Skip to main content

I can find 12v warm white LED strips on Ebay, but I am running conventional, so need 3 - 6 volt LED strips which will light using my 1033 PW transformer when I crack it on.  Aren't the 12v strips for DCS or TMCC  where there is contant voltage on the track?   Or perhaps I need to use single LED's that light up at a low voltage ??  I already know how to build the circuit using a bridge rectifier, capacitor, and resistor.   Any advice appreciated.                                

My youtube channel

 (19150dava)

Original Post

Replies sorted oldest to newest

I have used gunrunnerjohn's constant current driver with a CL2 on the roll strips that break off in sections of 3 leds. I just tested a caboose with only 3 and it lights at 6.6 volts ac track voltage. Fluke true rms meter.

Anyway, it's all about your driver circuit. the leds raw will light at 2.3 - 3.4 depending on the color. It's the driver circuitry converting the AC and controlling the voltage or current that uses the rest.

The 20110 LED Regulator was really designed with command operation and constant voltage in mind.  I like the module Stan shows for conventional operation.  Had I known then what I know now, I might have attempted to do a fancier module with the capability of operating over a wide range of voltages.  I think that ship has sailed at this point.  It would probably be a larger module than the one I came up with, but of course it would also be more capable.

This topic has rekindled my desire to convert some of my passenger cars to LED lighting. 

I found the following modules on Amazon for $12.98.

]WGCD 10pcs Mini 360 DC to DC Buck Converter Step Down Module 4.75V-23V to 1V-17V       

  • The input voltage: 4.75 V - 23 V
  • The output voltage: 1.0 V - 17 V
  • The output current: 3 A (Max)
  • Conversion efficiency: 96% (maximum)
  • Switching frequency: 340 KHZ

 

I realize I need to add a rectifier and capacitor for AC current input.  Does anyone know a reason not to use these units?

Last edited by GregM

Those modules only handle 23V DC maximum, rectified 18V track power will usually exceed that, something to keep in mind.  I use similar modules that have a 28VDC maximum in a little P/S module I created.  You'll need at least a rectifier and a filter cap, and for DCS compatibility, the 22uh choke.  Here's The board I used and my module perched on top that  adds the choke, bridge, and filter caps.  Note the hole in the upper PCB to allow for voltage adjustment.

This is the schematic for the module.  I have a jumper option to use full-wave or half-wave rectification.  That's useful as most TMCC stuff has a common AC and DC ground.  If I use this in a TMCC environment where I want the DC to have a common ground, I use the half-wave feature.  If I don't mind the DC requiring total isolation, I use the more efficient full-wave feature.

John H posted:

I have a good reason right here. $18 gets you two modules and the only thing to add is a light strip. That was my point in my post. I think it's a pretty good deal.

 In Greg's defense, he's talking about ten of the modules for that price. Of course, you do have to add the other components to make a complete solution.

Attachments

Images (2)
  • mceclip0
  • mceclip1
Last edited by gunrunnerjohn

GRJ, thank you for the information you provided above.  It has been a long time since l did electronic "stuff."  A couple questions if you don't mind, what is the reason for "common ground" setting with LED lights?  Also would adding a resistor between the unregulated DC and the module be sufficient to prevent voltage overload?

 

The reason for my option was as I stated, sometimes I need a common AC and DC ground.  Other times, such as my passenger car lighting module, I don't have that requirement and I use a bridge rectifier.

The resistor is probably not going to do the trick.  You can buy the proper 28V rated module for the same price on eBay, that's the way I'd go. 

You could use some back to back 5V Zener diodes to drop the voltage a bit for each module.  A kludge, but it works.  You still need the other components, rectifier and capacitor, also the choke if you use DCS.

If you're going to get into soldering components together, you can probable roll your own easier than using the modules.  For low currents for LED lighting, constant current is the way to go.  You can tune the brightness by varying the resistor value.  This is basically how my lighting modules work.

Constant Current LED Passenger Car Lights

Attachments

Images (1)
  • Constant Current LED Passenger Car Lights
Last edited by gunrunnerjohn

I go with 35V capacitors, they've always been sufficient.  The voltages on 18 volts end up being in the 26 volt range on the capacitor.  25V capacitors would probably fail pretty quickly with that voltage.  The circuit above is good for around 50ma without any heatsink, if you're driving 12V lighting strips with an 18V AC power supply, that results in .8 watts power dissipation in the LM317.  It'll get around 70C or so, but within it's operating ratings.  If you want more current, add a heatsink to the TO220 regulator package.

If you haven't yet placed your DigiKey order consider a 40 cent trimpot to adjust brightness.

http://www.digikey.com/product...1/3306F-201-ND/84621

In place of a fixed-value R1 in GRJ's diagram above you could use, say, a 27 ohm resistor in series with a 200 ohm trimpot to give an adjustable R1 range of 27 to 227 Ohms.  This would give a current range of about 5 mA to 45 mA.  This adjustment method is discussed/documented in one of GRJ's previous posts.

I realize you're not doing eBay but for anyone contemplating a roll-your-own you can get 200 ohm trimpots on eBay for about 10 cents each shipped.

200 ohm trimpots for 10 cents each

Attachments

Images (1)
  • 200 ohm trimpots for 10 cents each
Last edited by stan2004

Hi all - I did this conversion to my Polar Express but something seems amiss (don't want to hijack the thread): I don't have constant brightness (it still varies with power applied), and now there are tiny sparks/crackling where the wheels meet the rails when it rolls.  There's a very small hum too, changing pitch depending on power applied and whether or not the light strip is attached to the Buck (no strip = no hum).

I used these components:  Buck Converter,  Light string (3528)Bridge Rectifier, 220uf capacitor.  Piece of cake to solder up.  I removed both bulb sockets in there and ran the incoming power wires to one "~" of the bridge and the incoming black wires to the other "~".   "+" and "-" out to the Buck.  The power wires have a stripe on them (at least one does) and initially attached to the red wires of both sockets, opposite for the black wires.

The light string for the car lights up so I know the AC-DC conversion is happening.  Could something be improperly grounded or crossed (I don't see how, but...)?  Are my components wrong?  What should I check?

Thanks much

It has only been two years, but I finally sat down this month and learned enough about the KiCad EDA program to design a PCB to use the parts bought back in 2017.  Only ordered three PCB's from OSH Park to make sure I didn't screw up somewhere along the way.  Now to just wait for them to get here.

I believe I have the PCB matching the schematic, PCB layout passed all of KiCad's checks.  I just don't know if I left enough room around the components (using through hole for now) to be able to actually populate the board.  I think so, but we will see in a couple weeks or so.  I reworked the PCB a couple of times to get its dimensions under 1 inch by 2 inches.

