Replacement for DZ-1008 Relay for Ross Switch Powered Rails (04/23/21 Update)

@gunrunnerjohn   

Looks very similar to your "Locomotive Motion Sensor" schematic but with fewer components.  Could this be used to sense motion to turn cab lights on/off and to dim the headlight when stopped?

I hadn't considered that application, but the board is larger so it wouldn't fit quite as well.  I did start with the Locomotive Motion Sensor design and just hack it from there.  I also have a larger terminal block on this one for easy of connecting when you're standing on your head under the layout.

John,   

   Is this project of yours going to be available to the masses and if so by kit or assembled.  I see you are using SMT technology which some of us are not up to speed on.  Maybe through hole also?   It is sort of perfect timing as I saved my last 5 or so switches that require the relay until last and they are not done yet.  If you are making these available I would be interested in a few.  I just ordered two speeders and I guess you are saying they will not go through the switch.  Thanks - let us know.

Yep, my speeders, BEEPS, the Trackmobile, Wienermobile etc. all have issues with the DSS or the long curved switches, even with the relays installed.  When I investigate, I see the rollers firmly sitting on a powered rail, but only one wheel without a tire near a powered rail, and apparently that one wheel loses contact long enough to stop the presses.  The relays did help some paths through the switch, but not consistently.  Larger stuff that was stalling on the DSS passage are now making it across, but that doesn't help the little guys.

As for making them available, I was looking and having issues finding the relay I used.  I have my stock of the relays I used, but they're out of stock.  As for a thru-hole design, I also try to use parts I've already used in another design to minimize the number of parts I have to keep around.  It's already getting hard to keep up with all the parts now.  Taking time to design another board that I'll never build and having to verify the design all takes time.

Maybe it's time you stepped up to some simple SMT designs and tried your hand at building them.

Oh, OK, here's what a thru-hole version would look like.  You still have to contend with the opto, I didn't find one that matched what I needed in thru-hole.  It gets a little bigger at 1.3" x 1.05" of course.

  John,

Yeah - your right,  I should expand my horizons and learn to handle SMT.   JP

You'll have do deal with the opto for this project.  It's easy, big legs and only four of them.

I ordered prototype boards, if all goes well and I don't have any issues, the thru-hole one should work as well, it's based on the same schematic.  I did find that Mouser has some of the thru-hole relays in stock, nobody has the SMT ones right now, so I'm glad I have enough to finish my RR.

I have used Atlas double pole, double throw snap relays with great success on an 11 degree ross cross track. For this purpose they will certainly handle the voltage. If you look at the atlas schematic you can use one pole for + and - then crisscross the wires on the second pole to get just the opposite. You will need to put this relay as a separate switch since it cannot take constant voltage on the solenoid. Atlas showed this in a booklet titled, Fun with relays.

I have Atlas snap relays around, it won't work with the DZ-2500 as I don't have any separate switch.  I need a relay that is continuous duty and is activated by a very low current 5V signal.

I would commit to a bunch of them.

the continual  buzzing on the DZ-1008s  as used on the dbl slip switch is annoying.

The good news is, the relay I use on my board doesn't buzz.

I made sure the "normal" route through my DSS had the relay de energized, but that was just so it wasn't always on and drawing power.  My DZ-1008 relays don't make any noise, they just don't do the full job.  I hear the relays click as I change positions on the double-slip switches, but they're silent after that.

FWIW, I bought seven of the DZ-1008 relays, two were bad right out of the package!  Not exactly a good batting average.  For a little package that has a relay and a handful of components to provide the low voltage trigger, that's a pretty lousy batting average!

Great little project John! Should be the answer to this problem. You may want to contact Steve at RCS about offering these for him to sell with his DSS, and xover switches.

That smt board doesn't look too challenging to build. The unavailability of the relays though is going to be a problem!

Rod

Oh I forgot about the "clicks"  too. I had several that I thought were bad too.

Dennis provided me with some tests to perform on the DZ-1008 to show the relays were working correctly; like hearing the BUZZ and the clicks!.

Well, I assumed they're bad when I measure the contacts with a meter and they don't change when the green line has 5V or is open.  To me that's a 100% test of functionality, and two of them were bad!  I actually don't hear a noticeable buzz from my DZ-1008's, but I do hear the click when the actuate, if they actuate!

I had actually installed the bad relay with the "assumption" that I wouldn't get a second bad one, that will teach me to assume!

@Rod Stewart posted:

Great little project John! Should be the answer to this problem. You may want to contact Steve at RCS about offering these for him to sell with his DSS, and xover switches.

That smt board doesn't look too challenging to build. The unavailability of the relays though is going to be a problem!

Rod

I can't imagine it being a big challenge to build, all the parts are pretty good sized and pretty well spaced.  It's really getting harder to find thru-hole parts for some functions nowadays, so SMT is really going to be the norm.

