Yeah, I think you lucked into a working circuit. The idea that the 1K resistor with an LED drops 12V to 3.1V is not true, the 1K resistor limits the current through the LED to a safe value (quite safe, actually, less than half the typical maximum, which is why it's a good choice. That and most LEDs these days are way too bright, reducing the current reduces the brightness a bit). But, because directily across the cupercab is the LED and the 100 ohm resistor, what yoou get is the LED running at near maximum current - 20ma. That same 20ma goes through the 100 ohm resistor, resulting in a voltage drop of 2 volts (.020 x 100). 2 volts in the resistor plus 3.1 volts in the LED is - tada - 5.1 volts, the voltage across the capacitor can't go any higher than that because it's in parallel with the LED/resistor, and two devices in parallel get the same voltage. The only problem is that should the LED ever fail, the voltage across the capacitor is going to swing almost all the way to the full input voltage (the 1K resistor won't limit much voltage, but it will limit the current surge into the capacitor) and 12-15 volts across a 5.1 volt capacitor is very bad news.
So as long as all the components are connected and working, the circuit works as it should. But a single component failure can lead to a big mess - a cap that size can certainly blow apart an HO scale passenger car or caboose. If you want to see what can happen and have a spare cap, make sure you use a LONG wire and stay well back and use face protection, as bits of the aluminum can fly off like shrapnel. Better bet would be to put some bricks or cinder blocks between you and the cap, and then after it blows up go look at the mess it made. There is a LOT of energy stores in a supercapacitor, even though it's only 5 volts compared to say a 470uF 35 volt electrolytic, 1F is 1 MILLION uF. So even though the 470uF cap has 7 times the voltage, the supercap is 2500x the capacitance. The energy in a cap is voltage x charge (not capacitance), in the case of a 5V 1F capacitor that's 12.5 Joules, or 12.5 W/seconds - it can supply 12.5 watts for 1 second, or lesser energy for a longer time - that's why the supercap keeps the LED on so long, or why a supercap keep alive can keep a loco moving for several seconds. By comparison, a 470uf cap charged to a DCC voltage of 15 volts has a mere .053 joules of energy. A loco that draws .2 amps uses 3 watts at 15 volts. That puny cap can only keep it going .018 seconds. 3x 1F supercaps in series to get 15V will have a capacitance of .33F. At 15V, this is 37W/S, so that 3 watt loco will run for over 12 seconds.
It's too early to do this much math. Probably off somewhere. But a supercap is incredibly huge compared to ordinary value capacitors.
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
Randy - to be perfectly honest, I designed my circuit probably 20 years ago before I knew much of anything about electronics. I guess it's one of those cases where it works because I didn't know it shouldn't !
Here's the dumb logic I used .... I was using 1000 ohm resistors to power 3 volt LEDs, so the same resistor with "my math" would only allow 3 volts to the capacitor. Knowing what I know now, I probably would think different.
Again, mathematically, I figured the cap was charging to 3 volts, so a 100 ohm resistor would be sufficient to slow the discharge to the LEDs to keep them lit AND protect the LED current rating at the same time.
Today, I too would shake my head in disbelief at that logic.
Oddly enough it works. To this day I can't tell you how or why. But I can tell you I've had a number of them performing as they should for a good many years. I don't even have a switch on them either - whenever the layout is powered up, they are on (except for staging tracks where the track power itself is turned off).
Mark.
¡ uʍop ǝpısdn sı ǝɹnʇɐuƃıs ʎɯ 'dlǝɥ
Lots of great ideas! Thanks, everyone!
I like the idea of wiring the LEDs in series to eliminate the need for resistors (fewer resistors would mean less electricity "thrown away") but the LEDs are mounted in permanent lighting strips, so that's not an option. So I'll use something that can be spliced into the wires from the bogies to the lighting strips. The cars have interior details, so not much room for batteries, and I don't want to do anything really destructive to them. Maybe I could mount a battery or large capacitor under the car and paint it to disguise it as an air tank or something.
It'll take me a while to look at everyone's ideas more closely, but I just wanted to say a big THANK YOU to everyone!
gregc Mark R. which enable me to use a 5 volt 1 farad super capacitor which is quite small physically so you charge the cap thru a diode and 1k resistor? you don't need to use something to limit the voltage across the 5V supercap?
Mark R. which enable me to use a 5 volt 1 farad super capacitor which is quite small physically
so you charge the cap thru a diode and 1k resistor?
you don't need to use something to limit the voltage across the 5V supercap?
I am trying to figure out how that could possibly work - at some point as the cap charges up the voltage across it should exceed 5V, 1K resistor or not - Although there is the 100 ohm resistor and 3.1V LED across it as well. So that may be where the limiting is occurring, to little more than 3.1V. Personally I'd feel much better about a slightly more complex cirtuit to guaranteed no matter what the cap is never overvolted. I'm assuming those that added a switch to turn it off put the switch on the 1K resistor side? If you open just the LED side, the cap WILL hit greater than 5V and go boom.
gregc I think the schematic posted by gmpullman, without a bridge, is fine except that the LEDs can be in series, drawing a 1/3 of the current and increasing the decay period of the capacitor by a factor of 3. i recently wired passenger cars with 2 LEDs in series with 3k resistance to reduce the brightness after initially using a 1k resistor. I had the flicker problem, but connecting the 3 cars together helped a lot. I would think adding the capacitor to on car would fix it. Not sure I want to illuminate the interior of a passenger car to show the capacitor in it. richg1998 I just Googled car lighting keep alive. Here is one. There are more out there. http://www.dccguy.com/?p=3947 i understand the need for a full-wave bridge on a DC layout, but not a DCC layout. Also don't understand the need for the zener and like gmpullman's circuit, the LEDs can be in series along with the resistor bench test first.
