JNSF DIY lighting project (PCB based)

[Ochanomizu]

On 3/10/2019 at 1:12 PM, gavino200 said:

No circuit details here. But a testimony to successful results with 100uF

 

 

 

Hello Gavino200,

 

I have been quietly following this thread with interest.  I wonder if smaller capacitor will be acceptable, say 25uF or 50uF?  Have you tested such?

[gavino200]

1 hour ago, Ochanomizu said:

 

Hello Gavino200,

 

I have been quietly following this thread with interest.  I wonder if smaller capacitor will be acceptable, say 25uF or 50uF?  Have you tested such?

 

The smallest value that I've ordered is 100uF. But you have a point. It's worth testing lower values. Especially with a voltage regulator dropping the voltage to 3.3V we should be able to make do with less capacitance. I think I'll order a few.

[cteno4]

Lowering to the 3.3v won’t mean you can lower cap, it will behave the same with 3.3 or 5v. The amount of charge going to the leds is the same in each circuit. Going to 3.3v would let you go to a lower voltage cap like a 6.3v.

 

this is going to be a hard test to judge what dirty and bumpy track will give good flicker to overcome. Gavin thought he had trackmthat gave good flicker to try to negate. Then best to overkill it a little.

 

jeff

[gavino200]

2 minutes ago, cteno4 said:

Lowering to the 3.3v won’t mean you can lower cap, it will behave the same with 3.3 or 5v. The amount of charge going to the leds is the same in each circuit. Going to 3.3v would let you go to a lower voltage cap like a 6.3v.

 

 

But the resistance will be much lower. So less of the charge will end up being dissipated. 

[gavino200]

8 minutes ago, cteno4 said:

 Gavin thought he had trackmthat gave good flicker to try to negate. 

 

?

 

Quote

Then best to overkill it a little.

 

 

Agree in principle. But small size is always good as long as it's adequate. I don't mind testing it.

Edited March 12, 2019 by gavino200

[gavino200]

For a 3.3V circuit. It would be hard to beat this for size.

 

https://www.ebay.com/itm/50-pcs-1206-3216-Chip-Capacitor-1pF-100uF-SMD-Ceramic-Capacitors-Assortment/291870890748?ssPageName=STRK%3AMEBIDX%3AIT&var=590941858237&_trksid=p2057872.m2749.l2649

Edited March 12, 2019 by gavino200

[chadbag]

Note that those are 1210 size.  Except for the first cap pad the pcb will take 1206 (or 0805) but not 1210

[gavino200]

2 minutes ago, chadbag said:

Note that those are 1210 size.  Except for the first cap pad the pcb will take 1206 (or 0805) but not 1210

 

I switched the link to a 1206. Also cheaper.

[cteno4]

40 minutes ago, gavino200 said:

But the resistance will be much lower. So less of the charge will end up being dissipated. 

 

No you are planning on running the bank of leds on the same current in both. The capacitance is basically the volume of charge (ie number of electrons stored) and the voltage is the level of energy of each charge unit (ie electron). With the 5v the resistor sucks up the extra energy but the same amount of charge (ie the current) gets to the leds. With the 3.3v circuit less energy is taken off in the lower value resistor.

 

jeff

[gavino200]

6 minutes ago, cteno4 said:

 

No you are planning on running the bank of leds on the same current in both. The capacitance is basically the volume of charge (ie number of electrons stored) and the voltage is the level of energy of each charge unit (ie electron). With the 5v the resistor sucks up the extra energy but the same amount of charge (ie the current) gets to the leds. With the 3.3v circuit less energy is take off in the lower resistor.

 

jeff

 

Thanks Jeff. I like that description!

[cteno4]

1210 is just wider footprint than 1206, both are the same length. Second pair of numbers represents the width, first pair the length between contacts. Height can vary (ie the higher capacity caps will be taller with the same smd number).

 

handy guide

 

http://www.topline.tv/SizeChart.html

 

jeff

[cteno4]

44 minutes ago, gavino200 said:

?

 

Gavin I thought you had flicker on your current layout that you were able to negate with one of your led strip tests and a cap. Of did I dream that. 

 

This is the challenge to know how far to go on the caps!

 

jeff

[gavino200]

10 minutes ago, cteno4 said:

 

Gavin I thought you had flicker on your current layout that you were able to negate with one of your led strip tests and a cap. Of did I dream that. 