I did order a few of the thru-hole relays from Mouser, and I might get a prototype run of the thru-hole version once I have the SMT version working.  The thru-hole is the only game in town unless you want to wait for the relays to be available again.

Looks like the thru-hole relay that is available fits in the 1 x 1.05 size that I had for the all SMT version, so that's not bad.  Sometimes with thru-hole, you have trouble routing the fat traces for power.

@Rod Stewart posted:

Great little project John! Should be the answer to this problem. You may want to contact Steve at RCS about offering these for him to sell with his DSS, and xover switches.

I sent Steve a little blurb about why I was building this, but I don't know if he wants to get into the business of supplying this stuff directly.

GRJ-

Cool project as I've had issues with the DZ-1008 as well.  However, I'm curious what switch part the Negatives need to attach to.  Per Dennis's instructions, the 1008 is used to provide + power the straight lead and the curved lead.  Where do the - power connections go to on the switch?

There are two dead rails on a standard switch, the suggested wiring for the DZ-1008 connects one of those to the center rail power and leaves the other one unconnected, depending on the path through the switch.  The one that's not connected to center rail power is the one that the wheels will roll over for that path through the switch.  My additional relay contacts will connect that rail to the outer track for that path.  This will allow small powered units with limited wheel contact to better transition the switches, especially the double-slip switches and the longer switches like the curved switches, #6 and larger, etc.

Thank for the clarification GRJ!  I too would be interested in trying a couple of your new relays once they are ready.

Once I get the boards in, I'm going to stick them on a couple of the problem switches and see if they do the trick.  I'm sure they have to improve things, but I don't know if it's a 100% solution as stuff like the DSS has a lot of places where you still won't have more than one wheel on any live rail.  If that wheel has a traction tire, you may have a big problem!  However, my little track speeders would stop on the DSS, the rollers were energized, but one wheel would be sitting on that dead rail, and no other wheel was on a live rail, so in that case I suspect it would have had a beneficial effect.

In looking at the parts availability, I see that the surface mount relay is out of stock and has long lead times.  Based on the availability of the thru-hole relay, I shifted gears and I'll use that relay, those are still available.

I added two jumpers to do the cross connection between the two sets of relay contacts, this avoids having to wire them separately from the connector. This will wire just like the DZ-1008, power in, two contacts out to the dead rails.

I also enhanced the relay pick with an accelerator cap.  This allows a reliable pick of the relay while having a large enough value resistor to avoid overheating the relay when energized for a long period of time.  This avoids having to have a regulated power supply for the relay.

The 10-pin connector is optional, I suspect many will simply solder to the board for the connections and treat this just like the DZ-1008 relay board.

John

when you start making these I hope you going to sell some.

Much needed for those using DZ switch machines and other machines switching Ross-like turnouts.

I'll make something available, once I get a change to test the design well, I'll go to press.

Got my blank boards back from OSH Park, they having pretty quick turnaround nowadays.  I ordered on 2/27, they shipped on 3/9, and I got them today.

Built two of the 1.0 version with all SMT, no surprise they worked fine on the bench, it's a pretty simple design.  In testing I did note that accelerator cap would be useful if you run the board at 18V, I can increase the resistor in series with the relay to 1K and keep it from getting overheated.  I'll also chop the value of the input filter cap down to maybe 47uf, I don't need nearly that much filtering for just a relay.

The input trigger can be AC or DC from around 2.5 volts to 18 volts, pretty flexible.  It's also opto-isolated, so it can come from anywhere.

Given that all I've done is added the accelerator cap and the two jumpers to make wiring this to switches easier, I'm going to "shoot the dice" and order a batch of boards, with the thru-hole relays, what can go wrong?

What this does for the everyday railroader WHO USES ROSS SWITCHES is allow better control over powering (and not powering) certain isolated rails that can cause either a loss of power or a momentary short when wheels run over those isolated rails.

This replaces and improves on the functions of the DZ-1008 relay; that although nice was a bit of an overgrown after thought for the DZ-2500C switch machines. The DZ-1008 worked well when plugged into the DZ-1000 switch machine that it was intended to support; but not so much when used in conjunction with the DZ-2500C switch machines.

Also, most of the specialty turnout wiring diagrams displayed on both ROSS's and DZ's sites are for the DZ-1000s. It takes a bit of electrical engineering knowhow ( for me trial and error) to interchange those diagrams for the DZ-2500Cs. This relay looks like it makes that interchange a lot easier.

I looked at the combinations of surface mount and thru-hole and decided that it probably makes sense to go with the thru-hole layout.  This allows me to offer a kit of parts that more people can take advantage of.

The thru-hole board ends up being 1.25" x 1.05".

The all SMT version is 1.05" x 1.0", so not a ton smaller.

