Lighting control panel

[gavino200]

I have not done this YET!! but I think its a viable option. (This is why I lean toward Digitrax as they have a lot of components to work with).

 

Use a SEC8 and control the LED's via DCC. The SEC8 can control 32 LEDS, Normally they are used for signaling aspects but there is no reason you cannot us them for other things.

 

Using JMRI will allow you to control them via a computer and create a control panel on a screen. This is a lot easier than physically building one. If you dont like it you can redo it easily.

 

Also you can create scripts to operate the LED on a time table and much more.

 

Mounting a flat screen TV on the wall allows you to blend the control panel into the room and plus use it as a TV. ............win win

 

The table on the upper right shows the DCC address and which LED will illuminate based on the what condition is sent (closed or thrown). You can build a control panel in JMRI that has buttons that will send out the assigned address for a given LED. Notice that DRV6 position 4 is a flashing LED. With that address you can have lights flash based on your needs.

 

Inobu

 

 

 

 

Inobu! I was wondering when you'd chime in. I was thinking of messaging you to get your two cents.

 

I'm planning on using this set up when I get around to signalling. It'll be necessary because I want to incorporate block detection so that the signals are somewhat automatic.

 

I'd have to think through this carefully before considering it for the LEDs. Here are a couple of random thoughts for now. I'll add better questions later after I've done a bit of cogitation.

 

1. This may not be as much fun for an 8 year old, as a physical switchboard.

 

2. Related to 1. The physical board is a chance to teach some basic skills and knowledge to the boy. 

 

3. Concerned about boot up time etc. Is there a way to leave this running in a dedicated computer 24/7 so that it's already set up when I turn on the layout.

 

4. How many circuits can I run on a single SEC8? We have some fairly grand aspirations as to how many LEDs we will have on this essentially all urban layout. It's going to be a lot.

 

5. Is it possible to make a computer graphical interface that would look a lot like a physical switchboard. This sort of goes to point 1.

 

I'm definitely interested.

 

Edit: When using this kind of setup for signalling. Is it possible to use something like arduino to make the control units for the signals self contained and underboard. ie. Independent of a computer that has to boot up and make a connection.

Edited April 8, 2017 by gavino200

[kvp]

The SEC8 is a signal head driver and supports up to 8 signals (32 heads) but not with all leds on. A lighting acccessory decoder with individual on-off control is better. However it would get expensive very fast as a single 8 signal system would cost as much as the whole analog setup with a hundred leds. An arduino is not much better cost wise but at least it could be left running.

 

Imho the through hole screw mounted single pole double throw (on-off) switches are the easiest as the switch outputs could be soldered directly to the wires which is way faster and easier than routing them on a circuit board. Much cheaper too. Also no chips, software and programming so not much could break in it.

[inobu]

Its hard to call which will be beneficial. The physical build verses a graphical build. For me I've started to convert to computer controls and replaced pc's with Raspberry pi's running JMRI.

The JMRI boots up with the layout so the wifi and panels initializes the layout itself. The Raspberry pi has a hdmi output that I connect to a flat screen.

 

The programming and development time involved with the ardinuo isn't worth the time as everything is already done in JMRI. Learning Python or Jython is quicker then trying to integrate the arduino.

 

The SEC8 can control 32 LED heads or a total of 64 LED's.

 

Here that panel I build a few month ago. The guy listened to me, put up a flat screen with a Raspberry pi and he is so happy now (His words). When JMRI boots it clears all the switches automatically. You can remove all the switch maps

and leave just the toggle switches and LED;s or make your own icons. If you don't like it you can delete it and do it again. 

 

[gavino200]

Its hard to call which will be beneficial. The physical build verses a graphical build. For me I've started to convert to computer controls and replaced pc's with Raspberry pi's running JMRI.

The JMRI boots up with the layout so the wifi and panels initializes the layout itself. The Raspberry pi has a hdmi output that I connect to a flat screen.

 

The programming and development time involved with the ardinuo isn't worth the time as everything is already done in JMRI. Learning Python or Jython is quicker then trying to integrate the arduino.

 

The SEC8 can control 32 LED heads or a total of 64 LED's.

