does kinda' make sense. However, it appears that the TO220 package should be able to handle the 40ma at 30 volts without going into thermal shutdown, but it might be a bit "toasty".   Perhaps going with a higher value resistor across it would make more sense, at lower voltage drops, I don't need tune it as fine as almost all the current.  I'm thinking maybe provisioning for a 1W resistor and then see what makes sense at expected operating voltages.  I'm thinking more along the lines of maybe 1500 ohms of shunt resistance, I don't have to dissipate all the current, just some of it.

Another point to consider is the resistor has to be large enough where if the input voltage is maybe 15 VDC from a low AC voltage of perhaps 10 volts, I don't want the regulator be totally out of the picture.

Whatever - Just giving you some food for thought. I'm a theory guy, you and Stan are into the practicalities 

I sized the resistor so the "maximum" dissipation of the regulator was a minimum. At the end of the day, you need to dissipate the same power somewhere. 

Another point to consider is the resistor has to be large enough where if the input voltage is maybe 15 VDC from a low AC voltage of perhaps 10 volts, I don't want the regulator be totally out of the picture.

The current through the resistor is proportional to the voltage drop across the regulator! As the supply voltage goes down, the regulator current goes up! Using my example, with an 8 volt supply, the regulator drops 2 volts, so the resistor current is about 2 mA and the regulator would handle 31. Counter intuitive, but that's the way the math works out!

Yep, I was thinking "backwards" with the resistor.  I think I've settled on allowing for either regulator type, that should be flexible enough to cover most bases.  If you really felt the need for a resistor, just poke it in the holes of the missing regulator, problem solved.

Hi John

I am sure I can use these at the club & home layout.  For starters I will take 4.

Thanks,

Bob D

NJ-Hi railers

Here's some input, John, that might be helpful.  Depotronics has been gone for some time.  He had a lot of control boards that preceded train operation before all the electronics were placed in the locomotive/engines.   This board was kind of a universal isolated rail board that could be altered for several track side accessories.  I was able to match it to MTH crossing gates. You can see some of the alterations that could be done for flashers and other control. 

Green is isolated rail section.  Accessory and track common have to be the same.  If using a relatively short isolated rail the axle spacing, front to back of the car can cause the gate to go up and down.    Blue, yellow and white, match to the MTH crossing gate.   Never replaced the 2 amp fuse. 

Appears to be a 12 volt DC relay  Most of the board a bridge rectifier?? 

mike g. posted:

You know John, I don't understand a thing you just posted, but if it is going to make a simple guys life like mine easier, I want in for some also!

Well said Mike, my thoughts exactly!!

John,

I haven't had the opportunity to read the entire thread but from the little I have read I will take several of these. When will they be available? Will the relay handle the current of a twin coil switch machine at  about 20 volts? I realize the device is not designed for this purpose but I thought I would ask.

Dave Drake posted:

I haven't had the opportunity to read the entire thread but from the little I have read I will take several of these. When will they be available? Will the relay handle the current of a twin coil switch machine at  about 20 volts? I realize the device is not designed for this purpose but I thought I would ask.

Dave, the relays are rated at 2A, they should handle most switch machines.

Mike CT posted:

Here's some input, John, that might be helpful.  Depotronics has been gone for some time.  He had a lot of control boards that preceded train operation before all the electronics were placed in the locomotive/engines.   This board was kind of a universal isolated rail board that could be altered for several track side accessories.  I was able to match it to MTH crossing gates. You can see some of the alterations that could be done for flashers and other control. 

Green is isolated rail section.  Accessory and track common have to be the same.  If using a relatively short isolated rail the axle spacing, front to back of the car can cause the gate to go up and down.    Blue, yellow and white, match to the MTH crossing gate.   Never replaced the 2 amp fuse. 

Appears to be a 12 volt DC relay  Most of the board a bridge rectifier??

