Skip to main content

I looked at the issue, and to build a board to this schematic would cost about $4.50 in parts (including PCB), would still need to be assembled, and have to sell for at least $10 for it to be reasonable.  I suspect that's more than most folks would want to pay for it, but that's the realities of doing stuff here and not in China.  I have no outlet to purchase the parts on the boards they sell for $3 for anywhere near that price.  They buy in thousands and have really great prices at that.  Labor costs them next to nothing, and most of this would be automated anyway.

 

Here's a simple diagram of what the module might look like.  With those values, it could go from around 45ma down to around 5ma.  The pot would be surface mounted and accessible from inside the car.  You could make the resistor a bit lower, say 22 ohms and have up to around 57ma max current.  For passenger cars, I can't imagine needing more than 45, I light most of mine on 20-25 with a single CL2, either the 20 or 25ma model.

 

 

 

Pass_Car_Lighting_Module

Attachments

Images (1)
  • Pass_Car_Lighting_Module
Last edited by gunrunnerjohn

Questions for gunrunnerjohn, RailRide, electronics gurus:

 

Revisiting this topic. I ordered some parts and decided to play with the circuits outlined in this thread.  My input voltage is from a cheap Lionel transformer from the late 90's.  I turn it on full most of the time for this stuff to simulate 18v power bricks.

 

I first built the circuit below provided by RailRide.  It powered my strip of 12v LEDs (24 LED's) and they were very bright.  However, after a minute or so the LM317 got hot enough that I couldn't keep my finger on it for more than a 3 seconds and it started to get that "burning in electronics" smell.  I measured the voltage at the LED strip and it was 19-20v DC.  I measured the current several times and it showed around .465 A (almost ½-amp?).  

 

I tried different values for the resistors (R1=75 ohm, R2=680 ohm) to get the LM317 output voltage to 12...which worked, but the current stayed the same at right under 0.5 amp and the LM317 got even hotter to the touch.

 

Railride's Circuit:

 

railride

 

 

 

 

I then built and tested gunrunnerjohn's circuit below, using a 470uF cap instead.  No problems with heat, but voltage was measured at 8.4v DC and the current was 20ma as expected.  The problem with this circuit is I'm not sure the LED's will be bright enough.

 

Gunrunnerjohn's Circuit:

 

gunrunner

 

 

 

 

My first question is, am I doing something wrong with Railride's circuit or are my results to be expected as far as voltage, amperage, heat, etc.  Is it correct that these measurements exceed the common specs of a 12v LED strip?  With the heat generated from the LM317 in my test (assuming I didn't make a circuit mistake), I cannot imagine this would not melt or warp plastic passenger car shells if not kept away on a circuit board or heat sink.

 

Second, is there another CL2-type driver to modify gunrunnerjohn's circuit that would allow the LED's to be brighter but still not cause a heat issues or exceed the specs of the LED strip?  I like this circuit for it's simplicity but I feel it may be just a bit dim. (Don't say it!   )   

 

Thanks for any input.  You guys are great!

 

 

Attachments

Images (2)
  • railride
  • gunrunner
Last edited by Bobot

Bobot,

gunrunnerjohn's driver is a constant current driver. The voltage will increase or decrease with the number of leds that it's powering.

 

You can keep adding them on until you reach the voltage supplied by the track or transformer.

 

The part that's missing from the photo that was in the deleted thread is to put in a PTC resettable fuse. The hot wire(s) from the collectors or the transformer are connected to one leg and the power for the driver is connected to the other leg.  Circuit protection for your work. He now uses 220uf 50v caps. Your 470uf will keep the lights on for a while. It's there for flicker elimination. The choke is there to prevent signal interference for DCS, but may as well use it for TMCC and Legacy, too.

It doesn't get hot. The leds will be bright. Always getting maximum current. Buy them with the various lumens output for brighter or dimmer.

 

The LEDs should be connected in SERIES.

 

The thread that pine creek referred to was deleted for some reason. The remaining thread has the photo posted above.

 

I have used this circuit and it is really nice. I didn't get the axial caps the first time, so those circuits took up more space.