 

This is the challenge to know how far to go on the caps!

 

jeff

 

This is true. It's not 100% perfect but it's better than I've achieved with anything else. Definitely a "proof of concept".  That was with the rectifier-resistor-cap setup so it's anyone guess what that translates to with the improved arrangement, rectifier-cap-resistor.

 

Electrically my track is actually pretty good. I have a feeling that most of the flicker comes from the crappy brass strip pickup contact as the train trundles along. I plan on switching to soldering directly to the brass strips.

 

What I'd really like is a cap that's capable of buffering the train going over insulated joiners, if I wan't to play with block detection at some stage. The isolated joiner flicker is quite significant.

Edited March 12, 2019 by gavino200

[cteno4]

11 minutes ago, gavino200 said:

Thanks Jeff. I like that description!

 

Its good to think in these terms to think thru what happens in each component of a circuit as it then helps to understand the whole.

 

the old P = I x E then makes sense as power is how many electrons times the energy of each.

 

jeff

[gavino200]

I think I'll also solder all my tracks on my next layout. My goal is 100% perfect interior lighting. 

[cteno4]

3 minutes ago, gavino200 said:

What I'd really like is a cap that's capable of buffering the train going over insulated joiners, if I wan't to play with block detection at some stage. The isolated joiner flicker is quite significant.

 

Ahh then it sounds like a few joints with insulation joiners with each section wired may be the torture track to test on. Good as reproducible. Also use the flaky contacts to make the worst case scenario.

 

jeff

[gavino200]

1 minute ago, cteno4 said:

  

Ahh then it sounds like a few joints with insulation joiners with each section wired may be the torture track to test on. Good as reproducible. Also use the flaky contacts to make the worst case scenario.

 

jeff

 

Great idea. Actually I'm about to dismantle my layout. I could set up a temporary track just like that to experiment with.

[Ochanomizu]

22 hours ago, gavino200 said:

 

The smallest value that I've ordered is 100uF. But you have a point. It's worth testing lower values. Especially with a voltage regulator dropping the voltage to 3.3V we should be able to make do with less capacitance. I think I'll order a few.

 

Hello gavino200,

 

I look forward to seeing the results of a smaller cap.

[chadbag]

Here is the latest schematic, which added a few more LEDs to the circuit and layed it out a bit differently.

 

Also, the first PCB layout of the actual circuit is done, except for the power pickup pads (the track input solder pads), and the physical edge of the PCB s not defined.  And there should probably be more text and drawing on it for things.

 

The cap pads, Resistor pads, and LED pads are sized for both 0805 and 1206 sized components, except C3, which is the first capacitor which is sized for 0805 up to D-case Tantalum.

 

The voltage regulator has 3 sets of pads and can work with SOT-89, SOT-223, or TO-220 sized.  The big ones that are 1A or 1.5A with the metal legs are the TO-220.  The small ones I bought that are 0.15A are the SOT-89.  Another one I bought and which seems to use a common package is SOT-223.   This will accomodate any of those.  And in any voltage you want.

 

The LEDs are about 10.8mm apart except for the couple interspersed in the other components on the very left, where the spacing is more as-fits with that approx distance.  The two left most LEDs are basically center line centered while the others are staggered up and down a little.  DOn't know if that is helpful or not and I may make another version with them straight.  The idea was to get better light coverage.

 

I need to figure out how to make the solder pads for the track voltage wires (actually two sets -- one for track voltage for DC/DCC no decoder hooked to the AC inputs on the rectifier bridge, and 2 pads for DCC decoder outputs (which are more DC like) which are on the other side of the rectifier.  I also need to figure out how to define the PCB board physical edges.

 

The board is approximately 6.2 inches long (15.6 cm).  The idea is to fit Euro sized passenger cars etc. but allow the user to saw off individual LEDs of the right edge to make the board shorter.  We can also make a variant PCB that is shorter by a couple LEDs as well.

 

Schematic-v0.1.5.pdf

PCB-v0.1.5.pdf

[chadbag]

I've gotten the edges defined and even produced production files as a test and uploaded them to get an automated quote.  100 pieces comes in at 40 - 50 cents each.  (10 pieces is like $2.60 each). 

 

I was looking at a Fleischmann wagon and I think it is slightly too long. I am going to tighten up the LED spacing and make a few other tweaks.

 

[gavino200]

Wow!

