So I finally got a chance to play around with the stuff I bought. Works pretty well, at least using Freestyler
Had a few false starts because the Amazon listing for the USB adapter I got said to use a certain device type in Freestyler, but I couldn't get it to work. It saw the device, but didn;t seem to be sending commands to my driver. Then looking at the reviews on Amazon, someone mentioned selecting a different device type. I switched to that and boom, my RGB LED strip was under control. MUCH nicer transistions with 255 steps for each color, the simple little controller that came with the LEDs only did 15 steps per color.
I can see it will take a long time to figure out just exactly what color levels for each color will make the most pleasing lighting, and also to figure out how to smoothly transition between those nice colors and all white for daylight or all blue foor night, without really out of place colors appearing in the middle of the transition.
I stuffed everything in a box, I was considering tossing it in my suitcase to take along and work on during my down time. It's checked baggage, but I wonder if it will cause any issues going through security.
What I really want to do is write my own code to control it.
One possible downside is that in this driver, they use a PWM frequency that is definitely in my audible range. Depending on how bright a certain color is set (mid levels are the worst), I cna clearly hear the noice coming from the circuit.I don;t think I'd hear it if the unit were tucked up under benchwork, but sitting on my desk no more than 2 feet from my ear, I can hear it. Dogs didn't react though, sitting some 10 feet away.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Apparantly newer versions of JMRI have an optioon to use a DMX interface. That may be another option. I assume that would allow things like synchonizing the lighting with the fast clock.
Let me know how it worked! I'm going to try to follow your lead if your system works like you want. Thanks!
York1 John
Well the heck with it, just ordered a proper USB to DMX adapter from Amazon, Prime gets it here Sunday. So I will have time to play.
Well, my decoder for DMX arrived today. Looks OK (of course I opened it to peek inside). Not the greatest, but I've seen far worse.
Now I have to find my RGB reel and attach some wires to connect it to the decoder. And write some code to send DMX protocol packets through the USB to RS-422/485 adapter I have (it worked fine to allow JMRI to control an Arduino running CMRI software). I may need to get a proper USB to DMX adapter though, looks like you can;t just make this up ion the fly, at least fast enough to be sure, not in any higher level language on a multitasking OS anyway. I have a feeling I can do a quick and dirty and it will work with 1 or 2 channels, but as sooon as I try expanding it out to as many channels as I will need for the layout, it won't work without a proper driver. And a regular 8 bit Arduino appears to not be fast enough to meet the bit timing requirements, so I either need a proper adapter or one of the faster processor Arduino compatibles.
5 amps sounds rights for the RGB strips, less than 2 amps per color.
But that's at 12 volts. So 60 amps at 12 volts is 720 watts. Less than 8 100 watt incandescent bulbs, which could never light the whole space, never mind the heat.
rrinker Since each strip of 5 meters worth of LEDs gets its own connection, there's no issue with daisy chaining too many strips and having the far one dim due to voltage drop.
The specs on ebay state 5amp/5M to power these lights. 12 x 5 would be 60 amps to power your lights. Is that right?
Henry
COB Potomac & Northern
Shenandoah Valley
Blue fills are often used in theater to project the night - it may not be the ONLY light. The whites may be on dim, with blue on full (much lower light output per goot than the white LEDs).
I've actually always liked the bluish night shots of models. It's as much about setting a mood as it is the actualy overall color. Not so much John Allen's night shots with all the black light illuminated objects, all that was pretty much the only option available at the time. It probably all goes back to 8th and 9th grade, we had a planetarium in our school and one of the teachers that ran it was my science teacher, and I ended up learning how to work a lot of the equipment - outside of the star projector itself, most of it was home made as there was no budget for it. And in 9th grade, our English class presented a play. No way was I going to be on stage, so I did the lights. I quickly figured out the controls, and since the play ended with a scene that the stage directions said "the dawn of a new day" well, instead of just flipping on the lights like they expected me to do, I did what I think was a pretty good sunrise given the limited set of lights I had to work with.