 

Greg did say he's using a through-hole component design...a wise first step.  The components in his version would be about 50 cents per board in small quantities on eBay.  So even if the extra square inch increases board cost, in DIY quantities it might be a wash if eBay from Asia is an option.  Then there's the assembly and handling hassles of surface-mount parts which require the dexterity and eyesight of a teenager.  

 

Edited to clarify board dimensions.

This is the very first PCB I have ever had produced for me.  I only ordered three boards this go around for prototyping.  I hope I get better as I go along but for this time I was content when I got the board under two inches long, first layout was about 2.25 inches long.  Decreasing the board length did bring the cost down from my first attempt.  I really don't intend to produce that many boards in the long run, but if eventually these boards help others that would be nice.

 

P.S.  I never had the eyesight of a teenager when I was a teenager, let alone now.

 

 

Last edited by GregM

GREGM; I think it's commendable that you are brave enough to launch into designing and ordering your own PCB's from OSH Park. Hopefully they will work as you envision. I just received 2 orders of boards from OSH Park using grj's Gerber files which he kindly provided, and they look great. Kudos to grj for all the research and design work!  My first impression of the boards is how small they are. Yikes! I know they are going to be fiddely to build, but hopefully I am up to it. Time will tell I guess. I expect it will be one of those things where just as you finish the last one, you find the best way to do it. 

Rod

Rod, I always find something that could be done better AFTER I send the boards out to fabrication!

Ted, I just picked parts out of the PCB package, I'd have to scout them out.  I know they exist, because all the 3D models are of real parts.  The one I did specifically look up was the 470uf capacitor, I was trying to do it with a 13mm part, but couldn't find one that was the right value.

I have made a couple of additions to the circuit design.  I will wait until after I receive the ordered PCB boards to layout a new PCB to match this schematic.  Once I actually see the board I can work on placing the components closer together.

 

 

Let me know if you see something wrong.

Thanks for every ones input.

 

Attachments

Images (1)
  • mceclip0

John,

I reread your post at the top of this page again.  I see the confusion now, my board isn't 2 inches square, it is just under two square inches of PCB material.  I calculated the dimensions as 1.95 inches by .85 inches.  OSH Park calculated the dimensions as 1.96 inches by .86 inches.  My first attempt was (my calculation) 2.25 inches by .85 inches.  Decreasing the length of the boards reduced the cost for the three boards to under $8.00.

 

Rod,

Got distracted, meant to say Thank You when I typed this post.

Last edited by GregM

I'm confused about the PTC, the regulator has current limiting, the chance that the PTC will be used is pretty small.  You'd be better off with a universal design to toss in the 22uh choke for DCS in that spot.

----------------------------------

I missed that 10mm cap, I could have made that board smaller with that one.  With the 10mm diameter, it's now 1" x 3/4".

Attachments

Images (1)
  • mceclip0
gunrunnerjohn posted:

I'm confused about the PTC, the regulator has current limiting, the chance that the PTC will be used is pretty small.  You'd be better off with a universal design to toss in the 22uh choke for DCS in that spot.

----------------------------------

I missed that 10mm cap, I could have made that board smaller with that one.  With the 10mm diameter, it's now 1" x 3/4".

 

The PTC is there to prevent the wires from melting in the event of a derailment that causes a short between the rollers.  I have read that recommended here.

gunrunnerjohn posted:

I missed that 10mm cap, I could have made that board smaller with that one.  With the 10mm diameter, it's now 1" x 3/4".

John that is a petty nice compact board layout, and it looks like it would be simple enough to build, even for a hacker like me. Is that a complete ready to go Gerber file? And how did you create it so fast?

Rod

Last edited by Rod Stewart

It's not a Gerber file yet, but that's just because I didn't export it from the PCB layout package.  As for the creation speed, there are only a handful of components on the schematic, I just picked representative package sizes for the parts and then plopped them on the circuit board.

Before I committed it to Gerber, I'd probably want to insure I can actually buy the parts represented.  They are real part sizes, but they may not be readily available.  I refer specifically to the pot and the choke, the rest I know are available in that footprint.  I also thought that maybe trying to get a pot that is adjusted from above would be a better option.

Here you go, everything you need to build that board, including the BOM.  All the parts are available on Digikey, but you can also probably find even cheaper sources for some of them.

Have fun.

Pass_Car_Lighting thru-hole [top adjust) 3DPass_Car_Lighting thru-hole [top adjust) Schematic

Pass_Car_Lighting thru-hole (top adjust) BOM.pdf

Pass_Car_Lighting thru-hole Gerber (top adjust).zip

Attachments

Wow John, that is fantastic, thanks for taking the time out of your schedule to whip that up.  I especially appreciate that this is a leap of faith for you inasmuch as it competes with your kit already being sold by Hennings. But it is a great help for those of us who want to roll our own, and don't have the talent or time to whip up our own board design, and find thru-hole building easier.

It should be possible to make a 100 ma version simply by switching R1 & R2 to 12 and 50 ohm values respectively, correct? I may also poke around and see if Digi-Key has a larger value C1 cap, still in a 10 mm body style and 35V, accepting that it will be taller of course.

As always, my hat is off to you! 

Rod

Below is a link to a circuit that I just purchased which accepts AC in and outputs DC. It has an adjustable output as well. The cost is about $2.00 each The size is about 3/4 in by 2 in and is about 7/8 in tall. This should cover most the of questions being asked in this thread.

https://www.ebay.com/itm/AC-DC...p2060353.m2749.l2649

 

AC-DC-LM317-Linear-Regulator-Step-down-Rectifier-Buck-Power-Module-1-25-37V-1-Gh

 

 

Thanks;

idea-thinker

 

So with the latest incarnation, the bare per-board cost is down to ~$1.30...and if you do eBay for components, the total out-of-pocket is $2 per board!  Wow.

So I realize the thread is titled pax cars, but mate this with a 99-cent (free-shipping from Asia) eBay relay module, set the current to max (45 mA) and for $3 you have an insulated-rail relay module for gate-crossing mechanisms, dual-aspect signals, etc..  Again, this means you must know which end of a soldering iron to hold, but I believe this would be the lowest cost ever for an insulated-rail relay module.  

idea-thinker posted:

Below is a link to a circuit that I just purchased which accepts AC in and outputs DC. It has an adjustable output as well. The cost is about $2.00 each The size is about 3/4 in by 2 in and is about 7/8 in tall. This should cover most the of questions being asked in this thread.

https://www.ebay.com/itm/AC-DC...p2060353.m2749.l2649

That's certainly an option, but note that it's twice the size and also taller.  I use an LM317T that has the tab that can be folded down to reduce the height of the board to just over half an inch.  This board also lacks DCS compatibility, so you'd have to add an external choke.  Finally, my design is a constant current design for a very specific reason, it's very easy to get a precise intensity as the adjustment from dim to bright covers the entire pot travel.

