Lighting Trains

[cteno4]

Usually most components have a manufacture code, polarity (if needed and room) and, if room, a the main rating on it.

 

so manufacturer codes have some cross reference to standard component sizes. Some encode some of an old standard part number done by some biggie in the industry. There is no standard on all this. It can be hard as so many manufacturers copy the same component and sometimes just label it the same with a company mark or at times none just a clone or at times part of the original code with an addition that has nothing to do with a change in specifications.

 

yes it’s way overkill but diodes can be made pretty robust pretty small so you get a Big Bang for the buck here. No need to go smaller in value, I think 300v maybe the smallest I’ve seen.

 

i think you will find a dip 100uf tantalium being bulkier than a similar smd. You also need to take care in bending the leads near the base of some dip caps as they can crack the insulation dip. These days I think it’s usually an epoxy dip so not a problem but in the past I’ve cracked some bending them over when jclose soldered on a board. A 25v 100uf type D tantalum cap is 7.3 x 4.3 x 2.4mm and a lot cheaper than dips.

 

Kiha is spot on that using the dimmer resistor will limit the in rush, so that gets rid of doing a resistor for that. You would only maybe need a resistor in series with the cap if you put the cap across the rectifier output. I’m pretty sure your inrush ona 100mf will be pretty small and random and not matter.

 

if it works well on the flicker with the cap on the led leads and doesn’t change brightness and lasts long enough to get rid of Flicker the should be good. Just solder an smd resistor across the end of the led strip contacts (and a second on the center contacts for extra if needed as kiha notes). Then on other end just resistor between rectifier output and led strip on one side and just a wire on the other.

 

Curiois if you can see it slow the dim when you pull power from it. You could run a test on a led module and dimmer resistor with and without a cap and turn off the power simultaneously to them both. If you can see a big difference you could the. Test where the cap is placed before or after the dimmer resistor.

 

my bet is if the cap is upstream of the dimmer resistor you will get longer dimmer after power off and down stream will get shorter brighter after power off. Upstream means each led strip is only drawing 1.4ma from the cap discharge and down stream 15-20ma (don’t know what value current limiter for the 3 leds they have on the led strip).

 

even at 100uf I think you will see the extra on time with the cap as you are drawing low amps here.

 

but if you are planning on investing the time to make 1000 then I would suggest you spend a little $ on an ee to look at the circuit to make sure it’s the best design for your needs and nothing overlooked. Don’t want a headslap 500 in...

 

cheers

 

jeff

[gavino200]

Thanks, everyone for sharing your knowledge and expertise. I'm going to make up a few test strips wired in all the ways we discussed to see how they work in practice. One wild card is the brightness response of the LEDs themselves in response to varying current/voltage and our perception of it. Having played with LEDs and potentiometers I've noted that you can adjust the pot for huge amounts with no perceptible difference in brightness and then have a big change at a certain inflection point. 

 

Please feel free to contribute more on the theory side if you notice something. I can always make up another test strip.

[gavino200]

10 hours ago, cteno4 said:

 

but if you are planning on investing the time to make 1000 then I would suggest you spend a little $ on an ee to look at the circuit to make sure it’s the best design for your needs and nothing overlooked. Don’t want a headslap 500 in...

 

 

I'm not going to do these at once. I'm just thinking of an average ten car train, and assuming I'll probably buy a hundred more trains in my lifetime.

[cteno4]

Doesn’t matter if slow or fast on the production, it’s the amount. Don’t want to discover something after having done a bunch. I’ll see if I can mock this up here on breadboards this week.

 

Yep you are right, above like 5ma they are just blazing suns you don’t see any difference in directly but may if in a enclosed space where you are seeing reflected light spread out. But at a point source it just maxes out our receptors so you don’t notice any change. I’m thinking at the dimmer slower discharge you won’t see a difference but may be able to make the pulse longer to cover flicker better and maybe even reduce the cap size. But if it works fine with a reasonable cap right on the led strip then you’re done! But the inrush charging current w.o an incoming resistor may reduce the life of the caps some, not sure there.

 

the ultimate antiflicker light boards had a not so simple circuit on them (not crazy, but more than this simple one) with 2 super caps (I think 1mf), somethere are more things that can be done, but. But it may have been total overkill and those cost like $18 each! I was never able to trace it well enough innthe Picts to decipher all of what was going on. I thought of getting one to see, but realized recreating w.o a pcb for the circuit would be a challenge and recreating a new one simpler from it was probably beyond my ee skill set! 