 

Here that panel I build a few month ago. The guy listened to me, put up a flat screen with a Raspberry pi and he is so happy now (His words). When JMRI boots it clears all the switches automatically. You can remove all the switch maps

and leave just the toggle switches and LED;s or make your own icons. If you don't like it you can delete it and do it again. 

 

 

Inobu, this is brilliant. But alas, too brilliant for me now. I'm going analog for the LEDs. I'm looking forward to experimenting with this stuff when I get to signalling.

 

A couple of points on my limitations:

 

1. I'm already embarked on a number of learning curves with this project, with just the basics.

 

2. I like doing Arduino projects. It gives me a sense of how computer code works, without me having to learn a real computer language. When we get to more advanced projects, I'm content to be what coders call a "script kiddie" and just use other peoples code. If the boy gets into it and wants to take it further, I'd support that and maybe take it a little further myself. But I'm not planning to learn python or anything like that. I just have to many other hobbies that would take preference, to be able to learn it well. 

 

3. I didn't know about Rasberry Pi. It looks interesting. I'll look into it.

 

4. I'm thinking a little bit strategically about the order in which I go about learning things. The reason why I chose to light the station first, is that it's essentially completely pre-built. There's no modelling involved. It's exactly how it came out of the box with the exception of adding tiny stickers. All I have to do is add LEDs. Therefore it's the perfect vehicle for learning about LEDs and lighting. The next most important thing to do with the layout is to do some modelling - city blocks, buildings, and hopefully a tunnel with subway station. For those projects the main learning curve will be techniques of modelling, so I want to have LED knowledge under my belt. Only after we have a decent bit of modeling would it make sense to start working on signalling, traffic lights, and automation. 

 

5. I am, however, trying to include the infrastructure for future advancements. The track is set up for block detection. The junctions are wired but not connected. I'm going to add speakers as I go, but not connect them yet. I'll also give some thought to the possibility of switching to computer controlled lighting in the distant future and take that into account when laying the wiring.

 

While I recognize your level of knowledge is miles above mine, I find your ideas fascinating.

Edited April 8, 2017 by gavino200

[gavino200]

Imho the through hole screw mounted single pole double throw (on-off) switches are the easiest as the switch outputs could be soldered directly to the wires which is way faster and easier than routing them on a circuit board. Much cheaper too. Also no chips, software and programming so not much could break in it.

 

kvp, I think you may be overestimating my abilities. I'm just planning on using those brown plastic bread boards with the holes in them. It's just a strategy for keeping the switches, pots, and lights, in order. I'm still just soldering wires to switches. 

[kvp]

I'm talking about using those metal lever hole mountable switches. Drill a hole in the panel, put in the switch, put on the nut, tighten it, done. There are solder tabs on the bottom of the switches. For an on-off switch you would have two. There are screw mountable potmeters too with the same tabs.

 

It's possible to add all components to the panel board/box, then put the end of each wire through the hole in a tab and use a bit of solder to fix it. No breadboard required and the front is a solid piece of wood or plastic, holding the whole circuit together. Much easier than using the raster boards.

 

For leds, you can get through hole led holders, like the one Jeff linked above. If you solder up led+resistor pairs beforehand, you could just connect the two wires with a bit of solder.

[gavino200]

I'm talking about using those metal lever hole mountable switches. Drill a hole in the panel, put in the switch, put on the nut, tighten it, done. There are solder tabs on the bottom of the switches. For an on-off switch you would have two. There are screw mountable potmeters too with the same tabs.

 

It's possible to add all components to the panel board/box, then put the end of each wire through the hole in a tab and use a bit of solder to fix it. No breadboard required and the front is a solid piece of wood or plastic, holding the whole circuit together. Much easier than using the raster boards.

 

For leds, you can get through hole led holders, like the one Jeff linked above. If you solder up led+resistor pairs beforehand, you could just connect the two wires with a bit of solder.

 

Yes. I see it now. This is the way to go.

 

So the box is basically two boards - a front and a back. I had been thinking of building the circuits on the back board and then adding a front cover.