Different strokes for different folks.

stan2004 posted:

So with the latest incarnation, the bare per-board cost is down to ~$1.30...and if you do eBay for components, the total out-of-pocket is $2 per board!  Wow.

If you buy a quantity of boards, you can do considerably better than that for this size board.  Here's one estimate, total looks to be around 20 cents a board for quantity 100.

Attachments

Images (1)
  • mceclip0

For those of us running conventional AC, here's another boost-buck unit, similar to what Stan showed earlier, FWIW. 

https://www.ebay.com/itm/DC-DC...047675.c100010.m2109

The one thing it lacks is a trimpot to fine-tune the LED brightness - though I note that these modules seem to come in different flavors, ~3, ~4, 5, 9 and 12 vDC, so doing something at 9v might be an option.  I have the LED strip I've been thinking about using (another one of those "round-tuit" projects) so I could see how bright that strip is at 9 v*.  What caught my eye was the small size - 19 x 14 x 4 mm.  These are $2.85 each, which is a bit pricier than others, but given the number of passenger cars and cabeese I have, affordable.  Still have to have an AC to DC bridge in front of them.

- Rich

*ps  Most of the LED strips I've seen and used (undercabinet lighting in my kitchen) are rated at 12 vDC input -- is there any reason NOT to operate them at a lower input voltage - e.g. 9 v (assuming the light quality is ok at that voltage)?

stan2004 posted:
                                                             **************

....but mate this with a 99-cent (free-shipping from Asia) eBay relay module, set the current to max (45 mA) and for $3 you have an insulated-rail relay module for gate-crossing mechanisms, dual-aspect signals, etc..  Again, this means you must know which end of a soldering iron to hold, but I believe this would be the lowest cost ever for an insulated-rail relay module.  

Okay Stan, where is the eBay link/item number for "99 cent ... relay module" for those of us who do not keep up with this stuff.

You and GRJ are terrific contributors to the content here--THANK YOU BOTH.

richs09 posted:
*ps  Most of the LED strips I've seen and used (undercabinet lighting in my kitchen) are rated at 12 vDC input -- is there any reason NOT to operate them at a lower input voltage - e.g. 9 v (assuming the light quality is ok at that voltage)?

Gosh, no!  Although my modules use constant current, the running voltage of the 12V LED strips actually have from around 8.5V to 11V applied, depending on the current setting.  You don't damage LED's with low current, only with high current!

Pingman posted:
stan2004 posted:
                                                             **************

....but mate this with a 99-cent (free-shipping from Asia) eBay relay module, set the current to max (45 mA) and for $3 you have an insulated-rail relay module for gate-crossing mechanisms, dual-aspect signals, etc..  Again, this means you must know which end of a soldering iron to hold, but I believe this would be the lowest cost ever for an insulated-rail relay module.  

Okay Stan, where is the eBay link/item number for "99 cent ... relay module" for those of us who do not keep up with this stuff.

Dozens of choices on eBay searching for "12V relay module"...such as

ebay 99 cent relay module 12v

The pairing is described in greater detail in this OGR thread which includes a video proof-of-concept.

And as GRJ points out, at higher board quantities the cost is almost unbelievable!  Again, this is for the DIY enthusiast as assembly-wiring-soldering is required.   Though it should be pointed out that off-the-shelf relay modules for insulated rail triggering typically don't have capacitor storage which provides anti-chatter protection and companion inductor for DCS compatibility.

Attachments

Images (1)
  • ebay 99 cent relay module 12v

Gene, yes, I uploaded the zip file to Osh Park.  Price is batches of three shipped.  I have an account as I have ordered other boards from them.  I don’t know if GRJ has shared the file or not, but I’m sure he would if need be.  I could share it under my account also, I guess.(but is it mine to share?)  FWIW, I had to use an old windows laptop to upload file.  Copied it to the desktop then within OP’s site pointed the “browse for files” box to that zip.  Was automatic from there.  Really pretty easy.  Couldn’t make it work from the iPad I use for forum work.  Let me know if you need something additional.  Perhaps someone else has help for you also.

Last edited by TedW

Gene, like TedW, I downloaded GRJ's zip file and then uploaded it to OSH Park to order some boards.

Also, after I read your post and TedW's reply, I tried to use the link in his post, and copy and paste the zip file link that GRJ posted.  I had no success in doing this.

As GRJ said, his zip file is safe, and if you want the boards it's an easy way to order them.

And props to TedW for letting me know off forum of the availability of these GRJ/JWA boards.

Now, I'v off to re-read the link in Stan's reply regarding the relays, and then order some on the bay.

Once you have an account set up with OSH Park, it is very easy to do a repeat order or a new order. Uploading John's Gerber files to their site is quite simple. Once the site has processed the files you can see all the front and back views, the silk screening, etc. And they accept Paypal. My first two orders went very smoothly and I had the boards delivered in about 2 weeks.

Rod

Actually I find that the time consuming part of this is ordering all the needed components. John has kindly provided the BOM, but you still need to go to the Digi-Key site and find each component and add it to your cart, then when its all correct go ahead and place your order.

I usually set it all up in a spreadsheet first and I am just working on this for these new boards. If anyone is interested I can post it once complete. Another option would be a snip of my cart summary once complete. I could just post that here for anyone to duplicate. I wonder if its possible to export the electronic cart summary from their site, anyone know?

Rod

Last edited by Rod Stewart

I have another question for John/Stan/et al. - but first, Kudos.  I'm sure some of you have seen the missive from Alan Arnold (OGR head honcho) about his growing concerns about the escalation of negative posts in some of the discussions (something I recall Rich Melvin - his predecessor - worrying about).  Fortunately I haven't seen a lot of that and in apparent (and delightful) contrast to that concern has been the very helpful discussions and inputs in this forum on converting passenger car lighting to LEDs.  This is a topic that has been discussed quite a bit over the past several years yet the really knowledgeable and helpful folks like John, Stan and others haven't dismissively said 'we covered that back in "aught six" - go look it up...' but instead have engaged with the various 'new to the topic' folks and those of us that have been in the 'contemplation' stage for a couple of years.  So I just wanted to say "thanks" for both the knowledge and experience, the willingness to offer help, but also the attitude.  Nice goin'...