 

also stumbled on a paper that postulated the latest tanilum cap chemistry may not need to be downgraded as much on voltage, but manufacturers still don’t say much on this and it would be hard to know which ones have the improved chemistry anyway so better safe and do at least 50%. I know this increases cost and size at times but last thing you want are these popping on you down the road both to fix and potential car damage. I’ve never popped a tantalium but I have popped electrolytics and they do go pop and one cracked a plastic project case! You pop a cap as once you have heard one blow the distinctive pop will stay with you!

 

jeff

[chadbag]

I read that the tantalum ones shoot flame and fire and brimstone out while the electrolytic ones just pop and release the pressue.

 

[Kiha66]

I've had electrolytic ones fail and they shot plenty of flame out the end too.  Iirc it was for a rc boat lipo regulating circuit that had gotten water in the battery.

[cteno4]

The electrolytics do shoot flame and melt things! They are designed I think to blow their tops but I did have one rupture out from the side and brown a bunch of stuff beside it.

 

tantaliums are energy dense so it sounds like they are more of a intense meltdown as they fry closed usually! But I’ve never popped a tantalium as I’ve not used them for big storage like this on past hobby projects usually electrolytics as they are cheap and if you have space you can load up big ones! 

 

Maybe i should try over volting one to see! But the shorting when fried is not great so I guess fast fuse it and use a cheap power source! 

 

Cheers,

 

jeff

 

after posting this I thought someone must have YouTubed this and of course yes! Definitely why you want to have a good safety derating and solid circuit in your train car lighting design!

 

 

[gavino200]

I'm going to lay out the equipment tonight. Probably I'll wire everything up tomorrow or Friday. 

 

This is the setup I have in mind.

12V power supply. 

Multiple strips wired up to a junction connected to the power supply

Each strip is two units/6LEDs

Each strip will have an identical rectifier.

At the moment I only have 220uF Tantalum 40V Capacitors. All "capacitors" mentioned will be of this type.

All before/after designations are with respect to the power supply. ie the series "starts" at the power supply and "ends" at the LED strips.

The power will be disconnected at the power supply junction, so I'll be able to compare different circuit arrays directly/simultaneously.

 

The test strips will be the following. Feel free to suggest additions.

 

 

1. Rectifier and resistor alone with no caps.

2. Resistor after rectifier but before capacitor.

3. Resistor before rectifier, which in turn is before the capacitor (may have an assembly advantage)

4. Resistor after capacitor but before LEDs

 

5. Resistor after rectifier but before LEDs. Capacitors soldered to the two empty solder pads downstream.

 

If number 5. is successful then the following will be tested.

6. Same as 5 but with resistors in series with both capacitors.

7. Same as 5/6 but with resistor before rectifier (possible assembly/manufacturing advantage)

8. More than one capacitor side by side at the empty downstream solder pads (to experiment with greater capacitance - 660, 880uF)

 

[gavino200]

WANTED: Someone to ELI5 (Explain Like I'm Five) why I still need a 25V capacitor if I've already dropped the voltage with a resistor. Why don't I need to just follow the rule of thumb of using a capacitor with twice the likely voltage - using the voltage "left" AFTER the limiting/dimming resistor. I think the resistor is very unlikely to fail.

[Kiha66]

Just now, gavino200 said:

WANTED: Someone to ELI5 (Explain Like I'm Five) why I still need a 25V capacitor if I've already dropped the voltage with a resistor. 

 

The issue is the resistor acts as a restrictive device rather than a voltage regulator.  So it only lowers the voltage while there is current flow across it, if power is not being drained on the other side the voltage will eventually rise to that of the source before the resistor.  Its like a hose with a kink it it, the water might only slowly dribble out the end but if you put your finger over the end you'll soon feel the full pressure once it builds up.

[Kiha66]

Because the LEDs should always be draining that power out you don't need 25V, but its there on the off chance that if the worst happens (resistor gets shorted, LEDs burn out and stop draining power, lighting board gets shorted to track power, DCC station is accidentally set to HO/G gauge, ect).  Its a safety factor that means you're less likely to have a cap burn out and melt a possibly expensive model.  If you want to save money and can take the possibility of a car melting then you can certainly get away with 16v, but that would be something you'd decide yourself and not something most people feel comfortable suggesting to people in case the worst happens and someones car gets melted.

[gavino200]

30 minutes ago, Kiha66 said:

 

The issue is the resistor acts as a restrictive device rather than a voltage regulator.  So it only lowers the voltage while there is current flow across it, if power is not being drained on the other side the voltage will eventually rise to that of the source before the resistor.  Its like a hose with a kink it it, the water might only slowly dribble out the end but if you put your finger over the end you'll soon feel the full pressure once it builds up.