 

What you are suggesting is better. Fit the LEDs, pots, and switches to the front cover. Then build and wire everything onto the underside of the front cover. No roster boards necessary. Then I can add a removable back cover to close it. 

 

Have I got it right?

[gavino200]

This is the best schematic of a hole mount mini toggle switch that I could find. This one is pricey.

 

https://www.mcmaster.com/#standard-toggle-pilot-switches/=174299l

 

It looks like I'll need a panel board about .2 inches thick. I'll need a fairly hard wood.

 

 

These LED mounts suggested by Jeff should look good with them. 

 

http://www.ebay.com/itm/10Pcs-3MM-Round-Chrome-Metal-Plastic-LED-Lamp-Light-Emitting-Diode-Bezel-Holder-/361943265908?hash=item54457ff274:g:lS0AAOSwXYtY4O4i

 

Something like this for a hole mountable pot. With or without a knob. 

 

2K

 

http://www.allelectronics.com/item/pp-2ku/precision-3-turn-2k-ohm-pot-used/1.html

 

Or this 1K

 

http://www.ebay.com/itm/2-x-1K-OHM-Linear-Taper-Potentiometer-Round-Shaft-PC-Mount-USA-Seller-Free-Ship-/232288744548?hash=item36157cf464:g:VksAAOSwgkRVT9an

Edited April 8, 2017 by gavino200

[inobu]

Gavino,

 

i know that you stated that you were going another route but I wanted to give a better illustration for others that are following along. The graphical toggle switch work in conjunction with the graphical LED. The panel background was done in Illustrator. I just copied and pasted the panel text twice. This is the controls panel background. 

 

Bringing the background image into JMRI I place the icons on top of the image. The switches will flip back and forth via a mouse click. The switches are linked to a

table that has the addresses assigned. Theses addresses are linked to the SE8C which are wired to the LED's at the platform or what ever.

 

When the toggle switch is clicked on JMRI will send out the DCC address with the thrown commands and the SE8C will light the LED up.

 

This took about 20 minuted to set up. You can beat that. Changing it around is really easy.

 

There is a learning curve with JMRI but once you get it going skys the limit.

 

 

In any case keep it going!

 

Inobu

 

You can see how the LED's are crooked on the right side. You can move the leds by dragging them around so its easy to place things. 

Edited April 9, 2017 by inobu

[gavino200]

Gavino,

 

i know that you stated that you were going another route but I wanted to give a better illustration for others that are following along. The graphical toggle switch work in conjunction with the graphical LED. The panel background was done in Illustrator. I just copied and pasted the panel text twice. This is the controls panel background. 

 

Bringing the background image into JMRI I place the icons on top of the image. The switches will flip back and forth via a mouse click. The switches are linked to a

table that has the addresses assigned. Theses addresses are linked to the SE8C which are wired to the LED's at the platform or what ever.

 

When the toggle switch is clicked on JMRI will send out the DCC address with the thrown commands and the SE8C will light the LED up.

 

This took about 20 minuted to set up. You can beat that. Changing it around is really easy.

 

There is a learning curve with JMRI but once you get it going skys the limit.

 

 

In any case keep it going!

 

Inobu

 

You can see how the LED's are crooked on the right side. You can move the leds by dragging them around so its easy to place things. 

 

That really is impressive, Inobu. Can it also control light intensity?

[inobu]

Most lights are florescent which isn't variable but using a decoder with the LED connected to the motor leads makes it possible via DCC but not very practical. If it were me I would find the correct voltage levels and skip the variable option. The key for a good platform lighting is light diffusion. 

 

The example given was just to show everyone that JMRI is a convenient way to build a panel. It has its learning curve but once you get it you got something. 

 

Inobu

[kvp]

Yes. I see it now. This is the way to go.

 

So the box is basically two boards - a front and a back. I had been thinking of building the circuits on the back board and then adding a front cover.

 

What you are suggesting is better. Fit the LEDs, pots, and switches to the front cover. Then build and wire everything onto the underside of the front cover. No roster boards necessary. Then I can add a removable back cover to close it.

 

Have I got it right?

Yes. I'm sure there are cheaper switches on ebay and you can use ones that are only rated at around 1A as they are both cheaper and smaller.