So my question - looking back at the earlier part of this thread, which is a couple of years old - John, you made a comment about not having anything grounded to the frame (in response to a troubleshooting question).  I'm not sure how Lionel passenger cars are wired up - I'm a Flyer guy and in my case, the track power circuit is typically the 'hot' pickup through one truck which is insulated from the car frame and a wire lead from that truck (usually from the top of the truck rivet) to one side of the light bulb socket.  The other side of the socket is usually connected physically to the metal floor of the passenger car, which in turn is connected to the other truck (uninsulated).  My plan is to take the leads from the two trucks (I'll solder a lead to the rivet on the uninsulated truck or connect it to the frame) as inputs to the AC side of the AC-DC bridge circuit.  Nothing 'downstream' of that connection will electrically connect to the metal floor/frame.  Is that consistent with not grounding to the frame?  I suppose alternatively I could insulate the other truck from the floor/frame - which turns this into a somewhat bigger project.

Thanks,

Rich

richs09 posted:

So my question - looking back at the earlier part of this thread, which is a couple of years old - John, you made a comment about not having anything grounded to the frame (in response to a troubleshooting question). 

RIch, that comment is on the DC side of the lighting module. Since the frame & center roller go into the bridge and there's a diode between them and the output + or -, you MUST NOT ground either + or - on the output. For a Flyer 2-rail scenario, same caution applies, except it's the power from the wheels. You don't have to isolate anything, just make sure the DC out of the module doesn't connect to the frame anywhere.

TedW posted:

FWIW,  Here’s a cart share from digikey on the products for the GRJ-JWA pcb.  I’m very sure as GRJ sez, you can find cheaper prices from Asia.  YMMV

Edit:  I already had the chokes which is why they aren’t included.

digikey Parts

Did you see this?

Attachments

Images (1)
  • mceclip0
gunrunnerjohn posted:
richs09 posted:

So my question - looking back at the earlier part of this thread, which is a couple of years old - John, you made a comment about not having anything grounded to the frame (in response to a troubleshooting question). 

RIch, that comment is on the DC side of the lighting module. Since the frame & center roller go into the bridge and there's a diode between them and the output + or -, you MUST NOT ground either + or - on the output. For a Flyer 2-rail scenario, same caution applies, except it's the power from the wheels. You don't have to isolate anything, just make sure the DC out of the module doesn't connect to the frame anywhere.

TedW posted:

FWIW,  Here’s a cart share from digikey on the products for the GRJ-JWA pcb.  I’m very sure as GRJ sez, you can find cheaper prices from Asia.  YMMV

Edit:  I already had the chokes which is why they aren’t included.

digikey Parts

Did you see this?

Yeah, backordered, but they said estimated ship was 2/4/19, so I went with it as I’m in no rush.  The boards will take a week or so too.  BTW, thx again for the share, that should work for everyone directly to the order page.

https://www.digikey.com/produc...497-1575-5-ND/591677

Here's a version that is apparently in-stock at DigiKey and less expensive 67 cents (vs. 87 cents).  While I appreciate the convenience of 1-stop-shopping from DigiKey, given the ~2 week delivery time from OSH park, I wonder if going with an eBay Asia source is something one might consider.  IIRC the LM317 goes for less than 20 cents on eBay (from Asia).

Wow, a lot of activity on this thread since it was revived the other day.  

I had time this afternoon, so I revised my version of the PCB layout compressing it as much as possible.  Got it down to 1.08 by .86 inches.  Total area .9288 square inches.  Just like the previous time, OSH Park made the dimensions just .01 inches bigger in each dimension.  Not an issue, just made me laugh.  The price for three boards went down as expected.  Probably go ahead and order another batch of three boards.  Only reason I will get just three is to make certain I can actually populate that small of a PCB without problem.  It has been a long time since I have done any circuit board building.  Last time I did it I actually etched the PCB board using chemicals.

Yeah, that’s the order, but I recommend anyone to search elsewhere because of the price difference.  I even had some of the parts on hand but were not an exact match so I ordered the speced out part anyway.  The LM317T for example.  I have 30 in my bin I paid $4.35 shipped!   Oh well.  And from the page John copied it sez the regulators are available to ship.  I just got an email saying the order has shipped...  

Last edited by TedW
gunrunnerjohn posted:

You do realize you can actually truly delete the whole post, click on the Take Action link on your post and use the Delete function.

I find it therapeutic to manually delete incorrect information and publicly acknowledge the error of my ways--but, NO, I never knew I could delete the reply in its entirety--make it disappear which I have now done.

gunrunnerjohn posted:

Sounds like a more reasonable size, got a picture of the new layout?

Don't get me wrong, I truly appreciate all the help and advice I have received but honestly I don't understand why it's now "a more reasonable size."  I stated that I had never before designed a PCB that would be manufactured for me.  This is all new to me, I am only doing this to attempt to actually complete a project I purchased components for quite a while ago.    It was never my intention to compete with anyone with regards to board production.  I probably would use your board design as many others have if I didn't already have different components on hand.

 

As requested, here is a picture of the revised board layout.

 

PCB Ver 1.0.2

Attachments

Images (1)
  • PCB Ver 1.0.2

I think one of your limiting factors is your connection pads seem to be too generous.  I think there's where a lot of space could be gained.  Look at the space I leave for power and LED connections and the space you leave.  I think just shrinking those would do wonders for the size of the board, you'd have a much better chance if fitting stuff in a more compact manner.  The rest of your layout seems to be reasonably space efficient.  However, I think you'll find you can also squeeze the components a bit tighter than you have when you get actual components on the board.

What I do to size things if I'm not sure of the spacing is to print a 1:1 image of the board and then set the components on the image to see if I have any conflicts.

Attachments

Images (2)
  • mceclip0
  • mceclip1

What GRJ said.  For a first try it's quite good!   

If I understand it, your board design program does not have 3-D rendering (?) so it can be hard to visualize how the parts interact without actually assembling a unit.  Back in the day, when designing a board using through-hole parts what I found useful was to place the parts on a 0.1" grid perforated project board.  Of course this kind of measure-twice, fabricate boards-once practice has been turned on its head with what has become zero fixed $ penalty to fabricate revisions.  

perforated project board

Photo shows a couple other ideas.  Those screw-terminal connectors come in 0.1" pitch.  I see you're using 0.2" pitch which obviously take up a good chunk of real-estate.  I realize you already have the connectors but maybe next time if size matters.  Also, for space savings nothing wrong with standing up the resistor like black diode in above photo with pads spaced 0.1".

Separately, on your latest revision when you install the component you cover the silkscreen.  Your software should allow you to drag the white silkscreen reference designators (R1, RV1, J1, etc.) or labels/symbols (AC IN, +, etc.) so you can see them after the parts are mounted.  Note on GRJ's board how lettering is outside the footprint of the parts.  And for your RV1, it's usually customary to have a square-pad for pin 1; or for symmetrical parts like this (can be installed either way) at least some silkscreen indictor showing notch, tab or in this case maybe a small circle showing which end has the adjustment screw.