Click on graphic to expand.

So is this a double pole double throw relay?

GRJ-

One consideration in the relay is the speed that it throws the power after the DZ-2500 throws via non-derail.  If the train hits the switch with a good head of speed, the relay would cause a short because the power wasn’t thrown fast enough.  A simple fix is a longer non-derail section of track into the switch than the switch alone provides.  Sometimes this isn’t possible for various reasons.  

Can you look into a high speed relay for this power throw?  The inherent delays (high-speed non-derail followed by the command from the DZ-2500 to throw the relay) all add up.

Just a request since you’re designing a “better” solution.

@LionelAG posted:

GRJ-

One consideration in the relay is the speed that it throws the power after the DZ-2500 throws via non-derail.  If the train hits the switch with a good head of speed, the relay would cause a short because the power wasn’t thrown fast enough.  A simple fix is a longer non-derail section of track into the switch than the switch alone provides.  Sometimes this isn’t possible for various reasons.  

Can you look into a high speed relay for this power throw?  The inherent delays (high-speed non-derail followed by the command from the DZ-2500 to throw the relay) all add up.

Just a request since you’re designing a “better” solution.

I think you're looking for problems where they don't exist!  Just how fast do you think the train is going to cover the few inches between the non-derailing trip and the fixed rails?  Let's look at the relay specifications first.

We'll take the maximum trip or release time of 5 milliseconds.  This analysis assumes the #4 Ross switches, they have the shortest distance between the non-derailing rails and the fixed rails we're switching.  All the other switches have a longer distance between the two points.  In order for the train to cross the three inches between the middle of the non-derailing trip rails and the start of the fixed rails we're switching power on in five milliseconds, it would have to be traveling at over 1,500 scale MPH.  Even assuming it takes the relay two or three times the max amount of time to trip, we're nowhere near having a response time problem.

@romiller49 posted:

So is this a double pole double throw relay?

Yes.

Believe me when I say I have a repeatable problem using the DZ-2500 and DZ-1008.  Or don't believe me.

Logic execution does take time and you remember a conversation with an (unnamed) friend that said there were issues with the DZ-2500 firmware.  Maybe my problem couldn't be addressed by a hardware change alone and is part of the logic execution.

Like I said, something that I have a repeatable problem to the point I disconnected the DZ-1008 and the problem goes away.  You're re-engineering the solution and its a point to consider.

@LionelAG posted:

Like I said, something that I have a repeatable problem to the point I disconnected the DZ-1008 and the problem goes away.  You're re-engineering the solution and its a point to consider.

Gee, I thought I just considered it!  Five milliseconds maximum activation time seems to cover the "fast relay" comment more than adequately.  I believe it's a pretty fast relay right now, and I doubt you'll find a suitable relay with a significantly faster transfer time.  Also, if the relay operation was instantaneous it wouldn't make any difference in the results, your train isn't going to travel very far in 5-10 milliseconds!

I can't do anything about the DZ-2500 logic delays, and I'm not planning on recreating their non-derailing logic.  It sounds like my solution isn't suitable for your operating environment, but there's not much I can do about that.

Thank you for the kind reply.  I think I’d like to try 2 of your new devices when ready.

I'm getting a set of prototype boards from OSHPark for the thru-hole version, when that is verified, I'll order a quantity of boards and parts.

John I got to the party a little late, but the resistor circled below does not look to be padded. Have I missed something? Is it an extra component not included on the schematic?

Rod

Rod, that's actually an inductor, I increased the pad spacing and the 3D model didn't match the board until I fixed it.  I didn't notice until after posting the pictures.  Is this better?

Whew, yeah, that looks better! Just something did not look right to my eye. It would have come to light had I studied the schematic a little closer. Anyway, all good.

Rod

There are some undefined issues with the DZ-2500Cs and the DZ-1008s. The relay was designed to go with and specifically  "plug into" the DZ-1000s.

I think the DZ-1008s use with the DZ-2500Cs was as I said above "an afterthought"; like well here's a relay!

There's no doubt that the DZ-1008 was designed for the DZ-1000, I have to trim the leads sticking out of them to use them on the DZ-2500's.

I haven't seen any of the delay issues that cause shorts because of slow switching where I used the DZ-1008 relays with the DZ-2500 switch machines on my layout.  I have them on all of the 072/096 curved switches and the two double-slip switches.  I've transitioned the curves switches at 60 scale MPH and the relays switched long before there would be a problem with the switched power to the rails.  It's very unlikely that I'll ever be running faster than that speed, so I think I'm good.

Trying to thread your way through the DSS at high speeds is an exercise in futility in any case, so I certainly don't see the relay speed as being an issue there either.

that Turbo-Train  goes pretty fast not sure the SCALE SPEED is in "ms" though!