Tesla vs. Edison: AC or DC relays for block control?

Unless you have operating room clean track and wheel sets, the relays will experience chatter. You can add the capacitor(s) as you've suggested to avoid this condition. And as you have stated, this is easier with DC relay coils.

I would suggest that the cheap relay modules from eBay are SO CHEAP as to make them a compelling purchase. You can afford extras. In my experience, they work as advertised. They also use opto-isolators on the inputs that provide additional protection for the relays. And there is a back EMF diode across the coils that prevent voltage spikes. I doubt that you could do better than that.

While you're at it, you might want to look into AC/DC to DC buck converters. They are easily tuned to the voltage range that you'll need for the relay modules. The one shown below requires a meter to set the output voltage given a steady AC voltage input. Again these are available from eBay.

Wow, thanks, that's really helpful!  I didn't think of using those little relay modules because I think of those as an arduino project part (I was thinking of trying to do this with an arduino a while ago but I don't think it's worth the complication), and also because I wanted to use DPDT relays to also control signals.  But for the price of those modules I can afford to use 2 relays per block.

A few more questions:

It looks like the rectangular blue 10A Songle relays are only rated for 3A or 5A for inductive loads.  Does this mean I would be better off with the 30A (non-rectangular) relays to make sure they last a long time?  Or should I just use a TVS diode to protect the contact side of the relay and use 10A relays?  Because the diode on the relay boards just protects the coil, right?

That buck converter looks very handy for powering the relay boards.  Could I also use a second buck converter (or some other DC supply) to "power" the isolated track sections rather than connecting the isolated rail to my AC accessory terminal and then using a bridge rectifier for each relay input?  I guess the real question is can the DC negative output of the buck converter be "phased" with my transformer by connecting the negative output to common, since that would have to be the return path though the non-isolated outside rail.

How much current should I expect a relay board to draw from the control inputs?  Without a schematic (and without a general understanding of how one would design an optocoupler input for a relay) I don't have any intuition about that.

Thanks again!

OK, now I'm a little more confused about the relay choices.  Since I'm trying to do this without an Arduino to control the relays I need NC contacts (I want power to the center EXCEPT when the next block is occupied, which powers the relay coil via the isolated outside rail).  Looking at the data sheets for the various Songle relays I see (or don't see) on the modules:

SRD-12VDC-SL-C: 10A relay (rectangular, SPDT), rated for 3A for inductive loads, these are the relays I see on most of the modules.  3A probably not enough?  Can I use them at 10A if use a TVS to protect them from voltage spikes?

SRD-12VDC-SL-A: 10A relay (rectangular, SPDT), rated for 5A for inductive loads (best rating I've seen for NC), but I don't see any modules with these.

SLA-12VDC-SL-C: 30A relay (T-shaped, SPDT), NC contacts rated for 0.25 HP at 120 VAC = 1.6 A.  Interestingly, even though it's labeled 30A the highest rating for this relay (NO, resistive load) is only 20A.  Only the NO SPST relay in the SLA series is actually rated for 30A, even though they are apparently all labeled 30A.  Because why would anyone expect the printing on the relay to mean anything?!? 

So, based on the data sheets, it looks like the 10A relay that's actually rated for 3A inductive with a TVS added to the contact side would be my best bet.  Am I missing something (as I said I have never used a relay before), and will this work long-term?  Can I use a TVS to turn a motor into a resistive load as far as the relay is concerned?

The relay modules that I've found are rated at 10 amps at 125 VAC. With the low voltage AC used to run trains, these are perfectly suitable to supply power to the rails; especially when you have a layout divided into different blocks.

The diodes on the relay boards connected to the coils protect against noise from back EMF (electromagnetic force) which occurs when a relay is turned off and the magnetic field of the coil collapses. That will at least help with the electronics on the board itself. But it will not affect the contact side of the relay outputs.

A TVS is always a good idea to prevent voltage spikes on the tracks when you are running modern (electronics based) trains. The relay outputs from one of these modules might be a good spot for those. But you said you're running a conventional layout. TVS diodes won't help you much there.