 

#30 wire is used with snap-able connectors(when needed) to unplug the roof mounted lights when removing the roof from a passenger car. I lost the link for those.

 

here's another that got deleted Question on 300 led rope. The deletion stuff is starting to p!*^ me off.

 

Anyway, hope that helps you. grj's driver is a sweet circuit.

 

Edit: I forgot to mention that grj's boards with adjustable voltage are sold by forum sponsor Henning's Trains, if you don't want to roll your own and want variable voltage control.

Last edited by Moonman

I find a car with a full length strip of LED's is plenty bright with 20ma, which is why I went with that in my original upgrades.  Knowing that some folks like them brighter and dimmer, I did the passenger car lighting modules with the adjustable current output.

 

The thread Passenger Car Lighting Controller details the capability of my lighting controller.

The subject of the original post was something like "Lights sucked all the voltage out of my transformer" or something like that.  It contained many references and schematics.

 

Moonman, GRJ wired the LM317 for constant current.  They can also be wired for constant voltage, which is the route I took.  There is a manual that gives the ohmage for the two resistors to get different voltages as desired.

Thanks for all the input guys...I'm learning and really appreciate the details.

 

Based on the fact that John went with the LM317 in his production circuit board with adjustable dimming, I'm guessing that is the way to go if the CL-2 circuit isn't bright enough.  And now that I know his production board exists for sale, I would just buy that to retrofit my cars.

 

However, electronically speaking, I am interested in the circuit I built and how it compares to John's production version.  And my questions still remain:  is it correct that my LM317 getting hot and should I be seeing 20v DC with 0.45a when using a 300 ohm resistor to make the LM317 act as a current regulator for a 12v LED strip of 24 LEDs?  Why would it be OK to run a 12v strip off 20v, and why is it showing 0.45 amps?

 

Thanks for helping me understand.  I'm interested in playing with these components and understanding how and why they do what they do.

 

- Bob

FWIW, while I use constant current drivers for LED projects at work for the ability to adjust brightness easily, I have settled on two constant voltage circuits for passenger cars.

The common LED strips will in fact light with 9 volts applied. No modification to the strip is required and a miniature 9 volt regulator (LM78L09) will drive 3-4 times the LEDs required for any O gauge passenger car. The only downside is you get no light until approximately 9 volts are applied to the track (too lazy to do the math). This is fine for those who run Command only. You can still dim the lights to suit with a single series resistor. Once you find one that gives the brightness you desire, no need to test further.

 

The other circuit I use is for cars likely to run in conventional though it will work in Command as well. This uses an LM317L set to around 3.2 volts. Lights will come on at around 3 volts track voltage and stay relatively constant up to full track voltage. The downside of this circuit is 2 of 3 LEDs in a strip has to be shorted out which can be tedious. Still it puts out plenty of light and would be my preferred circuit for all cars. I have pics of the various cars but can't access them right now.

 

You can find the circuits for these my searching on the part numbers above at Digikey and downloading the spec sheets. Standard boiler plate stuff.

Last edited by Norton
Originally Posted by Moonman:

 

here's another that got deleted Question on 300 led rope. The deletion stuff is starting to p!*^ me off.

 

 

I checked this link (since it was a thread I posted on quite extensively) and found that it wasn't deleted--the URL references "Page 4", but there are only three pages in the thread:

 https://ogrforum.com/t...ight-question?page=4 

 

Delete the "?page=4" at the end and it will take you to the beginning of the thread, (corrected link)

 

---PCJ

 

Last edited by RailRide

My guess is he hooked up the LED strip to the "Vout" pin of the LM317 rather than the "ADJ" pin.  It's kind of confusing but for constant-current mode you connect the load to the ADJ pin.  In the constant-voltage mode, you connect the load to the Vout pin.  Also, as shown below, the ADJ pin is on the "left" side of the package which is not intuitive when looking at the component hookup diagram.

 

lm317 to220

Attachments

Images (1)
  • lm317 to220
Originally Posted by RailRide:
That still leaves open the question as to why his attempt to copy my arrangement of components was overheating the LM317. Mine certainly didn't get hot--I'm sure I would have smelled them while running the cars I installed them in.