Attachments

Images (1)
  • perforated project board

Stan,

  KiCad does have a 3D viewer but not all of the components have 3d models so when you view the board there are some components missing.  Yes, KiCad allows the silk screen data to be moved anywhere desired.  Since some of the information will be covered up, I put information on the bottom silkscreen to identify important information such as AC IN and DC OUT and which pin was + on the DC out.  As for the other designations, all the information needs to tell me is what goes where during assembly which is why I numbered the pins on RV1 so I know which way to mount it.  Now that I think about, it might not really matter which way that part is placed provided pin 2 is always in the middle.  (That's meant as humor in case someone miss interprets.)  Perhaps that's why there is no square pad for pin 1 on the footprint since it really doesn't matter.  I didn't create any of the footprints used, I just picked a matching (at least the best match I could discern) footprint from those available in KiCad for each component and placed the footprint on the PCB.

Greg, if you're thinking of doing more PCB work, I'd encourage you to give DipTrace a look.  They have a free version that's good for small projects like this, and it has a really extensive library.  The other big bonus is I have found it's very easy to create new parts for stuff that's missing from the library.  For the 3D views, you can pick a similar part and very easily stretch or shrink it in any dimension to give you a decent 3D view.  I usually get a pretty accurate representation of my boards with the 3D view.  This allows me to plan the layout and be reasonably sure things will fit.

Here's an example, my Super-Chuffer II board and the 3D view.  The biggest discrepancy in this view is the added socket on the real build, that allows me to easily change the uP firmware without a soldering iron.

 

Attachments

Images (2)
  • mceclip0
  • mceclip1
GregM posted:

...

Now that I think about, it might not really matter which way that part is placed provided pin 2 is always in the middle.  (That's meant as humor in case someone miss interprets.)  Perhaps that's why there is no square pad for pin 1 on the footprint since it really doesn't matter. 

Hmm.  I wonder if this explains why on the 99 cent voltage regulator modules on eBay that use similar looking blue multi-turn trimpots, some versions have CW increasing voltage while others have CCW increasing voltage. 

In case it is of any value here is my final as ordered BOM. It is an Excel file based on grj's BOM, just fluffed up a bit. Note the first is the standard 5-45 ma version and the second is the optional 100 ma version. The latter differs only in the resistors and the C1 cap value. Most stuff I ordered from Digi-Key but the trim pots come from Asia. The ebay item numbers are shown where applicable. My quantities are 50 and 30 respectively; I just want enough on hand to build up to either of those quantities.

BOM Snip 01-27-19

If needed email me offline and I can send you the spreadsheet which you can modify as desired. You can change your build quantities to suit. For now I have ordered only 12 boards from OSH Park, just so I can make sure all is well for fabrication, though I imagine they will be just fine.

Rod

Attachments

Images (1)
  • BOM Snip 01-27-19

  Along with Johns note on squeezing in the tiny resistors and diodes first, I'd include surface mount parts to that list of firsts too.

   I like to work from the center to the edges. On large, tight boards, I may fit the large suff first, "just incase".   If there is a problem fittlng something, it's usually because of part change, etc. (change in overall size, different case shape or even leg orientation, etc) it is most often easier to change position of a small part (like instead of flat, you can solder one end close to the board and stand it straight up, and use the long leg looped like a candy cane, for going back to the board.)

On small boards it hardly matters.

  Bending legs slighty to the side/opposite ways/etc. to hold the piece (or 4), solder, clip excess leg off ..... "Next!"   ...next,next next, clean solder flux of board, install and call it done.

Adriatic posted:

I like to work from the center to the edges. On large, tight boards, I may fit the large suff first, "just incase".   If there is a problem fittlng something, it's usually because of part change, etc. (change in overall size, different case shape or even leg orientation, etc) it is most often easier to change position of a small part (like instead of flat, you can solder one end close to the board and stand it straight up, and use the long leg looped like a candy cane, for going back to the board.)

One thing you clearly should do, especially with a new board design, build ONE and see if everything fits.  Don't ask me how I know this.

I don't want to flog a dead horse but just FWIW here is a snip of my revised BOM as ordered 01-27-19. All onshore parts have arrived; offshore are still somewhere out there. Waiting for the OSH Park boards to arrive so I can do a test fit. Looks like they are due to ship about Feb 5. 

My BOM Snip R2.12 02-02-19

You can see my all in board costs are $2.45 for the 45 ma version and $2.77 for the 100 ma version, not including fab time of course. The latter is primarily due to the higher cost of the 1000 uf cap C1. Interestingly I calculate that at 100 ma the load on the 12 ohm R1 resistor will be right at its rated 1/8 watt, so it will likely run a little warm. Had I known this at the time of ordering I would have ordered 1/4 watt resistors instead. For any about to order the 12 ohm resistors, maybe consider 1/4 watt. For the 27 ohm resistors 1/8 watt is fine.

I ordered the pots and the LM317's offshore primarily due to better pricing, and it looked like the 317's were out of stock at Dig-Key at the time. Longer delivery times of course, but I can't build any of these until we get home in April anyhow, though I can test fit the components I have. The offshore pots look like knockoffs of the Bourn 3362P units, so dimensionally they should be good. (The vendors even call them "3362P") These are rated at 1/2 watt, so there should be no power or overheating issues.

Rod

Attachments

Images (1)
  • My BOM Snip R2.12 02-02-19

TEDW; yes that does read kind of weird. The $63.34 is based on the build quantity of 50 in the upper yellow cell, and would be the total had I ordered the full 50 boards. I only ordered 12 boards to start with as shown to the right of that total, because I wanted a small quantity to test and make sure they meet expectations. I figured 12 would be good for starters as it will be enough for 2 sets of cars. If all is good I will then order more. Notice I ordered zero boards for the 100 ma version. Hope that makes sense. The spreadsheet does the math of course. 

Rod

Rod Stewart posted:

TEDW; yes that does read kind of weird. The $63.34 is based on the build quantity of 50 in the upper yellow cell, and would be the total had I ordered the full 50 boards. I only ordered 12 boards to start with as shown to the right of that total, because I wanted a small quantity to test and make sure they meet expectations. I figured 12 would be good for starters as it will be enough for 2 sets of cars. If all is good I will then order more. Notice I ordered zero boards for the 100 ma version. Hope that makes sense. The spreadsheet does the math of course. 