 

Thanks. This works for me!

[gavino200]

Surprised to find "zener" diodes are tiny and cheap. But I think I'm convinced about not going lower than 25V 

 

https://www.ebay.com/itm/Chanzon-30-Values-300pcs-1W-3V-47V-Zener-Diode-Assorted-Kit-1Watt-Assortment-Set/273318297950?hash=item3fa30a415e:rk:4:pf:1&checksum=2733182979500a2ddc8ee3c54e359fc8c7600a5ffbf6&enc=AQADAAADAFjVrDbVsZ8oH%2F8PNHtt9VX4%2Fw7FZcmMuqsX8uaFEduVRf281%2FTa%2Fsi%2BzqjQNgTimLEdKPfpozq50OAkupYUvPZNvfCl5GpDU%2BqLS17RqVPfI43nD58FlOhVHHQo7l5wfgfuvrJphQvVnJQ5Vt4GK%2BDdSyGzhNTv2QrVCAWMDwEpEzYFsVmNSe3nHLmsoXCfgaeu%2BHcZew8s9cOLgP3C7h3E8xiIQ5YFWsFU2J8pXP6pvS%2B8gp%2FiALPfUMT%2F%2FH29UUzxybZj%2B3eAb%2BmEUtHkupsB2BJKsvdSE6cgn%2FY242IqAEQVaReS7OjX7ghleNqjaYIMrxANPQpf%2FBLucHKZcBRzHpevKyUGrUvZ0%2F5B43o%2FaistOs5v4TBVw5cwUE3x162h1QWjjyDR0KnloKmwtRWOEbRm%2BaqAGg0LvsQXOIeX18ooM4tYOe64VGKaTtC45fGlC3R6YHQQfRMPgZKrZpJ5iFgbOCAEeRl4aWjjUavAHOi3IJY%2BRMkA%2FV%2FpLKJk25j0ub63tVnhYspP02pyUe4ppPt1aEMIVtMVH4EcOwp6z1MYZ90SZDf99gH3xeN9%2Fx3QAukFR6J4fnJ3ZFo4RoJY4XXrkuhHc%2FL7pZhDXPmoQHXbIK26b8BbmK4%2FNxG4WRNd%2BKLniwh2NKpLTuKYRdXqsy%2FCLGVkmjqFuQj3IfrZhQcpU48jPHmJhO6%2FqtKH%2FbR4pfgxK0RQ6DXGNmRmX%2FC1%2FI4T5H4czFVvUKGF81RwWM%2Bv5f7p6iMeeDiymIYErNPPiBBHw1oXUEy128gABD0z8wsymZfsghau7y%2FDcM1HrCuB9z2UIBccR8BCuuJfQlbMfV5IpCOMffN0dKbpeZmhrqfChSWZgKUcz4f7LKDXkcbbJwnBBcEenHs0DI%2B1ra2Y%2FL33dOU5dl6BnpxmufyPCqEPTs9EUVlYerNJtBCX5mZfb6UH%2BoX%2Bww7pZnweNO8Ee606HKxV5tQLLtHuF0ZOIeZ1STjscK8KdawL1nCRVK2vEKvtgv0qXJiGtP7mxg%3D%3D

[gavino200]

2 hours ago, Kiha66 said:

 

The issue is the resistor acts as a restrictive device rather than a voltage regulator.  So it only lowers the voltage while there is current flow across it, if power is not being drained on the other side the voltage will eventually rise to that of the source before the resistor.  Its like a hose with a kink it it, the water might only slowly dribble out the end but if you put your finger over the end you'll soon feel the full pressure once it builds up.

 

This comes down to V=IR, right? The voltage of the circuit is governed by the battery/power source/terminals of the rectifier. Higher resistance just lowers the current but doesn't affect the voltage. Voltage is like the water pressure at the tap in your example.

[Sheffie]

One other suggestion: add some sheets of paper of varying thickness to diffuse / lower the intensity of the LEDs. This should help you judge their brightness.

[cteno4]

Yes voltage is like water pressure and current is the size of the pipe.

 

Resistors don’t effect the full voltage the caps will see for the circuit as kiha points out, so the cap will see 12v. I would stick to the 25v to be safe. The point of the rating is that it can stand that voltage, but it’s at the limits and at which may break down. Thus the rule of thumb derating of 50% to be safe. In this application being safe is good as a failure will result in a melted car.

