realtrax switch with a relay?

How far are your switches from the AIU?

What is your switch power source (command track voltage, accessory AC, something else)?

Considering you have "dozens of conductors" did you try doubling-up or even tripling-up on conductors?  At first glance if you're going to buy additional items, I'd think different wire has got to be less expensive and easier than buying components (relays, capacitors, whatever).

But taking your question at face value, if you do plan to install relays out near the switches and run low-current, thin-gauge control wiring from the AIU, consider the inexpensive eBay relay modules (about $1 per relay and includes screw-terminals).    Here's an example of an 8-relay version; they come in 1,2,4,8,16 relays per module.  It looks like the relay you linked to is about $5 each and requires soldering.

Note that to duplicate or "relay" the momentary-on (1/2 second on-pulse) behavior of an AIU SWitch output port you need 2 relays per switch.

Running the switch from track power.  I did consider using multiple conductors per connection but then it takes 6 wires per switch and that's gonna take a bigger wire than I have.  You are right running the right size wire is the simplest approach but i have a few spots that are a PITA to fish a wire through let alone a dozen pair of 14-16 ga.  The relays you show certainly seem like a better idea than what I had thanks.

I'm missing something here:  

C&O line posted:

After a quick test, as expected, a 22ga wire won't carry enough current to operate the machine.  

How much current do these switches need to work?  I know that folks have routinely used telephone wire to control postwar style switch machines, and they had to draw more current than any modern design, right?  For what it's worth, 22AWG is good for about 7 amps, and will drop about 1.6 volts every 10 feet under a 5 amp load.  At a more likely 1 amp load it's about 1/3 volt drop per 10 feet.  FWIW, if you made your test on each end of a 500' roll of wire, then yes, I'd expect too much voltage drop, only about 2 volts coming through.  

How much voltage and current do these switch machines need, and how long are the wire runs?

JGL

I tried it with about a 50' piece of wire and all the switch machine would do was humm.  MTH recommends 18ga wire in their instructions.  The chart I looked at showed maximum amps for 22ga at 7 amps for chassis wiring and 0.92 amps for power transmission, 18ga was 16A and 2.3A respectively.  I don't know what chassis wiring means but power transmission is what we are trying to do I think.  Somewhere reading I got the idea that it takes 2A to make the switch work properly and that jives with MTHs recommendation and the power transition number.

Do you have an extra or secondary power brick. Track power is your enemy when running numerous switches. You should get a dedicated power brick just for your switches. I've got close to 25 switches running from 1 z1000, the 14v acc screws, and have no issues whatsoever. Using 20ga. 

When I first started laying out track, I had 5 lock ons, and 5 switches running off track power, and experienced what you are describing. That was all powered by 1 z1000 as well. 

I use 18 ga from my bus bar to my switches for the power supply. 

Wow that's a lot of useful information.  What does PTC stand for?  If there is plenty of power available running the switches from track power shouldn't be a problem correct?  That is why I was thinking if the AIU could be used to trigger another relay at the switch the diameter of the wire harness (that we are trying to hide/route) would be a lot smaller.

Yes everything is on command/dcs.  And yes the switches work with the jumpers and controllers in.  Trying to get the  AIU to control them now.

Don't underestimate your input.  the information you gave is enough to work back through and next time do the math myself.  Thank you for the time you spent on this.  I did some homework on the relays and it appears the AIU can relay dc as well as ac.  So the relays Stan found are looking pretty good.  Thanks to him too.

C&O line posted:

I tried it with about a 50' piece of wire and all the switch machine would do was humm.  MTH recommends 18ga wire in their instructions.  The chart I looked at showed maximum amps for 22ga at 7 amps for chassis wiring and 0.92 amps for power transmission, 18ga was 16A and 2.3A respectively.  I don't know what chassis wiring means but power transmission is what we are trying to do I think.  Somewhere reading I got the idea that it takes 2A to make the switch work properly and that jives with MTHs recommendation and the power transition number.

The RealTrax switch requires 10v @1.25 amp per MTH. The track power jumper needs to be removed when connecting AUX power. See attached. Chassis wiring is hooking stuff up.

Messing with the RealTrax switches on the Eagle Scout project, I found that 18 is a little large in stranded for the terminals without a forked terminal end. 20 gauge solid seems to be the sweet spot to keep the connection clean.

The problem is that these are power hogs. 5 switches per Z-1000 brick.