Rod

Gotcha

Hey Jim, others will have their favorites, but here is my take. Standard type 3825 strip LEDs in either warm white or bright/white depending on your preferences will work fine. You probably wont want the much brighter 5050 style; that would likely be overkill in a passenger car. They mostly come in 5 meter or 10 meter spools. 12 vdc is common for command operation, but if you want to run conventional 5 vdc strips would likely be better because they will start to light up at a much lower track voltage. 

gunrunnerjohn's constant current boards seem to do a great job and are easy to set up since they have a bridge rectifier, filter capacitor, and 22 uH choke all on the one board. Easy to adjust the led brightness to your liking. For those of us who want to roll our own and are OK with thru-hole board assembly it is fantastic that grj has made available his new board Gerber file that enables anyone to order them up on OSH Parks site. Kudos to John for opening up his shop, so to speak! For most of us cad design of circuit boards is not too likely to happen. 

Rod

 

Rod, I received similar emails today, also four minutes apart, for the revised version of my circuit board.  Yesterday, I received a pair of emails that applied to the first version of the board.  The first set of boards should get here Friday according to the USPS tracking info.  So possibly the second batch will get here Saturday.

 

Last edited by GregM

So FWIW   here’s the GRJ designed “roll ur own” board made up.  End result, works great.  Took maybe 15’ to make the first one and hook up to CW80 for testing.  Nice variation on current(brightness), holds lighting very well after power down(flicker control), and should be able to hide easily in the cars.  Also, for comparison, photos of GRJ’s own board and another I have used in the past that is fixed voltage out.71D7764A-D13E-4D08-BE5E-1234E4E17B9BB7954DDB-E5A4-49A6-B1F2-85F8E24B84E40C2CA135-A5BB-44EC-9306-600F76EDC0035B9DA766-1F3F-44FE-9E8E-14FC773B3821C4CB450A-AAEF-4F42-BE09-FB44119BAEAE 

Attachments

Images (5)
  • 71D7764A-D13E-4D08-BE5E-1234E4E17B9B
  • B7954DDB-E5A4-49A6-B1F2-85F8E24B84E4
  • 0C2CA135-A5BB-44EC-9306-600F76EDC003
  • 5B9DA766-1F3F-44FE-9E8E-14FC773B3821
  • C4CB450A-AAEF-4F42-BE09-FB44119BAEAE

Ted, looking good and great to know it works as intended, though I had no real doubts. Still giddy with excitement?  

I received my boards today also and did a test fit of the components I have, including the 1000 uF filter cap and 12 ohm R1. All good. Now just waiting for the R2 pots and the LM317's to arrive. These new boards are somewhat bigger than grj's first design for sure, but that is mostly due to the former being SMT instead of thru-hole I imagine. They are still very small and should fit most any pax car I would think.

IMG_3381

Rod

Attachments

Images (1)
  • IMG_3381

Well, I made something that's transformer powered, it has a high intensity 40ma LED on top, a button on the side for the AC, and spring loaded test pins to probe the board.  The pins are spaced to mate with the lighting module. I press the AC button, and it applies power to the module and the test LED on top lights.  When I release the button, I can check for the decay of the light to see that the cap is working.  I typically test these before they're split, so I have 25 completed units on a panel, and I just probe them one at a time for test before I separate them.

 

Attachments

Images (2)
  • mceclip0
  • mceclip1

Ahhh, ok.  The photo I remembered from a couple of years ago didn’t show the whole device, only the probe end.  That’s why I thought it might be a battery.  No matter, xfrmr is easier.  And I have some button switches for other projects.  It’s something I’m sure I can replicate.  The difference of course is your boards come assembled, so I’ll have to do mine individually as I complete each one.  Will work well I’m sure.  Thx again for the help.

I was thinking of a test device something like this, but I was going to include a 0-100 milliammeter. Then when I find the light intensity I liked the best, I could quickly set each module to the same intensity (same milliamps) and thereby have all cars in the set at the same brightness level, at least in theory. I thought this might be kind of good because taking one or two cars apart again to adjust the light intensity after completion would be kind of a pita.

Rod

Last edited by Rod Stewart

As long as you have the same number of similar LED's in each car, that will work.  You don't actually need an LED in series with the meter, this is a constant current design, so you can just measure the current with a direct connection across the output.

Ted, the reason mine come assembled is it would take me a LONG time to hand solder a thousand modules!

GRJ's application is to test large batches whereas the rest of us would be testing much fewer.  In case anyone missed it, what's neat about GRJ's tool are the spring-loaded test pins to make reliable contact to the 4 pads.  These so-called "pogo pins" (like a kid's pogo stick) are not expensive but hard to imagine using them if assembling/testing just a handful of boards.

As GRJ states, the LED confirms capacitor operation.  You see the brightness decay rather than going instantly dark.   Note that GRJ mentions a high intensity LED that can presumably take the full current.  So if using this for a 100mA board, use a suitable LED.

As for skipping the LED and hooking an ammeter directly across the board's output.  A digital meter updates a few times per second so its numbers would jump making it tricky to see a smooth decay.  An analog (needle) meter's spring slows the needle's return-to-zero and could be mis-interpreted as a smooth decay.  Separately, if using a digital mA meter, most are fused when making mA-range measurements.  Fuse values may be, say, 1/4 Amp or whatever.   OTOH, using the 10 Amp setting may not give enough resolution when testing or calibrating the range of these low-current boards.  So if the purpose of the mA metering is for test (before and as opposed to calibration), you may pop the mA fuse if there was an assembly error - fuse is replaceable but kind of a hassle.

Here are a few pictures of the revised version of the PCB for my lighting project(s).  These pictures were taken with my camera so they are a little better than the iPad pictures in the conversion thread.

Top and bottom of version 1.0.2.

Pic 1 - PCB Ver 1.0.2

 

Comparing length of this board to the original version.

Pic 2 - PCB Ver 1.0.2

 

Comparing width.

Pic 3 - PCB Ver 1.0.2

 

PCB populated with same components as original version.  I will say it was more work to solder the components to the board this time as it was harder to hold everything in place and apply the heat and solder.

Pic 4 - PCB Ver 1.0.2

 

Even though it was harder to populate the PCB it does work. 

Pic 5 - PCB Ver 1.0.2

 

I was a little concerned, but both connectors do fit on the revised PCB.  However, I plan on using only one connector on the remaining boards.  I will solder the supply wires directly to the PCB's AC IN pads along with a PTC to protect the wires.  Since the PCB is attached to the floor of the cars' with double sided tape, one connector for the DC OUT wires will be sufficient to allow me to completely detach the roof if needed in the future.  Paint the roofs white maybe?

 

Okay, breaks over, back to soldering.  Well maybe later.