There are a LOT of parts that aren't populated there, so I suspect there was a lot of documentation that accompanied that.  Was it something you assembled, or did it come as an assembled board?  If it was assembled, did they just have a bunch of different "products" that populated different sections of the board?

I have given some thought to a "deluxe" version with a variety of options, but of course that really runs up the cost and complexity.  If I were to do something like that, it would be microprocessor based and I'd probably go with surface mount parts and likely a jumper field to select operating options.  I'd also probably use a socketed PIC chip so that field updates wouldn't require sending the whole board back, similar to how I did the Super-Chuffer.

Every time I do some little project like this, I learn a lot that can be applied in the future.  The primary purpose of talking about them here during development is to take advantage of all the great minds that contribute to the design.

gunrunnerjohn posted:

Dave Drake posted:

I haven't had the opportunity to read the entire thread but from the little I have read I will take several of these. When will they be available? Will the relay handle the current of a twin coil switch machine at  about 20 volts? I realize the device is not designed for this purpose but I thought I would ask.

Dave, the relays are rated at 2A, they should handle most switch machines.

Mike CT posted:

Here's some input, John, that might be helpful.  Depotronics has been gone for some time.  He had a lot of control boards that preceded train operation before all the electronics were placed in the locomotive/engines.   This board was kind of a universal isolated rail board that could be altered for several track side accessories.  I was able to match it to MTH crossing gates. You can see some of the alterations that could be done for flashers and other control. 

Green is isolated rail section.  Accessory and track common have to be the same.  If using a relatively short isolated rail the axle spacing, front to back of the car can cause the gate to go up and down.    Blue, yellow and white, match to the MTH crossing gate.   Never replaced the 2 amp fuse. 

Appears to be a 12 volt DC relay  Most of the board a bridge rectifier??

There are a LOT of parts that aren't populated there, so I suspect there was a lot of documentation that accompanied that.  Was it something you assembled, or did it come as an assembled board? Board was assembled and had a dedicated purpose, but there were other similar looking boards, at the time, with different purposes, that used the unpopulated openings.

If it was assembled, did they just have a bunch of different "products" that populated different sections of the board?  Yes. Basic input all boards was an isolated rail section.  

I have given some thought to a "deluxe" version with a variety of options, but of course that really runs up the cost and complexity.  If I were to do something like that, it would be microprocessor based and I'd probably go with surface mount parts and likely a jumper field to select operating options.  I'd also probably use a socketed PIC chip so that field updates wouldn't require sending the whole board back, similar to how I did the Super-Chuffer. 

Every time I do some little project like this, I learn a lot that can be applied in the future.  The primary purpose of talking about them here during development is to take advantage of all the great minds that contribute to the design.

Best wishes with your project,

Mike CT.

John, another question if I may.  I have one of the MTH operating passenger stations.  Can I use your device to operate it.  This may have been covered in this thread, but I'm too tired at the moment to read back through it all.  

Dan, I can't imagine why it wouldn't work fine in that application.

Another day, another update.  In a different discussion, it was discovered that an LED being powered from track power was enough to trigger a malfunction in the Hotbox Reefer due to introducing a small DC offset.  Given that fact, I decided I better rethink my method of triggering the relay on this project!  I decided to go to a bridge rectifier to eliminate imparting a DC bias on the tracks when the relay is energized.  Also, since it's deemed that we need a TO220 for the 5V coil version (as soon as I get a good deal on relays), I figured I might as well bite the bullet and just make it standard.  That way I can change the relay and regulator while still using the same PCB layout.  I'll just have to eat the $7 I paid for the 200 LM78L12 regulators.

John, I have to give you and people like you credit.  That electronic stuff is intimidating to me at least.  As many times as I have tried to grasp it and was given explanations, I still come away with my head spinning.  

My knowledge of electricity is more in the practical aspect such as wiring a house.  I used to ask the electricians on the job, what's so hard about your work.  Only three things to know, black to black, white to white and payday is Friday.....LOL

I missed payday???

