a single resistor on common lead of signals?

[gregc]

do different colored LED intensities vary that much to really require separate and different valued resistors for each red, amber and green LED?

i need to adjust the LED intensity on a number of different signals and am thinking to just replace the separate resistors on each with a single resistor on the common lead.

a single common lead resistor would result in less current to each LED if multiple LEDs were active at the same time, possibly for testing, so no chance of damage to the LEDs

[wvgca]

individal resistors on each LED will give much more control over the intensities of thr LEDs, making them appear close to the same brightness

[richhotrain]

Yes , they do.

On Tomar Industries signals, the instructions recommend specific different LED values for each color.

They make a big difference.

Rich

[gregc]

i'm curious how they measured intensity

[richhotrain]

Maybe just visually?

Rich

[Overmod]

My two cents' worth would be to put enough resistance in the common return to ensure none of the LEDs will 'fry', set to the value for the 'dimmest' resistance, and then use trimming resistors (which might well be variable resistors!) to adjust the brightest values to match the dimmest.

And yes, you'd do it by eye.  Your own eye.  There are no tiny enginemen or FRA inspectors with micronized candela meters to fact-check your intensities.

If you're cheap, in the way I'm often cheap, you could measure the resistance across the variable, and replace it with the appropriate value or series/parallel combination of resistances, or burn a ladder to get your value.  That might then be a guide to making individual resistance adjustments for other heads.

[BigDaddy]

This is what RR Mel wrote some time ago:

He recommended the Bivar SMTL4-RGY LEDs, suggested using the following values to get the brightness of all three colours to be similar @ 5 volts:

Red - 1K ohms  Green - 560 ohms  Yellow - 180 ohms

"The values are for my (Mel's) Eye after cataract surgery (6 months) so I think the colors are pretty close to equal.   I ended up with the following @ 5 volts using my Arduino power.  

2ma for red  12.5ma for Yellow  4ma for green

The Yellow takes more current to equal the Red & Green brightness"

 

 

[gregc]

BigDaddy

Red - 1K ohms  Green - 560 ohms  Yellow - 180 ohms

that's a pretty big difference in values and i assume depends on the particular LEDs being used

one reason for considering this it to make signals from different venders have the same intensity and trying to match them.

wondering is an LDR could be used.  quite difference in sensitivity between red (700nm, yellow 600, green 550).   and then wondering about the relative sensitivity of the average human eyeball.

 

[gmpullman]

I probably have two dozen various signals around my layout with anywhere from one tri-color LED (an assembly of three individual LEDs actually) to signal bridges with as many as 24 individual LEDs.

I try to bench test each assembly but sometimes even after installation I still have to reduce intensity of particular LEDs for best visual representation.

Reds seem to be particularly "hot" as well as cool-blue that I use for the lunar white marker of a B&O style color position signal.

Quite obvious here I need to tone down the right hand arm showing the absolute stop aspect:

PRR_Signal-View-2 by Edmund, on Flickr

Same here, red is just too intense:

Signal_DB-east by Edmund, on Flickr

I use this three-color array from Evan designs for my searchlight signals.

LED_tri-color by Edmund, on Flickr

These tiny SMDs still exhibit what Henry shows that Mel pointed to above. Red needs most resistance, green, then yellow. Individual resistors is the only way I know to handle this. It seems the intensity of MY green LEDs are not the same as what Mel was using. I look for particular ones that give more of a "signal green" color than most yellowish-green LEDs.

The Lunar White on these B&O CPLs are still just a bit too bright. Haven't gotten around to adding resistance just yet...

Superliner_tail by Edmund, on Flickr

I use Logic Rail Signal Animators for nearly all my installations. They output 5 VDC to the lamps and allow for easy modification to add resistance to each cathode sink.

How would you use the LDR in your signal circuit? I could see lowering the intensity of the signal light output if your room lights are dimmed but I'd find this to be an unnecessary complication to the circuitry.

gregc

i need to adjust the LED intensity on a number of different signals and am thinking to just replace the separate LEDs on each with a single LED on the common lead.

I presume you mean "resistance"?

Good Luck, Ed

[mvlandsw]

"The Lunar White on these B&O CPLs are still just a bit too bright."