That buck converter looks very handy for powering the relay boards.  Could I also use a second buck converter (or some other DC supply) to "power" the isolated track sections rather than connecting the isolated rail to my AC accessory terminal and then using a bridge rectifier for each relay input?  I guess the real question is can the DC negative output of the buck converter be "phased" with my transformer by connecting the negative output to common, since that would have to be the return path though the non-isolated outside rail.

I'm not sure what to say when it comes to mixing AC and DC conductors. I try to avoid it myself just because I don't understand it.

You could feed the AC ground from the isolated rail into the inputs of the buck converter to power it up from AC aux power. That would give you a DC output as inputs to the relay module. But there are easier ways to achieve that.

I would want to use a bridge rectifier with a capacitor and maybe a voltage regulator to produce the inputs to the relay module.

The inputs to the relay module require very little current. I've run them straight off an Arduino which is about 20ma.

Thanks, your real-world experience (that these relays work for trains) is really helpful.  Those are exactly the relays I was talking about -- if you look at the data sheet, they are rated at 10A for resistive loads but only 3A for inductive loads (motors) at 120VAC.  As I understand it, the lower rating for inductive loads is because when the relay opens the circuit the energy stored in the magnetic field in the motor coil causes a voltage spike that can fry the relay contacts.  I think a TVS would protect the contacts by limiting the voltage spike from the motor, just like like the diode on the relay board protects the relay coil.  And maybe it's not important in our application since the relay contacts are rated for much higher AC voltage than the <20V we're using.  If you're powering a 120V AC motor you of course don't have as much headroom for voltage spikes from the motor.   In any case, I will give the 10A relay modules a try.

Thanks again for all the help!  I think I'm on the right track now.

I agree that using the TVS diodes on the outputs of the relay module will help to preserve the contacts. And at the same time, derailments (sparks, spikes) will also be dampened.

Good luck! Let us know how things turn out. This forum is a welcoming group.

bajinnova posted:

... I guess the real question is can the DC negative output of the buck converter be "phased" with my transformer by connecting the negative output to common, since that would have to be the return path though the non-isolated outside rail.

How much current should I expect a relay board to draw from the control inputs?  Without a schematic (and without a general understanding of how one would design an optocoupler input for a relay) I don't have any intuition about that.

Here's a diagram I posted on a previous OGR thread (which I can't find).  This shows one way to do the "phasing" between the AC track voltage common and the 12V DC relay voltage.  You need an electrically isolated DC power source such as using a 12V DC-output wall-wart.  These are about $2 on eBay (free shipping from Asia). 

As shown, tie the DC- to the AC outer rail.  Using a 12V relay module with "low" level triggering, a relay is then triggered when a train wheel axle straddles the outer-rail into the insulated trigger section.

The inputs to these relay modules are simply a resistor driving the opto-isolator LED - the current requirement is a few mA to trigger the corresponding relay.  As has been noted, you should be able to get a 1,2,4,8, or 16 channel relay module for less than $1 per relay on eBay (free shipping from Asia).

There are DPDT relay modules but as you pointed out, at $1 per relay (SPDT type) it's hard to resist just using two SPDT relays.  Obviously you just drive both relays with the same input - and since the input current requirement to trigger the relay is small, this should not a problem.

Separately, regarding the chatter issue.  

Again, another diagram I posted on a previous OGR thread on this very topic.  What you can do is add a 10 cent Resistor-Capacitor filter which will demote chatter when a consist enters/leaves an insulated rail trigger section.  That is, when you only have one or two axles in the trigger section you will frequently get intermittent contact and the relays respond in a fraction of a second and hence chatter on/off.  For "low" level trigger, the lower diagram applies.  The "trick" if you want to call it that is the DC input voltage to these relay modules do NOT direct drive the relay coil.  There is a transistor buffer so that you only need a few mA to trigger the coil instead of tens of mA if driving the coil directly.  This means you don't need a huge capacitor across the coil to "hold" the voltage during the brief losses of trigger voltage.  If this makes sense to you I'd have to go back an find the OGR thread where I'm pretty sure I came up with some suggested values.  I do recall if you try to do this with a 12V DC bare relay (no buffer), you need a capacitor in the 100's or 1000's of uF to de-chatter.  I'm thinking it was something like 10's of uF.  The low-value resistor (tens of Ohms or so) is used to limit inrush current to a discharged capacitor so you don't get sparking.