 

For the box, building the front and the sides as one pieces allows you to add the connectors in a similar drill and screw fashion to the sides. Then you just screw on the back plate. Very easy and the whole circuit is traceable visually by following the wires around.

[gavino200]

Most lights are florescent which isn't variable but using a decoder with the LED connected to the motor leads makes it possible via DCC but not very practical. If it were me I would find the correct voltage levels and skip the variable option. The key for a good platform lighting is light diffusion. 

 

The example given was just to show everyone that JMRI is a convenient way to build a panel. It has its learning curve but once you get it you got something. 

 

Inobu

 

Yes, I guess you can program in the voltage levels and change them as needed until you get them right. Or you could always combine your system with Jeff's method of putting pots locally in the buildings.

 

Regarding light diffusion. Can you point me to at thread or site where that's discussed, or expand on it a little?

 

Also, how much different would the basic wiring plan be (with the exception of the control panel, of course) when using SEC8? If, for example I wanted to try using SEC8s in future, would I need a complete rewire or can the changes be made on the off-layout side?

[gavino200]

Quick question about voltage. Somewhat moot, as I've already ordered a 12V transformer. But I'd like to know for the future.

 

Most people seem to use 12V. But some, including a few who have contributed here, use 5V or 6V. How much is enough? How much is too much? What's the thinking on this?

 

Edit: Nevermind. I have the information I need to answer this question from Jeffs previous post. Just need to mull it over.

Edited April 10, 2017 by gavino200

[gavino200]

Ok, everything's ordered except for the platform lampposts and the 25 pin serial cable. I still have to decide what length cable I want.

 

I hope to make a final circuit design over the next week or so. That is, I'm going to combine kiha and kvp's ideas into my own drawing. 

 

When I receive all the components, and know the exact size, I'll draw a design for the control panel box.

 

In the meantime I'm going to think about modeling and scenery around the station. Otherwise, I'd end up needing to go back and add LEDs to the panel later. It's amazing how one thing leads to another in this hobby. The modeling part is probably the most intimidating for me because of the artistry involved.

Edited April 10, 2017 by gavino200

[cteno4]

12v will let you do circuits with 3 leds in series with a resistor (or pot), but when you just do single leds from 12v you are burning most of your power in the resisitor, wasting 3x the power heating the resistor it takes to light the led. At 5v the resistor only burn 40% of the power it takes to light an led.

 

The most energy efficient is the 12v with units of 3 leds in series. But if you use a pot it dims all three together.

 

Most all the led strips use the 12v three led units. Some like the wire ones are parallel units that will need a high watt resistor.

 

Cheers

 

Jeff

Edited April 10, 2017 by cteno4

[cteno4]

Gavin,

 

I would suggest getting some components and playing with this some before diving in the deep end on it all. First get some leds and small 2k pots and play lighting a few things to see how you like to light things. Then you can figure out the best path for you.

 

For the control panel and the wiring the same thing once you see how you want to do things. If you don't go the computer approach and use a physical control panel I would suggest doing the panel in a modular fashion where you can keep adding units on it as you go. One big massive panel can get very messy to deal with and unless you do the whole thing at once it gets harder to open up and do more installs and potentially breaking something already done. Also easier to trouble shoot.

 

Cheers

 

Jeff

[inobu]

Yes, I guess you can program in the voltage levels and change them as needed until you get them right. Or you could always combine your system with Jeff's method of putting pots locally in the buildings.

 

Regarding light diffusion. Can you point me to at thread or site where that's discussed, or expand on it a little?

 

Also, how much different would the basic wiring plan be (with the exception of the control panel, of course) when using SEC8? If, for example I wanted to try using SEC8s in future, would I need a complete rewire or can the changes be made on the off-layout side?

There's no thread concerning diffusion but addresses the issue people complain about (stations looking too bright). The focal point of the LED's beam is too intense which radiates wide. The LEDs need diffusers

to control the radiated light. I bought a bunch of different LED to play with in order to find the right configuration for my station. Hopefully Ill have time to get it going and post something up.

 

SE8C connection.

 

The SE8C uses a 10 pin ribbon connection that you can fan out. The 8 10 pins connectors on the board.

