Skip to main content

WB47:  I may be corrected here, but I don’t think cutting(shortening) a segment of the strip would have any effect on the brightness of each led.  Of course overall brightness would be impacted just because you’re using less lights in the car.  I have used only a single 3 led segment inside a Lemax building with the module, with no problems at all.

Actually Ted, prepare to be corrected!

The lighting module is a constant current design, so the pot sets a current value for the LED strip.  The current gets divided evenly with all the LED's on the strip.  If you cut some segments off, the remaining LED's get more current, thus they are brighter.  You can demo this yourself, just connect a strip segment to the lighting module, then parallel another strip segment.  You will see a visible reduction in the brightness of the LED's that were already lit.

I'd think you can count on one hand the number of guys using GRJ's off-the-shelf lighting modules willing/able to replace the microscopic surface-mount trimpot and/or resistor.

That said, has someone proposed tacking on a 2-cent resistor around ~2000 Ohms to any pair of copper islands at the boundaries of each section of 3-LEDs?  This would "steal" current from the output of the lighting module.  So while the module itself would retain its output range of 5-to-45 mA of current...the amount of current going to the LEDs would be downward shifted to, say, 0-to-40 mA since the resistor would steal ~5 mA of the output.  There is some math to work out.  

stan2004 posted:

That said, has someone proposed tacking on a 2-cent resistor around ~2000 Ohms to any pair of copper islands at the boundaries of each section of 3-LEDs?   

Stan, you only need one resistor, right across the feed to the LED strip, I've tried this.  Values from 2.2K and down will vary the brightness of the LED's significantly without modifying the board.

cjack posted:

With silhouette windows you can put it on the floor...works well I found. But with the interiors you really have to put it on the underside of the roof. I haven't done a dome car.

Here's another thread where I did a dome and the observation cars.  Sometimes I can run the dome lights overhead if there's a roof panel, but for this upgrade, I had to do a sidewall mounted LED installation, no place on the roof!

Passenger car LED lighting (Again)

gunrunnerjohn posted:
stan2004 posted:

That said, has someone proposed tacking on a 2-cent resistor around ~2000 Ohms to any pair of copper islands at the boundaries of each section of 3-LEDs?   

Stan, you only need one resistor, right across the feed to the LED strip, I've tried this.  Values from 2.2K and down will vary the brightness of the LED's significantly without modifying the board.

John and Stan,

Your recent posts were interesting so I decided to do some experiments.  I found that a 3k resistor will just slightly dim the LEDS.  I also found that really low values, 270 Ohms for example, turns the LEDS off completely and immediately.  Anyone wanting to do this should start with John's 2.2K value and work their way down to get the desired brightness.

I also verified that the LEDS brightness does vary depending upon the number of LEDS attached.  More LEDS, less brightness.  That wasn't as noticeable when you just hook up different numbers of LEDS depending upon car requirements.  In some cars I used 15 LEDS and in a couple of baggage cars only six LEDS.

gunrunnerjohn posted:

The reasoning why low values kill the LED's is simple Greg.  Since the circuit supplies constant current, a very low value resistor shunt will simply use all the current available long before the voltage across the load gets high enough to light the LED's.

I actually thought any resistor connected across the supply would cause a short and kill the LEDS.  I was surprised when the higher value resistors didn't cause a short.  See I learn new things all the time around here.

TedW posted:

Where exactly would you place those ~ 2k +/- resistors to retrofit an led converted car?

E770B966-87AD-497B-9915-395F206EEDE0

Ted,

Here is a picture of where to connect the resistor to the LED strip.  I show a through hole resistor but a surface mount might be a better option if soldering to the LED strip.

image

Question for @gunrunnerjohn or @stan2004 does it matter if the board is constant current or constant voltage?

Attachments

Images (1)
  • image
GregM posted:

Here is a picture of where to connect the resistor to the LED strip.  I show a through hole resistor but a surface mount might be a better option if soldering to the LED strip.

image

Question for @gunrunnerjohn or @stan2004 does it matter if the board is constant current or constant voltage?

The behavior will be different.  If you have a constant voltage supply, the resistor would have to be in series with the power source. 

The reason the resistor is across the power source in this example is the supply is supplying a fixed amount of current, so by siphoning off current with the resistor, you have less current for the LED's, hence they are dimmer.

With a constant voltage supply, adding resistance across the power just increases the power consumed, but it doesn't change the voltage until you exceed the current capacity of the power supply, so the LED's are the same intensity.  Putting the resistor in series drops some of the voltage and dims the LED's.  The value of the series resistor will be dependent on the number of LED's you're powering, since the more current the LED's draw, the more voltage drop you'll have across a given series resistance.