 

Not sure if #3 will work properly, the resistor I think should be on the dc side of the rectifier.

 

jeff

[gavino200]

9 minutes ago, Sheffie said:

One other suggestion: add some sheets of paper of varying thickness to diffuse / lower the intensity of the LEDs. This should help you judge their brightness.

 

Interesting thought. Yes, it's hard to be objective while being being blinded by 48 little suns. 

 

However I'll probably just use one layer of thin paper if anything. When I used TORM lights I turned down the brightness with sheets of paper. In this case I'm controlling brightness with resistors. The paper is purely a diffuser rather than a brighness reducer. I actually want it to be thin and not decrease brightness much. So I'm not messing around with paper layers or thickness anymore.

 

As for the retinal damage, I guess I'll just take one for the team.

[gavino200]

11 minutes ago, cteno4 said:

 

Resistors don’t effect the full voltage the caps will see for the circuit as kiha points out, so the cap will see 12v. I would stick to the 25v to be safe. The point of the rating is that it can stand that voltage, but it’s at the limits and at which may break down. Thus the rule of thumb derating of 50% to be safe. In this application being safe is good as a failure will result in a melted car.

 

I'm interested in popping one deliberately just to see at what voltage it burns. Just out of curiosity. I do realize that there would be variance among individual caps, like any other mass produced item. I also think it would be fun. But currently I don't own a variable output transformer that goes that high. Pity. 

 

11 minutes ago, cteno4 said:

 

Not sure if #3 will work properly, the resistor I think should be on the dc side of the rectifier.

 

 

 

Dani puts his resistor in before the rectifier and it works fine. It seems like it should. Resistors don't have polarity so they should be able to limit AC. No?

 

 

[cteno4]

It’s not polarity if the resistor. It’s the fact that with it on the ac side it will have your current limiter jumping back and forth to either side of the circuit. I think it should work, just not sure if any other gotcha there. At times circuits will work but there may be better best practices. Above my ee grade to know.

 

jeff

[gavino200]

4 minutes ago, cteno4 said:

It’s not polarity if the resistor. It’s the fact that with it on the ac side it will have your current limiter jumping back and forth to either side of the circuit. I think it should work, just not sure if any other gotcha there. At times circuits will work but there may be better best practices. Above my ee grade to know.

 

jeff

 

I put no weight on my opinion. My only knowledge comes from faded high school physics and a bit of tinkering. But I think you can put a resistor at any point on the main circuit and have the same result. 

Edited February 27, 2019 by gavino200

[Sheffie]

I'm pretty sure you can put a resistor on any connection of a rectifier and get the same result. All it's doing is directing the current from the AC inputs to the DC outputs. A resistor on any one of those limbs is going to limit the current by the same amount.

[cteno4]

Yes it should work, there was just something nagging me about it, but it’s just old neurons. Sorry

 

jeff

[Kiha66]

3 hours ago, gavino200 said:

This comes down to V=IR, right? The voltage of the circuit is governed by the battery/power source/terminals of the rectifier. Higher resistance just lowers the current but doesn't affect the voltage. Voltage is like the water pressure at the tap in your example.

 

Actually in a roundabout way.  The thing is V=IR only applies when there is current flow (the I).  If there is no flow the voltage difference (V) across the resistor goes to 0 and the voltage potential to ground becomes the same throughout the circuit.  Sorry for my badly remembered explanations, who'da thought I'd actually have to use what I learned back in school!

[gavino200]

I'm  curious, what would be needed to actually step down the voltage? Is there a small component that could do that? 

 

I wonder if that's what Popondetta are doing. There LED stick seems to use 15 very small ceramic smd capacitors. They have to be low voltage caps.

 

There's a variable resistor switch to change from high to low "brightness", and a six pin component that looks like a microprocessor/chip. There are also two small black components immediately after the pick up leads. Any guesses what's going on there?

 

You can see it quite clearly at 0.07 on the video below.

 

https://popondetta.com/maker/?p=242

Edited February 27, 2019 by gavino200

[cteno4]

Im guessing the chip is most likely a voltage regulator that is dropping the voltage and it’s running the leds in parallel. If at a voltage near the led voltage then ceramic cap could be lower voltage and cheaper. Ganging up a bunch of small ceramic caps in parallel is smaller and safer (they are not as dramatic in failure) than bigger tantalium caps. I was looking at voltage regulator circuits for shorties like your project but put that aside to play with the current regulator circuit.

 

jeff