I think the schematic posted by gmpullman, without a bridge, is fine except that the LEDs can be in series, drawing a 1/3 of the current and increasing the decay period of the capacitor by a factor of 3.
i recently wired passenger cars with 2 LEDs in series with 3k resistance to reduce the brightness after initially using a 1k resistor. I had the flicker problem, but connecting the 3 cars together helped a lot. I would think adding the capacitor to on car would fix it. Not sure I want to illuminate the interior of a passenger car to show the capacitor in it.
richg1998 I just Googled car lighting keep alive. Here is one. There are more out there. http://www.dccguy.com/?p=3947
http://www.dccguy.com/?p=3947
i understand the need for a full-wave bridge on a DC layout, but not a DCC layout. Also don't understand the need for the zener and like gmpullman's circuit, the LEDs can be in series along with the resistor
bench test first.
The linked circuit uses a 5.5V supercapacitor, the zener is CRITICAL to keep the voltage from exceeding 5.5V or else the dynamite stored in the express operator's safe will go boom and make a big mess.
Since it's designed to work with DC< it will start charging at anything over the LED forward voltage, on up to full throttle, and it it has enough time to charge to the full 5.5 volts, the cap should keep a couple of LEDs lit for a nice long time during station stops. Overkill for DCC, since all you really need is enough capacitance to keep the lights from flickering over dirty spots.
Bernie, there are lots of good ideas above.
Here is what I do
I bought these DC/DC converters from China on ebay.
Mini 3A DC-DC Converter Adjustable Step down Power Supply Module LM2596s (the LM2596 is important, don't buy ones without this number)
very cheap, efficient and small. You need to adjust the output voltage with a screwdriver and you also need to buy bridge rectifiers and capacitors like in the circuits above. I bought 470UF/35V SMD Electrolytic Capacitor 12*13mm and DB104S SMD-4 DB104 1.0AMP Single Phase Bridge Rectifiers 400V 1A ,Diodes. I get two leds to stay lit for more than 30 seconds before I notice any dimming.
The advantage of this over the simpler circuits is that the DC/DC converter extracts all the energy out of the Cap before giving up, so there is no fade until the end. It can provide a 3Volt output for as long as the main cap is more than 3 volts.
It's only a bit more complicated than the other designs but gives a longer full brighness time.
Alan Jones in Sunny Queensland (Oz)
Mark R.which enable me to use a 5 volt 1 farad super capacitor which is quite small physically
greg - Philadelphia & Reading / Reading
Mark R.My two resistor values are reversed (1000 on the left and 100 on the right) which enable me to use a 5 volt 1 farad super capacitor which is quite small physically.
Mark, thanks for your "revisions". I have a bunch of 5V .1f supercaps. I was afraid to put DCC voltage (13.6V in my case) across a 5 volt cap. I thought I had to use 3 or 4 in series.
I'm interested in trying out your circuit. Just wanted to be sure of the 5 volt cap.
Most of my installations have been in 8 or 12 wheel Walthers passenger cars. Even with five seconds of "capacity" the flickering is completely eliminated. A longer stay alive time would be nice, but not entirely necessary.
gregcNot sure I want to illuminate the interior of a passenger car to show the capacitor in it.
Greg, In most of my passenger cars there is usually someplace where I can hide the capacitor, usually in the toilet or vestibule area.
IMG_9995 by Edmund, on Flickr
Thank You,
Ed
Mark's system works quite well. I have it installed in eight cabooses. Never flickers. The only addition I added was a latching reed switch so I can shut the lights off if the caboose is parked on a live track. I believe you can use a higher voltage capacitor if you wish in Mark's circuit.
Dave
I'm just a dude with a bad back having a lot of fun with model trains, and finally building a layout!
I use the same circuit basically as gmpullman, but ....
My two resistor values are reversed (1000 on the left and 100 on the right) which enable me to use a 5 volt 1 farad super capacitor which is quite small physically. In my cabooses which are lit with two LEDs, when removed from track power, they will stay lit for nearly three minutes !
richg1998I just Googled car lighting keep alive. Here is one. There are more out there. http://www.dccguy.com/?p=3947
I just Googled car lighting keep alive. Here is one. There are more out there.
Rich
If you ever fall over in public, pick yourself up and say “sorry it’s been a while since I inhabited a body.” And just walk away.
Hi Bernie,
I have been using this circuit on my lighted passenger cars for about two years now.
DCC_light2 by Edmund, on Flickr
I'll find the thread that has more information and update this thread with a link to it.
I show photos of a typical installation:
http://cs.trains.com/mrr/f/744/t/267418.aspx
Good Luck, Ed
Although I have a DCC system, this issue doesn't involve any of my locos or decoders. My locos all run fine!
But I have a few passenger coaches that draw power directly from the track (no decoder involved) to power the lights in which the light occasionally flickers. Not a big deal, but I'm wondering if there's a simple schematic available online somewhere that uses either a capacitor or rechargeable battery that I can solder together quickly. Or, better yet, buy pre-made and be done with it.
I've found some old threads from 2011 that talk about Keep Alive circuits for DCC locos, but I know some people get really pissed if you resurrect dead threads, so I'm starting a new one.