 

Edited***  I ended up using connectors for both the AC input and the DC output.  Since I am using the original wiring from the wheel sets in the PE cars, it was just easier than trying to solder the wires directly to the board.

Attachments

Images (5)
  • Pic 1 - PCB Ver 1.0.2
  • Pic 2 - PCB Ver 1.0.2
  • Pic 3 - PCB Ver 1.0.2
  • Pic 4 - PCB Ver 1.0.2
  • Pic 5 - PCB Ver 1.0.2
Last edited by GregM
Rod Stewart posted:

...For grj's new board, I am thinking that since the center rail is the one carrying the DCS signal, the center rail pickups should be connected to pin 2 of the AC input (ie the round hole) since this connector runs through the 22 uH choke. Does that make sense? 

Doesn't matter.  Choke performs same function either way.  

Consider that the board works with 2-rail where DCS "hot" could be either rail depending on car orientation.  

 

 

stan2004 posted:
Rod Stewart posted:

...For grj's new board, I am thinking that since the center rail is the one carrying the DCS signal, the center rail pickups should be connected to pin 2 of the AC input (ie the round hole) since this connector runs through the 22 uH choke. Does that make sense? 

Doesn't matter.  Choke performs same function either way.  

Consider that the board works with 2-rail where DCS "hot" could be either rail depending on car orientation.  

 

 

Stan and John; thanks for this confirmation. Kind of what I figured but thought I would ask anyway.

TEDW; I notice in your pictures the 22 uH chokes you are using are considerably bigger than the ones I wound up with, which are 285 ma. Do you perchance have a current rating for them, and perhaps a P/N? Thanks.

Rod

gunrunnerjohn posted:

As long as you have the same number of similar LED's in each car, that will work.  You don't actually need an LED in series with the meter, this is a constant current design, so you can just measure the current with a direct connection across the output.

Ted, the reason mine come assembled is it would take me a LONG time to hand solder a thousand modules!

I hadn't considered this John; great idea. I will likely just do it that way. My head is so engrained with the notion that an ammeter always needs to be in series with the load; couldn't see the forest for the trees.

BTW, that reduction in size of the newer Super-Chuffer board is really impressive. I am thinking the ones I bought a few years ago must be the older style because I recall the pcb material was green. Thinking I should order a couple of the newer ones; do I buy those direct from you, or through Hennings? Too bad the 8 pin DIP couldn't be rotated 90 degrees; that would have allowed even more of a reduction in width by the look of things.

Rod 

TEDW; I notice in your pictures the 22 uH chokes you are using are considerably bigger than the ones I wound up with, which are 285 ma. Do you perchance have a current rating for them, and perhaps a P/N? Thanks.

Rod

=======================

Rod, those were sourced from a recommendation for components on the little fixed voltage board I use(pictured earlier in the thread).  Used in this new board cause I had them.  As you see on the invoice they are 560mA.  I don’t know why, just following advice.  I don’t even have DCS, but in the event I do, or the cars are sold, it’s easy to add now and be done with it. They are way pricey from this supplier, I should have ordered offshore.  Hope this helps.  TW

C81153BD-287F-4DAA-9796-5D52CA064FA4

 

Attachments

Images (1)
  • C81153BD-287F-4DAA-9796-5D52CA064FA4
Last edited by TedW
Rod Stewart posted:
gunrunnerjohn posted:

As long as you have the same number of similar LED's in each car, that will work.  You don't actually need an LED in series with the meter, this is a constant current design, so you can just measure the current with a direct connection across the output.

Ted, the reason mine come assembled is it would take me a LONG time to hand solder a thousand modules!

I hadn't considered this John; great idea. I will likely just do it that way. My head is so engrained with the notion that an ammeter always needs to be in series with the load; couldn't see the forest for the trees.

BTW, that reduction in size of the newer Super-Chuffer board is really impressive. I am thinking the ones I bought a few years ago must be the older style because I recall the pcb material was green. Thinking I should order a couple of the newer ones; do I buy those direct from you, or through Hennings? Too bad the 8 pin DIP couldn't be rotated 90 degrees; that would have allowed even more of a reduction in width by the look of things.

Rod 

Rod, Stan does make a good point about the type of ammeter.  The LED is very good at showing you the decay of the light to verify the cap is doing it's job.  You can also put a resistor in parallel with a standard diode to split the load while providing a greater load on the board output for testing.

The S-C board reduction really helped in putting into places it wouldn't fit before.  However, any potential savings by rotating the DIP socket is an optical illusion, the socket measures 10mm in either dimension.  As I've stated previously, the reason for the DIP part is ease of programming.  If I didn't have a removable part, I'd have to provide a programming socket, so there would be no board area savings.  The removable part also makes it easy to do a code change in the field as you don't have to unsolder the board, just pop the new chip in.

Rod Stewart posted:

TEDW, those don't look too bad at 27.8 cents a piece imo. The ones I got from Digi-Key were 24 cents each and are only 285 ma current rating, about half the 560 ma of yours. Either is ok for these boards of course with the small outputs they have. I think for any future orders I will use the ones you ordered. 

Why would you pay that kind of money for a thru-hole choke?  285ma is more than enough for the application, why go bigger for capacity you'll never use?

How about 50 for $1.51, about three cents each?  eBay: 142908355800

Most of the components on the board can be sourced for peanuts if you are patient and wait a few weeks for shipments.

gunrunnerjohn posted:
 
How about 50 for $1.51, about three cents each?  eBay: 142908355800

Those are what I use for thru-hole rolling-stock applications.  Looks like your pads are spaced about 0.4" so here it is on a 0.1" grid project board.  All the listing says is "1/2 Watt" which is an odd way to spec an inductor.  In any case these measure in at about 0.7 Ohms (DCR) so no problem wrt current handling.

22uh ebay inductor 400mil

Attachments

Images (1)
  • 22uh ebay inductor 400mil

"Why would you pay that kind of money for a thru-hole choke?  285ma is more than enough for the application, why go bigger for capacity you'll never use?

How about 50 for $1.51, about three cents each?  eBay: 142908355800"

grj; I was really thinking of possible future applications where the current might exceed the 285 ma. For the constant current LED boards I agree, these are plenty big enough. Gonna go ahead and order some of those that you referenced on the bay, thanks! 

Rod

stan2004 posted:
gunrunnerjohn posted:
 
How about 50 for $1.51, about three cents each?  eBay: 142908355800

Those are what I use for thru-hole rolling-stock applications.  Looks like your pads are spaced about 0.4" so here it is on a 0.1" grid project board.  All the listing says is "1/2 Watt" which is an odd way to spec an inductor.  In any case these measure in at about 0.7 Ohms (DCR) so no problem wrt current handling.