Update: I found this video from this OGR thread showing an 18 Ohm and 47 uF filter.

Example of 10 cent R-C filter.  Note that while you can get qty 50 or qty 100 of a single resistor value for 99 cents (free shipping from Asia), if you're just getting in to all this electronic stuff I recommend a resistor assortment for $2 which gets you a assortment of values.  In this case the assortment does not have an 18 Ohm resistor but it does have 22 Ohm which is effectively the same.

Separately, in re the inductive load derating issue.  For any modern O-gauge engine with electronics, the track voltage does not directly attach to the inductive motor winding.  You have all kinds of components in between - at minimum a bridge-rectifier since "all" engines now use DC motors.  Assuming the engine has any smarts at all - such as reversing capability - there will be transistors that drive the inductive motor and protection components to clamp the wild voltage excursions during interruptions in motor drive current.  These semiconductors are much more sensitive to voltage excursions than a relay contact so be assured these have been clamped before they make it back to your block relay contacts!

Thanks, Stan!  I will have to think some more about how I want to handle the isolated tracks but this gives me a good start.  Hadn't thought about the need for a resistor if a capacitor is added (was just thinking about the RC time scale of the capacitor and coil resistance).  I have a postwar loco, and one from the '90s that still has a Pullmore AC motor, so I will use TVS diodes to be safe.

I ordered one of the buck converters and will check to see whether I can use a common common between the AC and DC sides, if not I will try the wall wart.  I guess I should be able to figure out if the common commons are possible but easier just to test it.

I found an 8 channel relay module on Amazon made by HiLetgo that has screw terminals for the inputs, the PC board silkscreen is in English, and they even provide a partial schematic.  Less than $10 and fast shipping.

It's going to be a little while before I get all of this together (starting with a simple tabletop oval for my son for Christmas so we will be able to run trains while working on the "big layout"), but will try to remember to update this thread when I do.  Thanks again!

bajinnova posted:

...

I ordered one of the buck converters and will check to see whether I can use a common common between the AC and DC sides, if not I will try the wall wart.  I guess I should be able to figure out if the common commons are possible but easier just to test it.

Well, I suppose it might be easier to just test it but I assure you it will be a sacrificial buck module albeit "only" a $1 investment.  Above diagram from a previous OGR post discussing the very issue of mixing DC common with AC common.  As shown, point A (the AC Ground/ common) is NOT the same as point B (DC common).  If you short A and B you will effectively be placing full AC voltage across one of the diodes inside the bridge rectifier.  This will blow up that diode - game over! 

Above is yet another diagram from a previous OGR thread.  I think this kind of represents what you are trying to do with blocks - ignore the "MTH signal" on the right side.  The wall-wart provides electrical isolation (via transformer windings) so that its DC- output is isolated from the AC Grounds/commons from the train transformers which of course have their own isolation windings.  It is this isolation between the train transformers and the wall-wart that allows you to then tie together the wall-wart DC- output and the train transformer(s) AC common output(s).

Thanks!  I know you can't do this with a bridge rectifier by itself, but I haven't found a schematic for this kind of transformerless low-voltage AC to DC buck converter so was hoping there was some magic in there. 

A buck converter uses switch mode technology to vary the power output. You can check Wikipedia here that has some schematics and a lot of calculus.

OK thanks, I was confused by the AC-to-DC part.  I'm guessing that this AC-to-DC buck converter is just the DC-DC type described on Wikipedia with a bridge rectifier and capacitor added on the input side.  A quick search suggests that inexpensive low voltage isolation transformers don't seem to exist, so it looks like a wall wart is a good bet.

A computer power supply will do the job. Not expensive and 300 watts of dc power. 

In reviewing your particular application with "only" 6 relays there is yet another option.

I can imagine situations where it's just plain inconvenient to install a wall-wart adapter which of course requires access to AC-wall-outlet power.  So let's say you really want to use the 14V-16V AC Accessory power output from your train transformer.

