How to wire a Lionel Gateman for DC ?

Wow, is that a gateman or an explosive device?

Seriously, good job. Btw, my favorite part of that clip is the PW Seaboard Switcher.

Thanks, John. Alot of hours to get this down, but thanks to Stan and GRJ and others, the wiring is figured out and anybody trying this will have a much easier time (and much less time consuming). 

And yes, that Seaboard is a wonderful engine. I threw that in there for the PW crowd.  I should have had more lighting on that 69 signal. It's in excellent condition. It's the first prewar item I've put on my layout. Most of them are beat up and rusty. I walked into this one at a swap meet last week and grabbed it. Little did I know, the silly thing would set off this time consuming project. 

ROGER1 posted:

...What would you do?

Tell me if I'm hallucinating, but it seems the 2 red wires from the 2 buck converters are tied together at the common GM terminal 2.  I believe these red wires are coming from the DC+ OUT of the buck converters.  In my recommended wiring, it's the DC- OUT (black wires) from the buck converters which should be tied together at terminal 2.  Then the red wire of the coil buck goes to terminal 1.  And the red wire of the lamp buck goes to terminal 3.

Since terminals 1,2,3 appear to be spring-clip terminals, it should be easy to swap without soldering. 

That's what I'd do....since you asked that is.  If you do this I don't think you will see any voltage change in the lamp when the door opens.

I can't see the lamp dimming when the door opens so if the voltage to the lamps is indeed dropping, it's not noticeable.

Stan,

First I'd like to thank you for all the help and time you put into it. I really appreciate it.  Projects like this for us electronic neophytes would go nowhere without the help.

As for those wires.......I thought I'd tried all the possibilities, but I will try it and report back.  When the bulb was dimming, there was almost 6 volts of voltage drop (down to 8.5) and the dimming was VERY noticeable. Now it's only a 2 volt drop.  When I got it down to 2, I switched the voltage back to AC to see how much of a drop there is with original PW setup.......1 volt. That's why I left it.  But I'll try the wire swap and see if there's a difference.

Somewhere in all this, if you have the time, I'd still like to know if you can dial down the coil voltage to 11V, 10V, whatever and reliably open the door.

Stan,

Well.......apparently I had tried it. I just went down and switched the reds to the black and vice versa. The door didn't open when triggered. The bell worked, but that's because it's line splits off early. Across terminal 2 and 3 the voltage remained at 12V (before and after triggering). Across the coils the measurement stayed at 0. I arrived at the present wiring arrangement by pure trial and error. It worked and I went with it. And it was the only one that worked. 

I didn't dial it in because I wasn't sure if I'd have to adjust one buck or both bucks to get it right. So, what I did was leave the base off. What you see in the vid is the green sheet metal top.....the bottom is not there. My plan was to run it and flip it up and actually feel the coil to see if it's getting hot. So far, it's not, but I haven't run it for an extended time.       We  have company coming over tomorrow who have never seen my trains and we planned to go down there and "play" with them. So, I was under a time constraint to get this done. I didn't want to bring them down there with the mess I had while working on it. Tools, meters, wires......all over the place.....typical "working " environment. But I got it in just under the wire (so to speak). So, I'll at the very least,  get a better "heat" test on it tomorrow night. 

Hi All

I have an update and a new issue. I started the thread as a result of the purchase of the prewar 69 signal. The noise of the Gateman's coil overpowered the bell, so.......I had to quiet the coil and succeeded. Dead quiet. And the systems worked perfectly. In the interim after my last post, I decided that I wanted to "dress up" the signal with a typical red flasher crossing signal system. I had Dave at Evan Designs make me up a custom alternator circuit that would "wig wag" the two bulbs. While waiting for that to arrive, I fabricated tubes for the bulbs made to look like typical crossing signals. I took all kinds of photos and was going to post them and a video today to show it all in operation. The circuit arrived yesterday. I tested it and it worked great. I then painted the bulbs red to match the prototype look and installed them in my " tubes". Connected the wiring to the bridge rec. that controls the bell ringing and then triggered the 153. The GM came out.....dead silent. The bell rang.  But the LEDs were flickering instead of alternately flashing. I emailed Dave earlier this morning to see what he thinks can fix this. In the meantime, I tried several things. I put a capacitor after the bridge into the bell and flasher shunts. No change. I wired the bell to the AC from the 153 IN FRONT OF the bridge. No change. I connected the bell to track power. No change. Every time the bell rings, no matter what the source of power for it.....the LEDs flicker. I'm thinking that its the wires from the LEDs that go down behind the coil compartment on the signal. Magnetic field interference?

Any ideas on a fix for this?   It all looks incredibly cool, but this flicker thing was an unwelcome surprise

Roger

Well, it's definitely the bell coil.  I moved the wires that go from the back of the LED bulbs in their tubes as far away from the coil box as slack allowed (I had the wires zip tied to the back of the signal). The flickering diminished by over 50 percent. But it's not fixed. But therein lies the problem.

Roger

Can you twist the wires to the LEDs?