[chadbag]

OK, here is the PCB layout.  I still need the solder pads for the power inputs.

It is about 5.89" long (from about 6.2") and fits with about a cm or slightly more when placed against the side of my test Fleischmann wagon.

 

The widest point is about 14.7mm (down slightly).  If I cut the border down on the wide part it could be made about 1mm less wide.   It should fit in a Euro wagon but be tight (about 15.875 mm).  A little more space space in a Japanese wagon (approx 16.66 mm).  [All wagon measurements based of single wagon examples].   The skinny part is about 10mm wide.  We could cut it down if we make the LEDs in a straight line.

 

PCB-v0.1.5-R2.pdf

 

 

 

Edited March 13, 2019 by chadbag
still

[cteno4]

Nice job chad!

 

jeff

[chadbag]

OK, pending a review to make sure it is hooked up correctly, and if no-one sees anything grossly wrong with this one. this is the one I am going to get some samples made from.

 

The very end with the solder pads and the rectifier is approx 14.35 -- 14.4mm wide.  The next section is about 11.05 mm wide, while the bulk of it is about 10 mm wide.  Overall length is about 149.6 mm or about 5.89 inches long.  If you cut all the LEDs off after the "cut marks" it is 100mm long and if you saw it up after the first LED after the last capacitor it is about 78 or 79mm long.   I chose a spacing of 1cm between most of the LEDs for better coverage (and because the TORM and Popondetta ones seem to have them spaced closer together).

 

There are spots for 14 LEDs, either 0805 or 1206 size.  The last 5 LEDs can be cut off at any point between them to make a shorter board.  I spaced this to hopefully fit in European passenger wagons.  Other wagons, including Japanese Shinkansen and EMU may need shorter ones.

 

You can use 0805 or 1206 caps and resistors as well.  And you can use one larger SMD cap if you want (up to case size D tantalum).  You can use it with or without a rectifier bridge.  There are solder pads for both choices.  You would use it without if you were feeding it from a DCC decoder (which converts it to a DC like power for you).  Otherwise, for normal DC or DCC track power, you would use it with a rectifier.  Any of suitable specs that is sized like the MB6S / MB6F will fit.  I'll label the pins on the board.   There are spaces for the "smoothing" caps before and after the voltage regulator spot, like is recommended for a voltage regulator.   You can leave them off if you want (at your own peril 🙂  ) without modification.  There is a current limiter resistor feeding into the caps (like 100 ohm or so).  If you leave this off then you need to bridge that position.  There is also the "dim/bright" resistor spot after the caps and before the LEDs.

 

There is space for up to 6 0805 or 1206 caps for the anti-flicker circuit.  The large tantalum cap location also supports the smaller sizes.

 

For track power you can solder to the two large pads on the back, or you can attach two 2.54mm spaced pins (from those breakable pin connectors) to the spots next to the top pad and hook up track wires to the other half of the connector so you can easily remove the shell from the body and break the connection without issue.  

 

You can use 3 different voltage regulator sizes as I mentioned earlier.  The circuit works with a voltage regulator in any of the positions.  Just make sure the pins are in the right order.  If you want to run straight 12V to the LEDs and use a larger resistor you should be able to by jumping to top most and bottom most pin positions on the voltage regulator position and use appropriate sized caps (25V or better).  

 

I'll probably get 100 made as a sample run (since the price between 10 and 100 is like $20 delivered or something -- 10 pieces is like $2.60 each and 100 samples are like 40 - 50 cents each, delivered).  So those who want to get some to play with can just cover whatever the cost is for me to get them plus postage to your place.  Once we like them we can do bigger runs (they will get used up fast since most trains are like 4 - 6 - 8 -10 or so cars or more and we all have arge fleets 🙂 and 1000 is like 25-30 cents each IIRC  ).

 

 

I won't claim this board layout would win any awards.  Rather ugly and inefficient I am guessing, but it should work and the parts are placed where they will work and where they made sense to ME.  This is the first time I have done this.

 

Please review the circuit schematic (in one of the replies above) and the PCB layout here:   (and let me know if you see any issues)

 

PCB-v0.1.5-R3.pdf

 

 

[chadbag]

One question for the team:   What thickness of board should we use?   The Ersatz-EM13 decoder is 0.6mm.  It is not bad but I thought maybe 0.8mm?  (Traditional boards in electronics seem to be about 1.6mm or so based on what I see to buy online for test boards).