 

 

http://www.digitrax.com/static/apps/products/detection-signaling/sdck/images/SDCK_jpg_2000x2000_q85.jpg

 

 

 

http://www.digitrax.com/static/apps/products/detection-signaling/se8c/images/SE8C_jpg_2000x2000_q85.jpg

 

 

 

If you terminate the LED to a central location that allows for interconnecting it will no be a problem. \\

 

Inobu

[kvp]

 

The LEDs need diffusers to control the radiated light. 

There are leds available with built in diffusers. Or you can just get wide angle leds.

 

For the SEC8, these boards don't have proper light intensity control and any brightness controllable decoders available today are usually single output only, meaning you'll need one decoder per individually switched led group, meaning a simple analog switch is traded for a really expensive DCC accessory (motor) decoder.

 

So unless someone comes up with a cheap DCC accessory decoder with continously variable light output that could be changed without register programming, the pure DCC solution is completly unusable for the task this topic is about. I know that there is an arduino based DCC decoder that could do this, but it's limited in the number of outputs and still fairly expensive. Not to mention, the DCC protocol is not really designed for this purpose. A loconet based system could be better, but proper software support for that is very limited and it won't be cheap either. (not to mention that moving physical switches is usually more fun than tapping a display)

[cteno4]

Led diffusion is one of those things mentioned in a number of threads but don't think it's has its own thread as inobu said.

 

Most of the diffusion talk has been in car lighting where the nature of the size and placement of the leds requires some sort of diffuser to keep the led mini suns from making too much of a point source and or moving the light from and led positioned at one end of the car to spread as evenly as possible down the car.

 

Leds are tough to diffuse as they are such a pin point of light. This is even true in real life. It's been a challenge to use them in exhibits due to this and take a while for manufacturers to come up with diffusers that don't cut the light output too low for the form factor of the bulb.

 

Another issue with leds and diffusers on model trains can be heat. You need a material that can take being a little warm over time and also not trap too much heat. This is another reason I like to run leds at lower currents to lower the heat generated at them (along with using more light sources to be more realistic).

 

One big way to diffuse leds and also affect their color some is to bounce the light off of something. This turned out to be the way i have to do the leds under our kitchen cabinets as the downfacing leds made bright shiny spots on the granite counter tops and due to the mix of colors gave an odd look to the countertop with the different while leds, I played with rgb leds to try to adjust this but was way too much work. Also diffusers on the rob strips heavy enough to knock out the hot spot reflections took out too much of the light. In the end I am going to mount the led strips on a 3/4" chunk of aluminum angle stock so the leds face the back white wall and bounce out. A strip of medium and a strip of warm give the right overall temp mix when bouncing off the flat brightnwhite back wall to light the counter well from the back where it's shaded by the cabinets from the overhead lights.

 

Bounce diffusion is used a lot in building lighting. It also wastes light like diffusers do well so the bounce has to be well directed to make it effective and economical.

 

I played very briefly with the idea with the idea of using some stryene channel stock to mount smd (or led strips now as the narrow led strips like Nariichi had were not around yet!) and point them at the roof of the platform so they bounce down to the platform off the roof. I've seen some platforms do this, but most are just light boxes mounted to the roof or framing and a big diffuser on it so it's not super prototypical. I never finished up the fiddling as I realized the channel stock needed ended up Bing like 2'+ wide and odd to have hanging there.

 

This bounce technique can be used to good adantge in car lighting by having a strip of several leds near the roof of the car facing up to the car roof to be bounced down, really helps diffusing the light down the car and coloring the papar up on the roof could help change the light temp.

 

 

Also in buildings I'm much more a fan of using really thick tee shirt paint to light seal the corners. Then use dimmed leds so the walls don't glow. If it's super thin plastic walls the walls can all be painted or heavy card stock internal wallsmput in with holes cut for the window areas. I like white paint not black as most interiors have a lot of whte walls which reflect light. I'm not a fan of aluminum foil formthe interiors as the reflections of the light are sharp and not as diffuses as using white.