Last edited by gunrunnerjohn
GregM posted:

Here is a picture of where to connect the resistor to the LED strip.  I show a through hole resistor but a surface mount might be a better option if soldering to the LED strip.

...

Question for @gunrunnerjohn or @stan2004 does it matter if the board is constant current or constant voltage?

1206 shunt on led strip

Right.  An SMD resistor might be easier to install and more tidy...of course you must have suitable eyesight to deal with tiny components!

So a couple comments about using an SMD resistor.  You want a power handling capability of 1/4 Watt or more (this also applies if using a leaded resistor as in GregM's photo).  In above photo, I show a "1206" size (.12" x .06") SMD resistor which is 1/4 Watt.  The value is not a 2K but this is for sizing only.  They are only a penny each on eBay but may be difficult to install since they just barely "straddle" the copper islands on the LED strips.  I'm sure you could do it but may get frustrating applying huge blobs of solder.  Your mileage may vary.

That said, you can also get slightly larger SMD resistors.  For example a "2010" size (.20" x .10") would nicely straddle a pair of copper pads and be easier to solder to the pads..as well as being physically larger and hence easier to see and manage!  A tad more expensive at 5 cents each so you may need to save up your allowance.

---

This resistor shunt method for dimming only works for constant-current boards like GRJ's lighting module.

Attachments

Images (1)
  • 1206 shunt on led strip
Last edited by stan2004

For those of you guys that want to earn DIY'er extra-credit, I found this video from an earlier OGR thread demonstrating how to apply the shunt resistor to dim an LED strip driven by GRJ's lighting module.

Idea 1.  Dim/bright lighting for night/day running.

So rather than a set-it-and-forget-it brightness.  If you install a simple 25 cent on/off switch with the new resistor, you can simply flip the switch to enable/disable the resistor.  You might have to cut a hole in the bottom of the car chassis to install the slide switch but you could at least then quickly convert the brightness between bright and dim.

Idea 2. Dim/bright (or off/on) lighting for passenger station loading/unloading.

Or using a 50 cent latching magnetic Hall sensor chip, a 10 cent magnet can enable/disable the external resistor to change the brightness of the LED strip.  As the video shows, as the train arrives at the station, the lights turn up so that passengers can get their luggage and load and unload at the platform.  When the train leaves the station, the lights dim so the passengers can get some shut-eye.

Last edited by stan2004
gunrunnerjohn posted:

Stan, where did you find the latching magnetic sensor, that's pretty cool!

In my parts bin.  LOL. 

Or, for Digikey disciples: https://www.digikey.com/produc...k=1&pageSize=500

What I think is cool about these devices is you can simply attach the "circuit" using 2-wires to any pair of copper pads on the LED strip.  Place the sensor on a wall of the passenger car.  Then use the "north" pole of a magnet to set the brightness to one level, use the "south" pole of a magnet to set the brightness to another level.  Lather, rinse, repeat.

Last edited by stan2004
stan2004 posted:
gunrunnerjohn posted:

Stan, where did you find the latching magnetic sensor, that's pretty cool!

In my parts bin.  LOL. 

Or, for Digikey disciples: https://www.digikey.com/produc...k=1&pageSize=500

What I think is cool about these devices is you can simply attach the "circuit" using 2-wires to any pair of copper pads on the LED strip.  Place the sensor on a wall of the passenger car.  Then use the "north" pole of a magnet to set the brightness to one level, use the "south" pole of a magnet to set the brightness to another level.  Lather, rinse, repeat.

Hey Stan.....you absolutely CRACK ME UP! I've been educating myself (reading all your fantastic post RE. Track Occupancy) to use them to trigger Atlas Signal System. Your posts are absolutely outstanding! 

Last edited by Junior

TEDW, that 1k R1 certainly does dim the leds way down. If you like running in simulated "night" conditions, that looks like the way to go. Most of my running is done at "dusk" or brighter conditions, so I like them a little brighter. Just personal taste.

Maybe what we need is a photocell that detects the room light level, and adjusts the leds accordingly? Grj, stan, what do you think, possible?

Rod

Rod Stewart posted:

TEDW, that 1k R1 certainly does dim the leds way down. If you like running in simulated "night" conditions, that looks like the way to go. Most of my running is done at "dusk" or brighter conditions, so I like them a little brighter. Just personal taste.