As discussed earlier it's that pesky bridge rectifier which stops you from simply combining DC and AC commons when the DC and AC are not isolated.  So if you use so-called half-wave AC-to-DC conversion, you use just a single diode.  This then goes into a generic DC-DC converter which has a common in- and out- as depicted by the dashed line.  This means point A and point B are the same.  But this is OK if the AC Accessory common is the same as the AC track voltage common as it will be (or can be wired to be) on train transformers.

I suggest this assuming we are only talking about a handful of relays so the 12V DC power requirement is relatively modest.  The relays under discussion require about 1/2 Watt of power each.  You would need to add a 25-cent capacitor as shown to provide energy storage during the half-cycle when the lone diode is not providing power (a bridge transfers power on both halves of the AC cycle).  Even with the loose components (diode and capacitor) I think you might be able to do this without a soldering iron by using screw-terminals or terminal blocks or the like.  If this is something you might want to pursue I can provide additional details. 

I might end up with more than 6 relays, if I also use them for signals.  But yes, I am interested in more details about using a half-wave rectifier and shared commons between AC and DC.  And I'm not opposed to soldering, but the more I can do without soldering the more I can get the kids involved.

If you add a 5-cent 1N4003 diode (instead of a bridge-rectifier) in front of a DC-to-DC stepdown converter, you can share the AC and DC common/ground.  The IN- and OUT- of the DC-to-DC converter is the same.  So the DC-to-DC module only receives AC power on the "+" half of the AC cycle.  This means you need to add a capacitor on the input of the DC-to-DC module to store energy to use on the "-" half of the AC cycle.  The task is to select a suitable capacitor based on the application.

Cutting to the chase, I think you'll be fine with a 1000uF 35V capacitor which are less than 20 cents a piece.  The 1N4003 diode is less than 5 cents.  You'd add the diode and capacitor at the input to the DC-to-DC module.  Both the diode and capacitor have polarity so need to be connected accordingly.  The module I show in the picture just happens to have solder-pads on the input to which the capacitor can be easily installed/soldered.  Or, you could get by without soldering using terminal blocks, wire-nuts, or whatever...there are only a few connections.

----------------

Technical details.  So I hooked up a 4-relay module (I don't have an 8-relay module handy for bench testing) to see what kind of DC power is required.  As shown, I used a DC-to-DC converter module that has an integral voltmeter.  I don't know what DC-to-DC module you bought but if it doesn't have an integral meter, any DC voltmeter can be used to set the module output 12V DC.  I powered the module with 16V AC from a train transformer assuming this as a typical AC Accessory voltage.

I measured the current/power required by the relay module with no relays active, 1-relay triggered, 2-relays triggered, etc..  The module draws about 5 mA with no relays.  Each triggered relay adds about 35 mA of current draw.  (Footnote: at 12V, the trigger current is about 1 mA).

I selected a capacitor that maintained 12V at the DC-to-DC module's output when loaded by 1/2 Amp (500 mA).  This supports over a dozen relays all ON at the same time (i.e., presumed "worst case" scenario).  That is, 12 relays = 6 relays 2-times-over to allow 1 relay for block power and 1 relay for signal lights if I understand your thinking.

As an aside, most of these DC-to-DC modules have a capacitor but these tend to be small and insufficient for this so-called 1/2-wave application where input power is only available 1/2 the time.  The module shown has a 100uF capacitor.  As it turns out this was large enough to power 1 active relay.  Any more relays and the 12V output voltage sagged.

And the resistor-capacitor de-chatter (if needed) would be something like this:

Above also shows example of having 2 relays triggered by one insulated rail section for the case of controlling block power with one relay and, say, a signal light with the other.

These are the 12v AC Ice-cube style relays that I use for my signals.

https://www.automationdirect.c...,_plug-in,_3a_-_15a_(78x-z-_qxx-z-h78x_series)/general_purpose,_15a_(781_-z-_782_-z-_783_-z-_784_series)/782-2c-12a

The additional plug-in base offered ($4 each) makes it easy to connect wires.

They have been very reliable, simple and effective.

graz posted:

These are the 12v AC Ice-cube style relays that I use for my signals.

https://www.automationdirect.c...,_plug-in,_3a_-_15a_(78x-z-_qxx-z-h78x_series)/general_purpose,_15a_(781_-z-_782_-z-_783_-z-_784_series)/782-2c-12a

... 