What may be happening is the large pulsing magnetic field from the bell coil is inducing small AC currents in the loop (yes, it's not a circular loop per se) of wire between the LED and the driving circuit.  By twisting the wires to the LED, you make the area of the loop smaller so there's less coupling).

But before twisting the wires, a simple experiment for amusement purposes only.  Disconnect the power input to the flasher circuit but leave everything else as is with the original LED wiring and placement.  Turn on the bell circuit.  I wonder if the LEDs will flicker very dimly (may need to turn off the room lights).  

Stan,

Thanks for popping up.  I did not expect this issue.  The flashers looked and worked great. Not sure what you mean by twisting the wires.  The wires fro the bulbs to the alternator Dave uses magnet wire.  So they will “twist”.  Loops or just a twist on their long axis?

i will try your experiment later and report back   

Thanks!

A couple twists along the length of the wire should do it.  It's a shot-in-the-dark but since moving LED wiring away from pulsing coil reduces flicker 50%, it's something simple to try.  

Stan,

i will try that.  And I was thinking.....going to try and shield the wires from the field with some brass sheet and possibly foil wrap

Well.......no luck. I didn't have much slack to do many twists, but it didn't have an effect. I put a piece of brass back there and it's effect was sporadic. Foil did nothing. Moving those wires causes it to improve but it's erratic. For a couple of seconds it's close to normal then goes faster. Up and down. Looking like this alternator doesn't want to play nice with this signal (the bell, at least). 

In doing all this testing, the door is open much longer than usual if a train goes by. As I mentioned, the base is not on the GM, so I can reach under and feel the coil. It's definitely getting pretty warm.....close to hot.  So, I guess it's time to mess with the voltage on the bucks. Should I just lower the buck that controls the coil or try to match both? And should I adjust them with the wires loose and no load?

Roger

ROGER1 posted:

1 ... The circuit arrived yesterday. I tested it and it worked great.

2...I tried several things. I put a capacitor after the bridge into the bell and flasher shunts. No change. 

1. Did you test it with a "pure" DC supply...or with the output of an AC-powered bridge rectifier with NO capacitor?

2. The bell needs pulsing-DC to ring.  If the bell still worked after adding the capacitor, then the capacitor was not large enough to smooth out the pulsing-DC.  I believe you were thinking a smoother DC supply might eliminate the LED flickering.

So that was my first thought too.  But crossed that off after you showed the flickering changed simply by moving the LEDs away from the coil.

In any case, if you have a photo that shows the circuit from Evans (I don't suppose they provided a schematic) I can probably guess on its susceptibility to pulsing DC.  So what I'd do next (depending on your answer to #1) is smooth out the DC supply to the LED flashers.  This may mean "isolating" the pulsing-DC going to the bell coil with a diode so that any capacitor you add to the LED flasher supply does not affect the voltage going to the bell coil.  I'm not aware of any DC-powered circuit that works "better" with a noisy/pulsing DC supply compared to a smooth/pure DC supply! 

johnstrains posted:

Wow, is that a gateman or an explosive device?

Seriously, good job. Btw, my favorite part of that clip is the PW Seaboard Switcher.

 

Me too! I love my Gatemen (3 of 'em so far), but that Seaboard is definitely my fav also!!

George

ROGER1 posted:

...So, I guess it's time to mess with the voltage on the bucks. Should I just lower the buck that controls the coil or try to match both? And should I adjust them with the wires loose and no load?

So here's the thing.  I've given the internal 3-terminal GM wiring some thought after your description of how they coil and light circuits interact (e.g., light dims when door opens).  So I hooked up a similar configuration of components/modules myself and discovered that there is an interaction - but it's beyond me to explain in a manner which gets us to a solution vs. an arcane discussion about circuit design.  And, yes, the interaction is a function of what the two regulators are set at.  And, yes, the interaction is different when you share the "+" of the coil and lamp vs. share the "-".

If I were doing it (and I'm mindful that no job is so easy as the one you imagine someone else doing ) I'd throw in the towel on the shared common terminal and install a floating lamp socket.  So the 2 wires from the lamp DC-DC converter output go to just the lamp.  Actually, since it appears that 12V or so is a suitable voltage, I'd take the DC lamp voltage and drive a 12V LED strip and be done with it.

As for adjusting the coil voltage, in all likelihood the voltage you set with no load might drop a bit (say, 1/4V) with the coil attached.  The math is somewhat complicated and depends on information not readily available (like the schematic to the regulator module).  But it would be useful to know if you can reliably fire the door mechanism with 11V or 10V or whatever.  As for the heating, it is what it is.  That is, I think the coil is about 8 Ohms.  12V applied is a power of 18 Watts which goes out as heat.  18 Watts is a lot of power to dissipate!  Consider that some small soldering irons might only use 18 Watts and they can melt solder.