 

Also if the led proves to bright in the structure it can be faced up at anpiece of white card stock on the ceiling and the led shielded with a bit of foil or cardstock so all you see is the bounced light off the ceiling. Putting a piece of colored cardstock in as carpet or flooring can also give some interesting color to the bounced light.

 

I got into playing with all this from way back when I made exhibit models and we had a couple of pro lighting folks come in for shoots and they were not on doing any direct lighting except some over all way pulled back floods. Real to the point lighting in the model was with almost all bounce lighting of some sort and making shadows in just the right places. They were masters of diffusion and It was incredible what it looked like thru the camera lens.

 

Cheers

 

Jeff

[inobu]

Thanks Jeff

 

That is what I was referring to in regards to diffusion. You must control the direction and dispersion of the light beam more so than trying to control the intensity. Controlling the intensity gives you the ability to change the light range but not its focal points. LED have viewing angles and getting the correct angles along with a diffuser can achieve more realistic models.

 

The application for variable lighting is limited. Modeling lighting is typically used to mimic public utilities or building illumination. Using variable lighting may not be necessary if the correct diffusion control is established. 

 

In regards of the SE8C it takes a bit of forward thinking to achieve what ever one wants to achieve. Being receptive to the possibilities along with innovation make it possible. Using toggle switch is pretty much left to the builders digression but here is what my posts are meant to do

 

Invoke Thought

Drive Innovation

Inspire to Achieve 

 

Inobu

[gavino200]

Gavin,

 

I would suggest getting some components and playing with this some before diving in the deep end on it all. First get some leds and small 2k pots and play lighting a few things to see how you like to light things. Then you can figure out the best path for you.

 

For the control panel and the wiring the same thing once you see how you want to do things. If you don't go the computer approach and use a physical control panel I would suggest doing the panel in a modular fashion where you can keep adding units on it as you go. One big massive panel can get very messy to deal with and unless you do the whole thing at once it gets harder to open up and do more installs and potentially breaking something already done. Also easier to trouble shoot.

 

Cheers

 

Jeff

 

Thanks Jeff. I agree with testing. I'm not going to rush this as pretty much every part of the project involves a learning curve for me. 

 

I've been thinking about modularity and not continually tinkering with "finished" work. I agree, it's better to not mess with it too much after it's built. With regard to modularity - if I make the modules too small, I'll have a lot of cables, which I don't want. I've already ordered 25 pin break out boards so I'm inclined to use them. That gives me one common wire and 24 individual circuits (or a permutation of that). I'm thinking of building a box with a few master control switches (Zones) and the remainder of the pins (21-24) wired to individual circuits.

 

The box will be built now. Circuits can be added to the breakout board on the layout as I add LEDs to the layout. Until then the unused switches won't be live and wont be labeled. (That description may not be clear. I can draw it sometime this week)

 

That brings me back to needing/wanting to decide/calculate how many circuits/LEDs I want to power with one transformer, so I can decide how many Zones to power with one transformer. Or I could have one zone per transformer.

 

My plan will be roughly this (below). I'm not going to rush it. It may take a long time.

 

1. Open the Station house. Tee shirt paint the interior. Work on the interior if necessary. Add wire lights and magnets. Test LEDs using a battery connected to a resister. Then leave it.

 

2. Do same with any other structures that I add (Yard sheds, switch control, ticket booths, etc.

 

3. Paint interior of stairs and passageways. Wire. Test. Leave.

 

4. Install LED strips in platforms. I'm planning to add a connector between platforms so that I can take them apart without removing the wiring. Then I'm going to run the wires up the stairwells and down to magnets below the station house. 

 

5 Same with open platform lampposts. Haven't decided whether to wired them with the platform canopy strips or not. Together would be more prototypical. Not sure if that's too much LED for one circuit. Probably is. Each of the two platforms will be on it's own circuit.

 

6. I may have some road/sidewalk by then. If so I'll wire the roadside lamps in the same way.

 

At this stage I'll have disconnected but tested LED wires below the layout and nothing else.

 

7. Build a  LED power bus under the layout. Input from one DC transformer fanning out to the LED circuits. Output from the LEDs converging on the 25 pin serial connector (break out board).