Ed,

  The marker lights on B&O CPL signals should not be lunar white. Lunar white is only used for the diagonal restricting indication on the main signal head. The marker lights use a clear lense with an icandescent bulb. A warm white LED would look right.

The marker lights to the left of the mast are yellow. They are used for the Approach Slow and the Medium Approach Slow indications.

 

Greg,

  If the relative intensity of the different colors is good and you want to reduce the brightness of all of them I would leave the existing resistors in place and add one to the common to adjust the brightness of all of them.

Mark

[gregc]

gmpullman

I presume you mean "resistance"?

yes, thanks

[jmbraddock]

As others have stated, different color LEDs have different brightness levels and it can be quite noticible to the human eye.

If you don't want use different resistors, you could go with all white LEDs with a single resistor and use either transclucent paint or markers to color the lenses.  

[Steve McDonough]

To test for a common intensity that you are satisfied with for any color you may make a test jig on a breadboard. A variable resistor or potentiometer should be used like a 5000 Ohm one no higher. As you test each LED make sure to center the pot or variable resistor to  the middle and go down from there. When you are happy with the brightness take a reading of the resistance of the variable pot to see what it is and write it down. This way you can put the same colored LEDs in a small container with the correct resistance on a sheet of paper with them. You can put a small linear pot in the breadboard but it is a little difficult checking what resistance it is at. Instead solder wires to two of the connections, one on one end and one to the center. Centralab or Calectro are easiest pots to do this with. For designing a simple circuit for LED needed resistance a search on the internet may help you to build a test jig with a breadboard.

[gmpullman]

Steve McDonough

To test for a common intensity that you are satisfied with for any color you may make a test jig on a breadboard.

I use a decade box similar to this one:

Decade Box

Mine is half-watt but there are quarter watt, lower tolerance models out there for a little less money. I install lots of LEDs in locos, structures, signals and passenger cars so I get plenty of use out of it. Others may not see the value in one but I find it a handy tool.

Regards, Ed

[Mike in NC]

Definitely different resistors values are needed.  I illuminate my structures with various LEDs but I like to vary the intensity so I use little potteniometers to make the adjustments.  

-Mike in NC,

[CSX Robert]

gmpullman

 

 

Steve McDonough

To test for a common intensity that you are satisfied with for any color you may make a test jig on a breadboard.

 

 

I use a decade box similar to this one:

Decade Box

Mine is half-watt but there are quarter watt, lower tolerance models out there for a little less money. I install lots of LEDs in locos, structures, signals and passenger cars so I get plenty of use out of it. Others may not see the value in one but I find it a handy tool.

Regards, Ed

 

I looked at that Amazon link and just had to laugh at part of the description:  "Resistance obtainable is from 1 Omega to 11, 111, 110 Omega in 1 Omega increments"

The unit of measure is ohm, not omega.  The name of the greek letter that represents ohm, Ω, is omega, but not the unit of measrue.

[mvlandsw]

Lighted signals, or any lights for that matter, are hard to photograph. If you adjust the exposure for the overall scene the lights are overexposed. When doing time exposures with film I would turn the lights off after a short time into the exposure which made them look more realistic.

I don't know if this can be done with digital photography. 

[Overmod]

mvlandsw

Lighted signals, or any lights for that matter, are hard to photograph. If you adjust the exposure for the overall scene the lights are overexposed. When doing time exposures with film I would turn the lights off after a short time into the exposure which made them look more realistic.

I don't know if this can be done with digital photography.

I think that LEDs being driven off continuous DC may have a different 'apparent brightness' to a given CCD imager than to a human eye -- the actual exposure time can be very short and hence the effects of the filters for color separation might be exaggerated.  In my opinion, you could easily test using a given camera both on- and off-axis, and if you were using double-pot trim on the two 'brightest' colors you could quickly (and cheaply) determine the optimal settings for photography... and with only a little work rig up something like a multipole double-throw electrical switch to change between 'viewing' and 'photo' display mode.

The other thing is if the LEDs are multiplexed or driven off some sort of PWM driver with 'off time' that leaves the LED darkened for infinitesimal intervals (with persistence of vision giving the 'brightness').  There are problems photographing modern LED locomotive lights with many digital cameras as the very short 'capture' may make a multiplexed device look dim or out in the 'taken' picture.