I'm kinda' blown away with the current requirements of the Realtrax switches, that certainly comes from left field.  I would probably think twice about ever using those from track power based on my testing with decoupling tracks with similar power requirements.

I connect decoupling tracks (Fastrack in this case), and powered them from track power at 18V.  They seemed to upset the TMCC/Legacy locomotives with the large noise spike generated.  The power requirement was around 1.5A measured, so they seem to be in the same ballpark as the Realtrax switches.  I'd expect similar issues.  The Fastrack switches sip power, it's hard to measure, but less than 100ma at any point.

gunrunnerjohn posted:

I'm kinda' blown away with the current requirements of the Realtrax switches, that certainly comes from left field.  I would probably think twice about ever using those from track power based on my testing with decoupling tracks with similar power requirements.

I connect decoupling tracks (Fastrack in this case), and powered them from track power at 18V.  They seemed to upset the TMCC/Legacy locomotives with the large noise spike generated.  The power requirement was around 1.5A measured, so they seem to be in the same ballpark as the Realtrax switches.  I'd expect similar issues.  The Fastrack switches sip power, it's hard to measure, but less than 100ma at any point.

I think that's because of the motor and only momentarily when switching. The motor plugs into the side. They're nasty.

MO985 has the solution.

Well, I know why it draws power, I just didn't expect that a modern design would need that kind of power.

gunrunnerjohn posted:

Well, I know why it draws power, I just didn't expect that a modern design would need that kind of power.

I get it. Not so modern though, as MTH rushed to beat Lionel with a molded roadbed track system in 1998/1999?, the development of the switches took an easy route with a coil motor. The system is worthy of a switch redesign.

I guess that explains why the Realtrax has issues, they didn't think a lot of stuff through.

I tend to believe you, I know that the old Lionel tubular switches draw quite a bit of current if you hold them on.

I am guessing this would apply to all twin coil switch machines and is not just due to it being the Realtrax or Lionel brands?

Well, Atlas switches will go up in smoke if you leave the current on them for more than a few seconds, they don't use as much power to switch either.

Thanks, I knew about the up in smoke part, although I haven't yet smoked one of those (fingers crossed).  

I didn't know about their current usage though. I may do some fiddling like PLCProf was doing above. So far, I have been very careful with the Atlas switch machines and the pushbuttons.

That's good detective work.  Since there appears to be no apparent phase shift between voltage and current simple Ohms law applies so coil resistance is about 1.5 ohms?  Should be simple to verify.

Let's see. 11.3V x 6.3A = 70 Watts, applied for 25 milliseconds is ~2 Joules of energy.  So if one tried to do this with Capacitive discharge using 18V AC track voltage to charge a cap to ~25V DC that would be a cap at least 10,000 uF / 35V; kind of bulky, kind of expensive.  It would take that peak current down a notch by requiring several seconds to charge up the cap between firing a switch.  OTOH it's not clear the brief 70 Watt pulse materially affects DCS operation in the OP's situation.

The Lionel O22 switches do draw less current.  From a sample of one, 3.8 ohm coil resistance and 4 amps at 18 volts, and 2.4 amps at 12 volts.

FWIW, the Atlas switch machines are 27.4 one way and 27.8 ohms the other. Haven't hooked any up for other measurements, still cleaning off my workbench so there's room to fiddle. It's a huge mess!

Well, there are electronics that are always consuming power in the Fastrack switch, I suspect that's where it comes from.  Truthfully, I'm surprised it takes an amp to throw it, that tiny motor can't possibly draw that much power!

Using a photo GRJ posted from another thread about FT switches, you can see the AC-to-DC conversion to drive the DC motor vs. the brute-force solenoid coil in the Realtrax turnout.  That explains the non-linear relationship between voltage and current in PLCProf's scope shots.

But it seems the issue at hand is the OP's Realtrax turnout issue.  To which I think even if he uses a separate AC voltage source (rather than track voltage) there's still the issue of voltage drop if all he has is 22 gauge phone cable and 50 foot runs (or whatever) to the AIU.  I think the Capacitive Discharge would work but for the same cost or less I'm still liking eBay 12V DC relay modules at $1 per relay...2 relays per turnout.

I didnt want to start a new thread since it seems like this one is currently in line with my question.  I believe in Barry's book, it states to run 20-22 gauge stranded wire to the switches (2nd Edition page 58.)  I'm reading here that people are using 18, will 20 work??   

Also, the power source for my reatrax switches will be an old Lionel KW.  How do I know how much power to apply to the transformer (18 volts??)