Dan Padova posted:

I used to ask the electricians on the job, what's so hard about your work.  Only three things to know, black to black, white to white and payday is Friday.....LOL

I used to tease them about having to have a '+' or a '-' on the handles of their screw drivers (Klein's).  They were always more than willing to return the compliments when the chance arose. I was fortunate to have worked with a really good bunch of folks over the years. We always had a little fun here and there. 

gunrunnerjohn posted:

... a different discussion, it was discovered that an LED being powered from track power was enough to trigger a malfunction in the Hotbox Reefer due to introducing a small DC offset...

I searched "hotbox reefer" in OGR and couldn't find the discussion.  That is really curious if you're talking the few mA of an LED?!   Even a bridge-rectifier driving a constant load will draw uneven "positive" and "negative" currents due to mis-match in the diodes within its bridge.  I'd like to read more. 

Here you go Stan, you had to search for Hot Box Reefer.

ERR Mini Cmd ACC in Hot Box Reefer not working properly...

If the mismatch of diodes in the bridge causes a problem, then the circuit that's affected is working too hard!

John,

I believe the full-wave bridge is an excellent idea. I was going to suggest it earlier on, but wasn't sure how the extra cost would be justified - now you have. I really don't know enough about the 'other systems' that may co-habit the layout, or how they may be affected by our own added electronics.

Alex

PS. I know several people have offered to test your devices, but in any case, I would also be glad to do so on my layout running TMCC-Legacy-DCS.

Alex, I added the 22uh choke initially and now the bridge with the hope that any signal impact would be minimal running with this powered by the track.

I went with the simple diode at first because it's sufficient, and I figured it had a smaller footprint, not to mention it is cheaper.   However, after the discussion with Chuck on the LED polarity affecting the Hot Box Reefer the other day, I had second thoughts about introducing any DC bias on the track.  The 5V relay will draw 40ma, more than an LED, and the inrush current through the 47 ohm resistor will be more, might as well err on the side of safety.

I also decided to ditch the TO-92 regulator and just have the TO220 for the 12V version (I already have 200 relays!) and later the 5V version (still looking for the proper spec relays at a good price).  I figure the TO220 will not see more than 30V on the input at up to 22 volts track power, at 65 C/W and a 30C ambient, I'm still well within the ratings of the bare TO220 package with no heatsink at 5V output.  Obviously, at 12V output for the higher voltage relays, I have less of a heat issue.  Sure, the regulator will be somewhat toasty, but as long as you aren't grabbing it, all is well.   Also, the power dissipation is only when the insulated track is energized.

I think the Hot Box Reefer issue is a red herring.  My interpretation is that's a situation where the attached load affect its own operation.  If turning on the positive-only LED on one reefer then affected behavior on another reefer then that would be the relevant concern wrt this relay board.  Anyway, it's a moot point if you're going with a bridge on this circuit.

Years ago I recall seeing published guidelines on track voltage behavior for 3-rail O-gauge which was remarkably detailed including allowable DC offsets and what should or should not trigger a whistle/bell.  If a piece of rolling stock (like this Hot Box Reefer) induces such a DC offset onto the main track voltage then it is not meeting the guideline.  OTOH if turning on a LED in a piece of rolling stock tricks itself into "thinking" a DC offset exists (when there is none) on the main track voltage then that's just a poor design.  In my opinion of course...

Could be, but I like to cover any possible, even if imagined, problem before I get a ton of negative feedback on the board.

I was truthfully quite surprised that turning around the LED fixed the problem, that wouldn't have even come up on my radar before Chuck mentioned it as the solution.  I assumed that something external to the car was unbalancing the load a bit more significantly and causing the issue.

Additionally, I was wondering why a previous car didn't have the issue since I had wired the LED the positive cycle direction. And I remember now that I had not put a diode in series with the LED back then, just the one I had the issue with. So that added .5 or so vdc apparently made the difference. 