Your link dropped some characters so gives and error.  In any case I put the following together from their website with what I believe are the 12V AC relays and sockets you refer:

So $8.75 for a 12V AC SPDT (15A contacts) and socket.  $9.75 for a DPDT version and socket.  The DPDT version means one relay can handle both the hi-current block power switching (on one pole) and the low-current signal switching (on other pole).  Orders of $49 gets you "free" domestic shipping so these would come much faster than anything from eBay-Asia.

These relays have a built-in LED indicator as well as a push-to-test button.  Nice! 

I wonder though if 12V AC is available as an Accessory AC voltage for the typical layout.  A layout train transformer might only have only a fixed AC Accessory output such as 14V or 16V AC.  Or there may not be a spare throttle that can be dialed down to 12V AC and dedicated to powering relays.  The spec sheet says the maximum coil voltage should not exceed 110% or 13.2V AC for relay power.

Also, I have not experimented with AC-coil relays with respect to contact-bounce or chatter from dirty/intermittent wheel-to-track contact.  The DC-coil approach allows application of inexpensive capacitors for DC energy storage to demote relay chatter.

Thinking a bit more about the topic title, there's another consideration for using 12V DC relays for block control.  Once you have 12V DC available, you can apply low-cost eBay ($1 to $2, free-shipping from Asia) timer modules to add interest to block control operation.

For example, the Lionel #253 timed block signal, #132 station stop, etc. are/were fun accessories that, for example, could stop a train for 15 seconds (or whatever) at a station and then re-apply track power to send it on its way.  The timing was based on a thermostatic bi-metallic switch which would heat up from AC track power heating a resistive wire. 

Here's an OGR thread using 12V DC eBay timer modules, in conjunction with 12V DC relays, to control block power in a sequenced trolley station-stop.  Here's the wiring diagram I suggested.  Note how I used an isolated 12V DC wall-wart.  But now that I've fiddled with the half-wave DC-DC converter, I see I could have also used AC Accessory power as provide the 12V DC with a common AC and DC to allow insulated rail triggering!

The wiring diagram is a bit tedious, but it also uses two of the four relays to implement a latching relay function which can be useful in block power control.  Yes, there are latching AC and DC relays but are typically expensive.  So here we're talking applications where an insulated rail trigger could be momentary (as the train passes by) but you want this trigger to "set" the latching relay to continuously apply block power to some section of track until another insulated rail trigger "resets" the latching relay.

Stan, the relays graz posted and you commented on above are really nice relays, and they are also available in 12VAC versions. I have a few different kinds of these exact relays and they are of good quality, nice features, fit on DIN rail and work well too. I think I may even have some 12VAC versions. Only drawback is their price as compared to the quite inexpensive ebay modules.

If you go to the Automation Direct site and search for 781-1C-12A you should get the 12VAC SPDT version. From there I think you can find the rest of the line. The socket is 781-1C-SKT and is linked at the bottom of the relay page. They have them up to 4PDT in 12VAC. I linked the above part number, but I have had trouble linking to items on this site in the past so it may not work?

Just in case you wanted some more info on these relays...

Nevermind, I see you have found the site and the relays....I totally misread the above...Duh!

Stan, do you happen to have a larger version of wiring diagram image for the timers?

I might have the original diagram which may have more resolution...but before I dig around can I ask what you're trying to do?  That is, I was trying to illustrate how timer relays can enhance block-control applications for not much additional cost...maybe $1-2 more per relay. 

And somewhere in the cloud I recall discussing using one of these 12V DC timer relay modules to implement the classic Lionel 132 Stop Station block control.

I also posted that diagram to show how to cobble together 2 standard non-latching relays into a latching relay which can have useful application in block control as shown in the dual trolley application in the linked thread.

And if you are doing signaling with some of your relays, I've shown in other threads how timer relays can implement timed-aspect behavior such as the red-to-yellow-green delay used in 3-aspect heads.  I just re-posted this video on another OGR thread showing a timer relay module for a PRR type signal

In any case, I looked at that thread and see that it tied AC common to DC+ (instead of DC common).  But before getting bogged down with why and when you'd want to tie AC common to DC+ instead of DC-, let's figure out what you're trying to do and come up with a practical solution to your application!