 

---PCJ

Take a close look at your picture, the resistor placement seems to be difficult to see, perhaps his is wired incorrectly?  Maybe a schematic should be included as well?

 

resistor placement

Attachments

Images (1)
  • resistor placement

Would one or two clear photos posted by bobot of his assembled components make it easier to diagnose the overheating problem?  I ask because I will shortly be undertaking the LED lighting of LIONEL PW-style 15" passenger cars, and if I run into a problem, I'd think posting photos of what I've assembled/installed would make it easier to diagnose the problem.

Originally Posted by gunrunnerjohn:
Originally Posted by RailRide:
That still leaves open the question as to why his attempt to copy my arrangement of components was overheating the LM317. Mine certainly didn't get hot--I'm sure I would have smelled them while running the cars I installed them in.

 

---PCJ

Take a close look at your picture, the resistor placement seems to be difficult to see, perhaps his is wired incorrectly?  Maybe a schematic should be included as well?

 

resistor placement

Well, looking at the LM317 with the heatsink facing away from you, the resistor is bridging the left and center pins together.

Here's a recycled photo from a previous LED thread.  In this case the tiny surface mount resistor value was 62 ohms (well, 61.9 ohms) which sets the constant-current output to about 20 mA.  If you have good enough eyesight and can manage these tiny parts, the LM317T plus 62 ohm resistor performs the same function as the constant-current 20 mA CL2N3 discussed above.

 

The resistors are a couple pennies each.  The out-of-pocket cost of a LM317T + resistor is maybe 1/2 the cost of a CL2N3.  The LM317T takes up more space but can handle more power; this can be a factor in train applications where transformers set to command voltages mean the input voltage to the regulator be 25V or more.

 

LM317T

Attachments

Images (1)
  • LM317T

Here are some pics of MTH cars I have done. Room lights off and camera set to manual in attempt to avoid compensation by the camera. LEDs are "warm" white.

 

9 volt constant voltage and no series resistor using a 15" LED strip in an 18" car this just reaches the vestibules.

 

Pass_9V_white

9 volt circuit with a 51 ohm series resistor and LEDs painted with red magic marker.

 

Pass_9v_red

3 Volt circuit driving 1/3rd of the LEDs compared to above. Overall intensity similar to the original incandescents.

 

Pass_3V

 

Pete

Attachments

Images (3)
  • Pass_9V_white
  • Pass_9v_red
  • Pass_3V
Originally Posted by subway vic:

Pingmam; If you're doing PW Lionel that have bayonet type bulbs, you can buy bayonet type leds that just plug in like a regular bulb from Town and Country Hobbies for $2 each, cheaper if you buy 60 pieces. I just did my 2500 series cars and work fine.

I even put them in my 2343 F3s. Vic

Vic, can you post or send a photo (my email is in my profile) of this application?  One of the virtues of the LED strings in the roof is the even lighting they provide compared to the "hot spot" effect of the OEM LIONEL lighting in 2500 series passenger cars.

 

I posted a photo showing an OEM 2500 car and an LED equipped equivalent and the difference was very noticeable.  If I can get the same evenness and brightness with the LED replacement bulbs you referenced, I'm there.

Originally Posted by subway vic:

Pingman, I don't have any pictures, but in my case I'm runing DCS & TMCC the cars were very bright. True they light up like the original design, not like the modern ones. Vic.

Vic, this is the photo I posted a couple of weeks ago; a handheld, low light shot at ~18 VAC which shows the even lighting of the lower car and the hot spots in the OEM LIONEL upper car's lighting.  (Note:  the lighting in the upper car in its vestibule is due to the absence of the interior door.) 

 

DSC01871

Attachments

Images (1)
  • DSC01871
Last edited by Pingman
Originally Posted by Gene:

Pingman,

 

How did get the lighting in the lower picture? I like the evenness and the color.

 

Thanks

 

Gene

It's gunrunnerjohn's recipe; i.e. the LED warm white LED's that come on a reel of 300 LED's.  A forum member I contacted provided me with the actual amazon and ebay items to buy; and, then did the installation on one of my cars so that I could have an actual installation to follow when I do my 25 or so cars.