Hard to say what is happening there, I figured it was worth making sure it didn't happen to me.

stan2004 posted:

I think the Hot Box Reefer issue is a red herring.  My interpretation is that's a situation where the attached load affect its own operation.  If turning on the positive-only LED on one reefer then affected behavior on another reefer then that would be the relevant concern wrt this relay board.  Anyway, it's a moot point if you're going with a bridge on this circuit.

Years ago I recall seeing published guidelines on track voltage behavior for 3-rail O-gauge which was remarkably detailed including allowable DC offsets and what should or should not trigger a whistle/bell.  If a piece of rolling stock (like this Hot Box Reefer) induces such a DC offset onto the main track voltage then it is not meeting the guideline.  OTOH if turning on a LED in a piece of rolling stock tricks itself into "thinking" a DC offset exists (when there is none) on the main track voltage then that's just a poor design.  In my opinion of course...

I think this is correct. As GRJ pointed out to me that the way Lionel provided for the car to be turned off, having no TMCC in it as delivered, was to push the Whistle command. So the LED ckt WITH the extra diode was enough to offset and cause retriggering the car electronics. The car I added Mini ACC some time ago had no such issues and I remembered now that I did not put the diode in to "protect" the LED. That made the difference in this latest one. Turning the LED ckt around made it the "Bell" command I guess which does nothing bad to the car electronics.

Truthfully, the fact that the diode is or isn't there shouldn't have made any difference.  The LED would only draw current in one way.  I can see a remote chance that turning the LED around would do something, though the puny current it draws shouldn't have upset the car, that is somewhat surprising.

The boards are ordered, so it should be a week or two and I'll have a box of blank boards.  I have most of the other components, the remaining stragglers should be here soon as well.

John:

How are you going to handle orders?  Should we just email you offline?  Or should we wait until you make a formal announcement?

Thanks.

Steven J. Serenska

I am wondering the same thing about ordering?

I'll figure that out when I actually have all the parts.

John,

Your latest rendition with the full wave bridge rectifier and hefty voltage regulator looks a lot like the classic linear power supply designs. That's all good; you know that will work. Dale H had a similar post on a different forum that follows along these same lines; allowing aux power and the use of a large capacitor to hold the relay coil activated during discharge. It's obvious that you've given this a lot of thought and that is reassuring to those of us who have an interest in the outcome.

I just wanted to give you a thumbs up for your efforts and I remain curious about how this all gets done; layout, design, adjustments, re-design, prototype and finished product. We await your reports.

Thanks -- Leo

Well, hopefully the "design" is finished, I'm just waiting on the PCB blanks.

GRJ, 

Finally drawn up what I have been bench testing for a few months and I've swapped out the resistor value to what your schematic featured and works even better. Although using through hole components, I think I will switch to SMD on the resistor and choke just to save board space.  J1 and J3 are going to become pcb screw terminals.

The schematic is tiny, so it's tough to read.

I originally started out with all SMT, but then I figured if I didn't do hundreds, I'd have to hand assemble them, thru-hole is easier.  I eally don't care about the size that much, after all they're under the layout as a rule.

That's true it is under the layout! This any better?  

Looks familiar.   I added a couple of options, the jumpering for ground or power to the common relay contact, and connections for a "helper" cap if desired.

Yes it is familiar. Like anything now a days. Vehicles may look different but they all accomplish the same thing. 

This is just a base model to what my master plan is. I plan to add a micro controller to it for more control for signaling. 

At some point I'm going to do a uP based model.  I'd have a number of inputs and outputs so that it could be programmed for a variety of tasks.

When do you think they are going to be available to purchase John?

Paul

I'm waiting on the Chinese to send the blank boards, that's my holdup right now.

John, put me on the mailing list for this.