22uh ebay inductor 400mil

So check my math on these guys. Based on a resistance of 0.7 ohm, and power limit of 1/2 watt, I am getting a max allowable current of about 837 ma. Sound right? 837 ma across 0.7 ohms give a voltage loss of 0.586 volts, and this times 837 ma = 0.50 watts. Did I get that about right?

If correct, these guys should work for lots of misc. applications, but not engine power pick up of course. For that purpose I use chokes rated at about 3 amps.

Rod

Patrick1544 posted:

John

If I want to order your board and use it with TMCC operations only, do I eliminate the choke and put a jumper in its place?

Or, for 3 cents each, DigiKey will sell you 0 Ohm resistors  

https://www.digikey.com/produc...4ZT0R00CT-ND/1830304

 

gunrunnerjohn posted:

You can obviously do that, though for three cents, I'd put in the choke in case the cars find their way on a "guest" layout.

What he said.  Just think of it as a 0.7 Ohm resistor which in this context is the same as a jumper! 

gunrunnerjohn posted:

You're right, of course, I just divided by two.  I should have thought about it.  In any case, I'd limit my planned power dissipation to around half the rated capacity.

Got it. Good advice!  It's hurting my head thinking about it.

I actually confirmed it using Ohm's Law:      V=I * R  and   W=I * V

                                    So substituting:    W= I*(I * R),    or  W = (I)^2 * R

                                    Re-arranging:       I^2 = W/R       or  I = (W/R)^.5

Now if you use W = 0.25 watt and R = 0.7 ohm; I works out to 597 ma, or call it 600 ma for simplicity.

Whew, time for a gin and tonic! 

Rod

Last edited by Rod Stewart

Ted, I am sure I am well over the total you listed for my lighting boards but at this point I am happy with the way they are turning out.  Like you, I usually source my parts from the US but that may (probably not but never say never) change since DigiKey's minimum shipping is now $4.57 for less than eight ounces using USPS.

I have enough parts for 15 boards (need to order more boards) that should be sufficient to upgrade the highest priority cars.  After that maybe order enough for 25 more cars.  Not sure I have 40 passenger cars but there are many cabooses I could upgrade someday.

Last edited by GregM

Greg, I think you are being too modest in your ability.  Designing and producing your own pcb was no small feat it seems to me.  Certainly not something I would tackle even for an experiment.  Cost may be one factor, but you should take no small amount of satifaction in your first successful attempt at building the module.  It is a fun part of the hobby to put these little projects together, and they actually work!  That little fixed voltage module I pictured earlier can be made for a $1.10!! 

Ted/Greg, I think both of your projects have come along really well, and provided others with ideas for moving forward. My cost per module may be a little lower, mainly because I went offshore for the pots and the LM317T's. But on the other hand, you both already have operating modules, whereas I am still waiting! Haha. 

gunrunnerjohn; do you have any further info for that board supplier where you said it looked like they could be had for around 20 cents a copy? I will need to order more boards at some point, and fast delivery is not especially important. Can they run with your Gerber zip file as is?

Thanks, Rod

TedW posted:

So, FWIW, with all the high end computations being put forth(SNL’s Chevy Chase said, "It was my understanding that there would be no math.") here is the price breakdown on the lighting pcb, sans labor, for my assembly.  The parts were US sourced, so we all realize savings can be achieved with off shore sourcing. YMMV

So I am not providing anything "new" here, but it occurs to me that someone stumbling on this thread next month or whenever might be looking for a summary on exactly what/where to buy.  As I see it, there are 3 orders:

1. Bare boards from OSH Park using GRJ's link earlier (I can't seem to consolidate reply-with-quote from previous thread pages) so duplicating the link here:   https://oshpark.com/shared_projects/mMzHpYbk

$3.80 for 3 boards ($1.26/board) shipping included.  You will need to create an account.

2. Components.  I used TedW's DigiKey cart link from earlier and updated it to include the 22uH choke/inductor listed in the parts list (BOM) from GRJ's OSH Park link. https://www.digikey.com/short/p3mr70

I also notice you can select a "photo" view of a DigiKey cart:

dk cart

I entered quantity 25 for all the components.  So with new 2019 USPS shipping rates, the total with shipping would be about $71.  Or $2.85 per board.  You don't need to set up an account.  As TedW says, you can get this down to about $1 per board if you do eBay or similar.

3. LED strip.  Numerous OGR threads on options.  When you buy a "standard" 5-meter reel which is what you'd do for 25 cars, you'll end up spending a few pennies per LED or, say, 50 cents per car.

Attachments

Images (1)
  • dk cart

So has anyone else built some of these boards and have them in service?  I have some in a small 4 car Milwaukee Road consist.  They really do work well.  Am wondering however what component would be changed to dim the lighting slightly more.  Realizing of course that leds are either on or off at the minimum so adjustments any lower may not be possible.  Thx in advance for the thoughts.

36A6B65E-20E3-427D-9942-A3FAF46A5934

Attachments

Images (1)
  • 36A6B65E-20E3-427D-9942-A3FAF46A5934
TedW posted:

...Am wondering however what component would be changed to dim the lighting slightly more.  Realizing of course that leds are either on or off at the minimum so adjustments any lower may not be possible.  Thx in advance for the thoughts.

36A6B65E-20E3-427D-9942-A3FAF46A5934

TedW, are you saying you did set the adjustment to minimum...but you want that minimum value to be dimmer?

If so, you need to increase the value of R2 (the adjustment component itself).  It is presently 200 ohms.  You can get the same component in 500 or 1000 Ohms.  This would reduce the minimum brightness proportionally compared to the present 200 Ohms....so about 2-1/2 times dimmer (using 500) to 5 times dimmer (using 1000).  Changing R2 does not affect maximum brightness value.  So if this is what you're after, one of us can come up with an exact part number to order from DigiKey or wherever.

Stan:  Yeah, minimum lower.  I have the pot turned all the way down; I like my lighting dimmer than most folks due to glare.  Looks more natural to my eye.  I have some 1k pots, based on your info I’ll build a couple boards with that pot to see how it looks.  If, as you say, it doesn’t change the maximum, then no harm done.  GRJ, I didn’t know led were 0 to max adjustable.  Thx for that.

gunrunnerjohn posted:

You can also just swap out the 27 ohm resistor for a larger value that will lower the adjustment range.

So to be sure I “got the gist” of what you advise, if I use a larger value pot, I increase the range of brightness control to the lower end, but if I use a higher value resistor at R1, I move the whole range lower.  Which means not as bright at the highest level.  Which for my taste is not a problem.  What would the value of R1 change to?  50 Ohm; 100 Ohm or more/less?  Thx again for the help.

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.
×
×