Thanks, Stan.  I looked at the thread you linked and the version of the diagram there is big enough to read clearly.  The application I have in mind is a trolley with stops and maybe a slow down right before the bumpers (by switching in diodes), that I think would be more interesting than a continuous bump-and-go.  But that's just an idea for the future.

Got it.  When you get closer to implementing the trolley system be sure to continue the dialog.  You would probably only need a handful of relays so the method described above with ~25 cents of components (diode, capacitor) in front of a DC-DC converter module set to 12V DC-output would apply again.

The you'd have 12V DC that can drive ~$1 relay modules that can be directly triggered by insulated rail sections.  And then for $1-2 more you can add timing capability to the relay to stop the trolley at the station for X seconds.

There are so many possibilities!  For example, if you have several trolleys running at once you could put a 10-cent magnet on certain trolleys.  Then a reed-switch on the track could sense that the "express" trolley or the "local" trolley is approaching the station(s).  This would control whether the relay system activates that drops the track voltage (via diode drops).  Thus, the "express" trolley would skip a station but the "local" trolley would slow down and stop for X seconds before starting back up.  Yeah, the guys will tell you to use an Arduino or similar microcontroller and "just" write software.  If this is in your comfort zone then so be it.  And there are off-the-shelf suppliers of trolley "systems" that can do all kinds of stuff if your wallet is fat enough!  OTOH, you can do plenty of neat stuff with 12V DC relay modules, timer-relay modules, and the like for next to nothing in out-of-pocket $. 

Yes, I will be in touch when I'm ready to go with the trolley.  The magnet idea could be fun as well.

An arduino is certainly a possibility, and will probably get used somewhere on the layout just for the sake of writing a little code (a big part of the motivation for the layout is teaching the kids, ages 6, 9, and 11, some new skills) but I think for block control and maybe trolley timing it would be fun to just use relays.  But I think we will at least use an arduino for some low power things that won't require relays, for instance creating a camp fire lit from below by flickering red/yellow LEDs.

Thanks for all the help!  There is a lot of good stuff in this thread.

An Alternate Method (Z-Stuff) ??

I realize this thread is a few months old.
There's much interesting information, much of which goes "over my head".

Will This Z-Stuff Alternative Method Work ??

However, I'm curious to know, if anyone thinks that this "Z-Stuff" method I'm about to describe -- is an acceptable ALTERNATIVE to the above discussion.

If the main objective is to control multiple trains (or trolleys) on the same track, this method seems to require a LOT LESS wiring.

Referring to the below screen capture from the video:

The LOGIC is essentially the same as the classic "insulated rail" method -- where:

• The 1st train is stopped in a "Stop Block"
• A 2nd train approaching from rear, crosses the insulated rail section (or the Z-Stuff detector) and "energizes" the stop block.

But . . . as demonstrated in the video, the Z-Stuff DZ-1012 infrared detector with adjustable time delay, and a Z-Stuff DZ-1008 detector -- can essentially replace the insulated rail section -- with a lot less wiring.

The video that demonstrates this is here:

 https://youtu.be/CiwxULos5IM?t=42

Drawings

James Ingram posted:

...

Will This Z-Stuff Alternative Method Work ??

However, I'm curious to know, if anyone thinks that this "Z-Stuff" method I'm about to describe -- is an acceptable ALTERNATIVE to the above discussion.

 

I wouldn't call it the "Z-stuff" method.  As I see it, you're asking if the Infrared (IR) or optical detection method is an alternative for the insulated-rail method for occupancy detection.  This topic has been exhaustively covered in previous OGR threads - but the answer is a qualified yes. It simply depends on the application but for basic stop-block automatic train control, then of course.  Other companies (Lionel, MTH, etc.) offer off-the-shelf optical occupancy detectors.

Additionally, for the DIY'er, there are OGR threads showing how to roll-your-own optical occupancy detector for, say, $1-2 and then how to mate it with a $1-2 relay.  From what I can tell, the 2 items you reference from Z-stuff would set you back ~$60.