Lou

Foxchaserr

Well test fitted the relay and screw terminals and going to order a shot of prototypes. As soon as these come in Ill assemble them and interface them with the micro-controller and see what happens. who knows this design might be obsolete in a few months!  

I'm just waiting on the circuit boards for mine, all the other parts are sitting here in readiness.

Anxiously waiting to hear & see in action!

The only thing slowing me down is the FCC, and if having a micro-controller based relay board  (if needed) to comply with section 15 of the FCC guide lines. 

Other than that my board should be in within two weeks. I've got enough things to keep me busy till they come in.

christhetrainguy09 posted:

...

Other than that my board should be in within two weeks. I've got enough things to keep me busy till they come in.

Curious your intent on making any/all of your bare-board designs order-able by a DIY OGR reader using, say, OSH Park boards?

For small quantities and prototypes, OSH Park is great.  However, for quantities, it's a horse of a different color.

Take my board...

Detected 2 layer board of 1.92x0.84 inches (48.74x21.44mm). $8.05 for three, that's $2.68/ea.

If I order at least 60 of them, I can get them for $1.68/ea from OSH Park.  However, if I go to my Chinese board supplier and buy at least 100 of them, I get them for $120 shipped.

It is nice to know I always have an alternate source of boards if I need it.

christhetrainguy09 posted:

The only thing slowing me down is the FCC, and if having a micro-controller based relay board  (if needed) to comply with section 15 of the FCC guide lines. 

Other than that my board should be in within two weeks. I've got enough things to keep me busy till they come in.

 

 

 

 

So that could be about RF radiation limitations?

gunrunnerjohn posted:

For small quantities and prototypes, OSH Park is great.  However, for quantities,it's a horse of a different color.

...

You mean purple? 

Anyway, from what I can tell most guys want an assembled board.  But are you going to offer the links for the DIYer to order $2.68 bare boards...and the full parts list?  The 2 week turn time on OSH Park is probably faster than the 2-3 weeks to get eBay parts from Asia...15 cent 7812 regulator, 3 cent DCS inductor, etc.

On a separate note I'm seeing purple PCB board modules on eBay though nothing for trains from what I can tell!

Stan, I'm considering offering a full parts kit with all the parts needed.  By the time you assemble all the parts individually, it is a lot more trouble and expense I would imagine.  I have also already ordered all the parts for the boards, so what would I do with those?  It's also somewhat of a PITA when I get calls and letters asking me why the parts they bought don't work.

I will post the design on OSH, maybe someone wants to go into competition building these.

Of course, there will also be a fully assembled product as a majority of folks seem to be looking for that.  I learned that with the Watchdog generator, I ended up with no kits, all assembled boards.

I am one of those guys that would need a fully assembled board with instructions. then give me a week or 2 LOL.

Instructions???  We don't need no stinkin' instructions!!!

Ok, just simple stick pictures will do!

Well, the PCB blanks finally arrived.   Here's the first "production" model.

They operate on AC or DC, 12V or more.

Since there really wasn't room to put legends on top, I annotated the connections and jumpers on the bottom.  Note the secondary set of relay contacts that can be connected on the left.

The units are built on 2 x 4 panels, 8 units for a panel.  The first panel shown completed.

Great looking work John, have you had a chance to test drive it? Any thought on a cost per unit?

Yep, I tested the first batch, but I need to build myself a slightly faster method of testing.  I have special spring pins that I use to test stuff like the lighting modules while still in the full panels, and I need to build a similar test rig for these so I can test these track sensors much faster.

They do what was intended, it's a pretty simple function.  I also have to write up instructions to go with them.

If you are now taking orders, I am ready to order 4 of them!!

I will be soon.  Still some prep work to do, but there's light at the end of the tunnel.  Of course, since these are thru-hole designs and my assembly house does SMT, I also have to assemble them, that will keep me busy for a spell...

You certainly don't seem to have any problems keeping busy!!  I will await further instructions.

Do those panels easily split apart after the components are installed?