Maybe what we need is a photocell that detects the room light level, and adjusts the leds accordingly? Grj, stan, what do you think, possible?

Rod

Rod, the 1k pot is at R2.  I’m sure you have a typo there.  You are right, night time running is subdued lighting.  I don’t think I will use the minimum brightness setting with the 1k pot.  I will try the 500 ohm too when they get here tho.

Rod Stewart posted:

...

Maybe what we need is a photocell that detects the room light level, and adjusts the leds accordingly? Grj, stan, what do you think, possible?

Certainly possible.  That is, a circuit that uses a photocell to shunt a variable amount of current (depending on room light level) from the LED strip would only cost, say, $1 in parts.  OTOH if fiddling with brightness control, I'd think most guys would want direct access to set lights bright vs. dim, or on vs. off.  In my opinion of course!

 

TEDW, yes I meant R2 = 1K of course, my mistake, duh. Anyway glad that is working for you.

I have my first few cars done using grj's roll your own modules with R1 of 12 ohms and R2 of 100 ohms. This gives an output range of about 12 to 100 ma. The cars are 15" so I am using 14" long led strips, with 21 leds per car. For a first run I have them set at 90 ma output, so just over 4 ma each, and they are quite bright. For now I like how they look, I may run them a while like that then crank them down a little. Time will tell.

In thinking about the photocell variable brightness idea, I guess it would be overkill and unnecessarily complex. Plus everytime the train went under a bridge or through a tunnel the lights would dim, which would look kind of odd. 

Rod

 

Last edited by Rod Stewart
Rod Stewart posted:

...

I have my first few cars done using grj's roll your own modules with R1 of 12 ohms and R2 of 50 ohms. This gives an output range of about 12 to 100 ma. The cars are 15" so I am using 14" long led strips, with 21 leds per car. For a first run I have them set at 90 ma output, so just over 4 ma each, and they are quite bright.

If you're using the standard 12V LED strips, note that they are 2" (5 cm) sections of 3 LEDs...with each LED within a section receiving the same current.  So your 14" strip is being split into 7 x 2" sections with each section driven by about 13 mA (90 mA / 7).  Hence each LED is also 13 mA.  And, as you and others have reported, 13 mA LED current is mighty bright for O-gauge passenger car LED strips.

Rod Stewart posted:

...

In thinking about the photocell variable brightness idea, I guess it would be overkill and unnecessarily complex. Plus everytime the train went under a bridge or through a tunnel the lights would dim, which would look kind of odd.  

I don't think $1 for a few generic parts (photocell, 2 resistors, 1 transistor) is complex...especially if it attaches to the LED strip with just 2 wires.  Ironically, by using a 5 cent capacitor like on GRJ's lighting module, you can implement flicker rejection on the photocell circuit.  In other words, it would average the ambient lighting over many seconds or even minutes.  So short term variations don't change the LED brightness. 

Junior posted:
 

Hey Stan.....you absolutely CRACK ME UP! I've been educating myself (reading all your fantastic post RE. Track Occupancy) to use them to trigger Atlas Signal System. Your posts are absolutely outstanding! 

Well, as long as you're laughing WITH me and not AT me!

 

I typically run the LED's in 18" cars with 25-30ma from my constant current regulator, that's 24 LED's in 8 groups of three.  It gives nice illumination, even with fairly bright room lighting that's about 3-4ma through the LED's.  I can't imagine what they'd look like if I ran the LED's with 13ma, but they'd be WAY brighter than I like them!

Here's my Superliner upgrade, this has 39 LED's in 13 groups of three.  I cranked this one up to a measured 35ma from the regulator, so that's around 3ma for each LED.  I am considering popping the top and cranking them down a bit.

Attachments

Images (1)
  • mceclip0
gunrunnerjohn posted:

Actually, those are warm white, they just look whiter in the photo.  FWIW, that's in a brightly lit room as well.

We may have different definitions of "decent brightness level".

John, I wouldn't change a thing. Those look great to my eye, realizing of course that pictures can be deceiving.

Aside, I had one dud of the 12 boards I prebuilt while in AZ. Turned out to be a bad solder joint on the 317. Replaced it and all is well. One thing I don't like about OSH Park boards is that the pads are not tinned. This makes them challenging to solder, but doable. The bulk lot of boards I ordered from Allpcb has arrived and I am happy to see the pads are tinned. They should be easier to deal with. These boards look really good BTW. 

Rod

Add Reply

Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
Link copied to your clipboard.
×
×