But if you want to do something about the heating, I can imagine installing a 25 cent PTC resettable thermal fuse that would cut power to the coil after, say, 10 seconds (or whatever).  Or there are so-called Delay-off timer relay modules (about $1 on eBay) that will open a relay after some settable time-interval (like 10 seconds or whatever).  If you are just having the time-of-your-life messing with this mechanism, you might look into a "trick" used when driving perhaps the most common spring-loaded coil mechanism which is a relay itself.  What happens is you apply the "full" voltage to turn the relay on, then you drop the voltage to a much lower one that is enough to hold the relay closed.  This cuts relay power (heating) dramatically - routinely more than 50% reduction.  Point being, I can imagine something like this allowing the coil to remain on indefinitely.  But it would take some experimentation to confirm the GM mechanism would play-well with such a scheme.

Finally, when all's said and done, after all bells, flashers, etc. have been worked out, I'd re-visit the design to simplify where possible.  I pretty sure this has long since gone beyond what anyone else would attempt to duplicate.   But given the amount of time you've invested, perhaps there's a list of "lessons learned" that could help anyone else considering this modification.

1. Did you test it with a "pure" DC supply...or with the output of an AC-powered bridge rectifier with NO capacitor?

I tested it with AC first, then hooked it in to my Gateman circuitry. Worked perfectly. As soon as the bell goes off, it's goes nuts.  Regardless of the power source, it seems.

2. The bell needs pulsing-DC to ring.  If the bell still worked after adding the capacitor, then the capacitor was not large enough to smooth out the pulsing-DC.  I believe you were thinking a smoother DC supply might eliminate the LED flickering.

A couple of things here.  The bell works with AC and pulsing DC.  Also.....I tried adding a cap after the bridge and there was no improvement. I had just taken shipment that day of 5 new 470uF caps (to replace what I used). But I thought.....why waste time trying various sizes.....let's go with a big one. Put a 4700 in there. No difference. I think I clipped it in correctly. + out from the bridge to + on the cap.  If you think this could be corrected by a cap bigger than 4700, I'll order one.     I was definitely starting to think that I'll have to chose between running the setup (on trigger) with the bell ringing OR the flashers. If I had to choose, I'd pick the flashers. The setup looks very cool

I don't have a schematic of the alternator (and it's sealed up in shrink). I emailed Dave yesterday, but it's not a working day for them. I'm sure I'll get a response tomorrow as to his thoughts on this.            I started looking at EMF shield tape on the Bay last night. Too expensive for this silly thing. LOL

Also, if you can, let me know the best way to "adjust" the voltage out of the bucks to prevent heating of the coil. Both the same, or just the buck controlling the coil. I'm thinking they have to be the same, but need to check.   I do have a Fluke Infra Red temp meter I use typically to compare the motors on dual motor PW setups after a tuneup. I haven't yet, but want to get a reading off the coil. 

Thanks!

Stan,

I guess I was typing when you entered the second post.

If I were doing it (and I'm mindful that no job is so easy as the one you imagine someone else doing ) I'd throw in the towel on the shared common terminal and install a floating lamp socket.  So the 2 wires from the lamp DC-DC converter output go to just the lamp.  Actually, since it appears that 12V or so is a suitable voltage, I'd take the DC lamp voltage and drive a 12V LED strip and be done with it.

Thing is.....the Gateman ONLY works with the two + leads on the left, black on the right. Reversing it and nothing happens.  I arrived at that wiring setup not through basic knowledge. It was trial and error. It was the only arrangement in which the GM worked. FWIW.   The lamp also dims a bit when the GM is run with normal wiring and AC. A bit over 1V drop. With my setup, I get a bit of dim, and it's a bit less than 2V drop (but DC). Does that matter?       And yes, I know I'm over engineering this.  And I do have PTCs laying around. I used to use them with my ZW (until I discovered 91 breakers).  I tested those bad boys in series with my friend's PM bricks and the 91s trip first every time.

Where's a good spot to put in a PTC?

Since you have a dialog with Evans I'd say that's your best bet; I realize now that their flasher circuit operates with AC or DC power but I'd only be speculating and having you perform shot-in-the-dark experiments whereas they should know right away how to troubleshoot.  I don't think it's worth spending more time on capacitor values to additionally smooth the bell/flasher pulsed DC voltage until you hear from them.

What's important is to determine the operating voltage range of the coil to reliably open the GM door and slide out the man so you're measuring the DC voltage under load.  If a PTC is installed, put it right on the wire that goes to the coil terminal on the GM frame.

I understand your combination of components only works when you tie the "+" outputs to the common.  I've already opined about how to approach this so I'll leave it at that.  If trial-and-error shows that it works to your satisfaction with different voltages from the two converters then so be it.  

Stan,

I totally understand what you say about that terminal. I know that it's best to disconnect the bulb and set up another circuit. I'll see how it goes after I deal with this flicker issue.  And I do appreciate your keeping after me on that issue. LOL

In the meantime, I've got some rough photos and videos of the setup in operation. I don't have the bell connected. The wiring is not hidden yet, and the roof popped up, but.....you'll get the idea.

Dave at Evans is sending me a new circuit. He said the original might be scrambled as a result of setting the flash frequency. Their customer service is second to none.      I put a 10,000uF cap in there last night and it had no effect. I also tried putting a diode before the bell circuitry as Stan suggested. Also no effect. 

Hoping the new alternator does the trick, otherwise this bell signal will become just a flasher signal. 

Roger