 

8. T break out board then connects to a 25 pin serial cable. Ground pin connects back to the Power supply to complete the circuit.

 

9. The control panel box will be a 25 pin break out board with wires running to (25-x) sets of {Switch, pot, LED}, where x is the number of master switches chosen. The box will be built, tested, closed, and hopefully never opened again.

 

Before I build any component I'm planning to build a small mock up for testing. I bought extras of most of the cheaper components to facilitate this.

 

Anything obvious, or not so obvious, that I'm not considering?

Edited April 10, 2017 by gavino200

[cteno4]

Gavin,

 

Good to start with small bits and grow. You can always add some wiring later.

 

One thing on the serial cable and using a common ground. The common ground will Carry the sum amperage of all the individual circuits. So if you run 24 circuits and use one wire for the common ground that wire will carry the cummof the other 24 circuits, so if 1 led each the that would be 24x0.02 or about half an amp. 30g wire you only want like .15amp for transmission, for 0.5amp you want 24g wire. So I would use at least 4 of the wires in the serial cable for common ground or 5 to be safe and the other 20 for individual led circuits. That's assuming 30g wires in the serial cables, some use 32g.

 

Not sure why you are doing a led power bus and the 25 pin cables. An led power bus would be if you were to power whole sections together and trim them locally if needed and turn whole buss(es) on and off. The 25 pin would be if you wanted to control each circuit (or a few combined) back at the control panel to turn on/off.

 

I don't see the need to control individual or small sets of leds on/off unless you want the mega control panel. Having a block of buildings or all streetlights or some other set on individual power busses and then you can turn each buss on and off. Each of the busses could get powered by a pair of like 20g wires back to the control panel where you could switch them on/off and have indicator led and wholesale pot if you want and that attached to individual power supplies or one big one.

 

But you are smart in doing things in steps. You can temp wire up bits as you go to test and play with ideas of what you want to control. That 30g wrapping wire is less than a cent a foot you can test wiring things up in various ways. 30g wire can power 7 led circuits (with 1 led or 2 or 3 leds in series).

 

Many ways to skinnthe cat.

 

Cheers,

 

Jeff

[gavino200]

Gavin,

 

Good to start with small bits and grow. You can always add some wiring later.

 

One thing on the serial cable and using a common ground. The common ground will Carry the sum amperage of all the individual circuits. So if you run 24 circuits and use one wire for the common ground that wire will carry the cummof the other 24 circuits, so if 1 led each the that would be 24x0.02 or about half an amp. 30g wire you only want like .15amp for transmission, for 0.5amp you want 24g wire. So I would use at least 4 of the wires in the serial cable for common ground or 5 to be safe and the other 20 for individual led circuits. That's assuming 30g wires in the serial cables, some use 32g.

 

Not sure why you are doing a led power bus and the 25 pin cables. An led power bus would be if you were to power whole sections together and trim them locally if needed and turn whole buss(es) on and off. The 25 pin would be if you wanted to control each circuit (or a few combined) back at the control panel to turn on/off.

 

I don't see the need to control individual or small sets of leds on/off unless you want the mega control panel. Having a block of buildings or all streetlights or some other set on individual power busses and then you can turn each buss on and off. Each of the busses could get powered by a pair of like 20g wires back to the control panel where you could switch them on/off and have indicator led and wholesale pot if you want and that attached to individual power supplies or one big one.

 

But you are smart in doing things in steps. You can temp wire up bits as you go to test and play with ideas of what you want to control. That 30g wrapping wire is less than a cent a foot you can test wiring things up in various ways. 30g wire can power 7 led circuits (with 1 led or 2 or 3 leds in series).

 

Many ways to skinnthe cat.

 

Cheers,

 

Jeff

 

Thanks. Good point about wider gauge/multiple pins for the common wire. It should be approx the same gauge or gauge equivalent (pins) as the transformer cable wires.

 

Regarding the "bus" I'm probably not using the word correctly. I'm really just talking about whatever wire goes from the transformer cable socket to the wire divergence point.

 

I'll try to draw up a plan soon. That should make it clear whether I'm understanding this stuff or not. 

Edited April 12, 2017 by gavino200