 

One thing I notice in the photo are the spot lights below the car where the light sockets have been removed.  Probably just put some electrical tape on the frame to cover them.

Originally Posted by gunrunnerjohn:

I keep a roll of ductape around for those holes in the chassis.

 

Ductape will no doubt be superior to electrical tape--thanks for the tip; no home (or car) should be without ductape.  Back many years ago, a buddy's car broke down late at night.  He called me; I met him; and, after applying more than a few loops of ductape between his car and mine, successfully towed him about 10 miles to his house.

OK men, reporting back as promised.  Per Stan2004's post, I played around with the circuit and moved the wire going to the LED's from Vout to Adj on the LM317 when using it as a constant-current regulator.  Initially I found reference online to the pinouts of the LM317 that differed from what I now believe is correct per Stan2004's diagram above.  The LM317 doesn't seem to heat up much and I let it on for a while without any obvious and immediate problems.  I ended up using a 47 ohm resistor and measured around 8.5v and 25ma to the strip with decent brightness.

 

I spent some time and fashioned up the circuit on perfboard as seen below.  Imagine how cool I would be if I actually knew what I was doing!  Yikes. 

 

 

Bobot-LED-Circuit

 

 

 

Bobot-Perfboard-Closeup

 

 

 

Now just a matter of wiring up to the pickups and installing the board and LED strip in the car.  Since I was just playing I have a feeling my LED strip is too long for my car so when I cut a group or so of 3-LED's off the strip I wonder what to expect and if I will have to adjust the resistor?

 

Thanks to all for the info and guidance on this guys.  I learned a good bit in the process.  It's a neat project.

 

- Bob

Attachments

Images (2)
  • Bobot-LED-Circuit
  • Bobot-Perfboard-Closeup
Originally Posted by Bobot:
... so when I cut a group or so of 3-LED's off the strip I wonder what to expect and if I will have to adjust the resistor?

It appears you have 9 3-LED sections.  In a constant-current configuration, if you cut off a 3-LED section the total light output will stay constant which means the remaining 8 sections will each be a tad brighter.  Stated differently, the 25 mA which was previously split 9-ways will now be split 8-ways so each section gets proportionally more (about 12% more) current and will be proportionally brighter (more light output).

 

I don't think you'll notice a 12% change in brightness.  But if so, change the 47 to 56 ohms to drop overall current to about 22 mA which split 8-ways will give each 3-LED section about the same current as the original 25 mA split 9-ways.

 

When you cut off a section of 3, you'll see the voltage on the strip rise from 8.5V to, say, 8.8V (or so).  This is because to deliver the same overall current into the strip, the LM317 increases the voltage so the LED sections see a slightly higher voltage and thereby operate at a slightly higher current.

 

OTOH, IF you had gone with a LM317 constant-voltage configuration the voltage on the strip would stay constant as you cut-off sections.  The current in each section would stay constant as would the brightness of each section.  But the overall current would decrease.  For example, if the 317T was configured for 8.5V constant voltage, you'd measure the same 25 mA overall current when driving 9 sections.  If you cut off 1 section of 3-LEDs, the voltage would stay 8.5V, the current per section would stay the same, but the overall current would drop to about 22 mA.  And the total light output would drop by 12%.

 

Now if you followed all that, you'll understand why I guessed you initially hooked the strip to Vout instead of ADJ.  In a constant-current configuration, the 317 adjusts its output voltage until it measures the programmed current (25 mA or whatever) flowing thru the resistor.  This means the load has to be on the "other side" of the resistor (i.e., on the ADJ side) rather than on Vout side.  With the strip hooked up to the Vout side, the current flowed directly into the LED strip and not thru the resistor.  So the 317 kept increasing its Vout voltage to no avail.  And putting 20+ Volts on the 12V LED strip, the LEDs go very bright and the current got very high.  That large current flowing thru the 317 is what made is so hot.

 

Long winded, but you did say you wanted to understand how and why they do what they do.

Add Reply

Post

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

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