I think when the light level gets below a certain level, the color receptors don;t fire, however, the brain fills in so much of our vision it isn't funny. My vision loss in the right eye is so bad that I pretty much have no way to read text with just my right eye. In parts of the field I can still see shapes and motion. In unfamiliar situations, it really messes with my depth perception, but in familiar places, reaching for familiar things - it's nearly unchanged. How can that be unless my brain is filling in the details I can no longer see. Hammering nails is now a bit of an adventure - I recently drove a pair of nails in the wall to hang a picture. While holding the nail - I didn't miss once. Once I was able to let go to finish driving the nail - I kept missing. Mu brain was somehow able to compensate by figuring out how far my left hand was extended holding the nail and combined with the clear vision from my left eye, was able to guide my right hand to strike the nail squarely. Take away that extra input of the left arm and vision alone was not enough to keep consistent.
don't we humans use our eye's rod cells at night and see things in black & white?
greg - Philadelphia & Reading / Reading
I've always wondered why the use of blue lights for night-time. Never in my life can I recall being outside at night and everything being blue ! All the model pictures under blue lights don't look like night to me.
Why not just dim your room lights down to just barely on and let your layout lighting create the proper night effect ? A full moon at night doesn't illuminate evrything in blue either, it's just very dim pure white light - probably cooler on the Kelvin scale than the sun itself.
Mark.
¡ uʍop ǝpısdn sı ǝɹnʇɐuƃıs ʎɯ 'dlǝɥ
carl425 rrinker Most areas I see as having 2x white strips, 1x blue, and 1x RBG. I'm curious how/why you came up with that combination.
rrinker Most areas I see as having 2x white strips, 1x blue, and 1x RBG.
I'm curious how/why you came up with that combination.
Mostly from putting LED strips under the bottom shelf to light my work bench. One strip of white is not bright enough. Even the higher brightness 5035 type I got.
The effects lighting, the RGB for dusk and dawn, cam be a lot more subtle, I'm not going to those once ina lifetime sunrise and sunsets you see, it needs to be more of an 'everyday' look. And one string of blue - technically 2 if I turn the RGB too full blue - is plenty to simulate a night scene. The whites will be dimming or brightening up as the RGB set goes through the colors for the sunset/sunrise. ANd if it's not quite enough, the RGB set can also be a third string of white.
I've seen a much more complex system with some strips behind the layout at the base of the backdrop to shine up on the 'sky' but I don't think I will go that far. It does look nice in the video I saw though. But that was a diorama, not a room-size layout with multiple decks.
All assuming this controller with a true 256 steps between off and full bright for each color allows for a smooth enough transition. If nooot, I'm only out $10, and I can probably find another use for it, for holiday decorations or backlighting my TV.
York1 Please post more when you get things working. I've wanted to know how to program lights for daytime and night to cycle. It sounds like you have much more electronic knowledge than I do (that's not hard -- I know next to nothing about electronics). Maybe someday there will be an easy solution for people like me, or there will be a cheap version to buy. At any rate, I'm interested in how yours works out.
Please post more when you get things working. I've wanted to know how to program lights for daytime and night to cycle.
It sounds like you have much more electronic knowledge than I do (that's not hard -- I know next to nothing about electronics).
Maybe someday there will be an easy solution for people like me, or there will be a cheap version to buy.
At any rate, I'm interested in how yours works out.
That's the thing - I found these ready to use pieces for the same/not much more money than I could design and build my own, so that takes most of the electronics knowledge right out of the equation. You don't really need to understand the theory to use these things. It's maybe a little more complex with LED strips because you have some wires to attach but seeing as how the terminals are labeled R, G, and B just like the terminals on the strips.... If programming is not your thing, there are off the shelf coontrollers, and while they have all these slide controls to adjust the brightness fooor each channel, they also have some program capability, in the sense of "move the sliders how you want, and it remembers what you did and repeats it by pressing one button" not in the sense of writing code. And there's a cabling standard for interconnecting it all, the cables can be purchased instead of made. The light riggers who set up theater lighting using this same technology aren't programmers and engineers. Outside of John Deacon of Queen, who's retired from performing, I can't think of any other electrical engineers in rock music.
I knew OF this before, but always thought it was probbaly far too complex for simple layout lighting, given that it runs all those fancy displays and light shoows and smoke machines and so forth. I would never have considered it had it not be mentioned on the EE forum I follow, I was much surprised to find it's really quite simple, even on the technical side, and also there are now inexpensive components available, not all $1000+ professional theater gear - at those prices, forget it, but the drivers I've found, plus the dropping costs of LED strips, it's almost a no brainer. Probably why even some small time local bands are able to have decent lighting.
rrinkerMost areas I see as having 2x white strips, 1x blue, and 1x RBG.
I have the right to remain silent. By posting here I have given up that right and accept that anything I say can and will be used as evidence to critique me.
I don't mean type - I've already decided (barring the introduction of some new technology between now and when I actually get something built) to use LED strips, a combination of white, RGB, and blue strips (so I can do daylight, dusk/dawn, and night lighting). This is about controlling it all.
My original idea was to build my own controllers, using Arduinos and some PWM controllers (native Arduino PWM is too slow, too flickery) with MOSFET drivers. Linked by some sort of communications protocol. This topic recently came up on an electronics forum I follow, and several suggested just buying off the shelf DMX512 components. I always thought this stuff was expensive (after all it is often used for professional theater lighting). Well, yes, there are some rather expensive controllers, but like everything else, there are relatively cheap Chinese made versions. Which are exactly what I was considering building, a micro, PWM controllers and MOSFET drivers. $50-$60 gets me one with 36 channels - that's 12 RGB strips or 36 strips of a single color. Each channel has the drive capability for slightly more than a single 5 meter reel of LEDs typically draws - perfect. DMX512 is a protocol that rides on top of RS422/485, and supports up to 31 devices (plus the controller). I almost can't build my own for that. And the protocol has been around forever, it's really pretty simple, easy enough to generate with a VB program on the PC, or whatever. There are even free programs that have the sliders like a physical controller console, although I'd want to program the effects, not manually adjust sliders. I ordered a simple 3 channel one, less than $10 and it ships from the US, not China, so I will have it next week.
I have a reel of RGB LEDs I got a long time ago to experiment with, it came with a simple controller and a remote, and power supply, but there are only 16 brightness steps for each color so it's very glaringly obvious with each step change - a standard DMX controller is 256 levels per channel, there are some more expensive ones that link 2 channels to a single output and offter a 16 bit out, 65535 levels per channel - but I don't think that will be necessary.
Since each strip of 5 meters worth of LEDs gets its own connection, there's no issue with daisy chaining too many strips and having the far one dim due to voltage drop. Drivers are available with varying numbers of channels, so distributed installation with communications via the DMX protocol keeps the high current wiring lengths down. Most areas I see as having 2x white strips, 1x blue, and 1x RBG. That's 6 channels - so a 12 channel drive smack in the middle will give me 10 meters of coverage. With up to 31 drivers per string (DMX calls them "universes" - the physical controllers like you see at a concert generally can control multiple 31 device strings, multiple "universes"), the gets me a total of 310 meters of covereage - for you non-metric people that's over 1000 linear feet of lights, which is far more than the combined linear length of BOTH decks of my layout.
Seems almost a no-brainer to me, much as I would like to build my own drivers, it seems nearly pointless. Arduinos are borderline performance-wise, at least the regular 8 bit ones. DMX runs at 250kbps - over HUGE distances. Which just goes to show how conservative Bruce Chubb was with CMRI, using the same phycial layer, he recommends slower speeds for 'larger' layouts which still would have a fraction of the distance DMX can support.