For some time I have considered incorporating LED lighting in my layout. I have seen after market offerings for LEDs operating between 5 -19 VDC, prewired universally for AC or DC power sources, (http://www.modeltrainsoftware.com/bl-212.html). The circuitry includes bridge rectifier, capacitor, resistor and lamp and I have been tempted to make a purchase. However, I would rather learn something in the process as opposed to simply hooking up a finished product. I am no electronics or electrical craftsman but can reason and follow instructions. I have 50 ea. 3.3 volt 0.066 watt LED lamps from an extra Christmas string rated at 3.2 watts and 0.027 amps that I would love to play with. Is there a generic schematic that shows the correct placement of the various components necessary to wire an LED using variable AC and how do I determine the proper specifications for the components?
Hej, Swede. This is a more complicated problem than you may realize. If you can find it, an article I wrote for the September, 2008, CTT may get you started.
Wayne, I’m Swedish on my father’s side. My immigrant great-grandfather, one Anders Svenson, contracted his patronym to a middle initial, adopted his father’s surname, “Nilsson”, and anglicized the whole thing, to become Andrew S. Nelson. My Swedish is very rudimentary but improving–my son lives in Sweden with his Swedish sambo.
God mórgon Bob. I thought you were Texan. That is wonderful that you can converse in Swedish. My ancestors were from Czechoslovakia, but only my grandparents could speak Czech. I wouldn’t even recognize it if I heard it.
God morgon till er båda, Bob och Wayne. My parents were born in Sweden but met in the US and married in the 20’s. Except when other Swedes were visiting, or my parents didn’t want us kids to know what was said, English was the only language used in the household. My siblings and I picked up words and phrases but none of us, sadly, are proficient in the language.
Wayne, thank you for attaching the circuit diagram. I have ordered the reprint suggested by Bob and will study it before I commit to an approach or, probably, different approaches depending on the lighting requirements I want to satisfy (car lighting versus static lighting, single lamp versus string).
I have used these to re-light two of my flyer locomotives and they have been great. Haven’t used them on any passenger cars or other components though.
Hi Bob, I received, reviewed your article in the 2008 CTT September issue, and have wired several successful LED circuits using a DC wall wort with resistor as well as an AC train transformer, with resistor. I’m sure you or others have also wired AC powered LED circuits that included a bridge rectifier and capacitor. There was no mention of these components in your article and I was hoping there would be. Have you seen any good postings referencing their use and have you seen any good posting outlining a flashing circuit for LED use?
not sure what may have been said in I assume swedish
leds, like all diodes are current devices and operate over a very narrow range of voltages. They are dark if too litle and burn out if too much. (I often use an LM317 as a current regulator, shown above, but not needed in this case).
a common way to control brightness is to toggle the voltage rapidly such as using a 555 timer to output a PWM signal. That signal can control an LM317 to provide sufficient voltage/current to an LED circuit. LED brightness depends on the percentage of time the PWM output is high.
an LM317 configured as a voltage regulator can be used to provide voltage to one or more LED circuits using series resistors. A 555 timer open-collector output can be used to ground the LM317 adjust terminal to lower the LM317 output to 1.25 volts, which is low enough to turn off the LEDs.
Greg, you didn’t miss much–the Swedish was mostly greetings and such.
Swede, I think the simplest answer to your question is that storing energy in capacitors pretty much means that you will be dealing with a DC source to power the LEDs. That is, the problem falls naturally into two parts: making a DC supply and then powering your LEDs from that DC supply.
Since my article already covers LED operation from DC, I will just say a little more about making the DC supply. About all you need for a simple supply is a capacitor and a diode. For example, connect the capacitor and the diode in series. Put the series combination across the AC input voltage and use the DC voltage across the capacitor as the output. Connect the diode’s cathode to the positive capacitor terminal; or connect the diode’s anode to the negative capacitor terminal. The capacitor’s voltage rating should be at least 1.4 times the highest RMS AC voltage that you plan to use. The diode’s voltage rating should be at least 2.8 times the highest RMS voltage. The capacitor is going to be charged through the diode once every cycle of the AC voltage and then discharge for 1/60 second until the same point in the next cycle. So the peak-to-peak ripple voltage on the output will be 1/60 * 1/C * I * V, where C is the capacitance in farads, I the load current in amperes, and V the DC voltage. For example, a ripple of 10 percent with a load current of 20 milliamperes implies
1/60 * 1/C * .02 = .1
C = 1/300 ~= 3333 microfarads
If, instead of a single diode, you use a bridge rectifier, the capacitor is recharged twice as often, so you need only half as much capacitance.
On a lighted toy-train car, you can use a polyphase bridge rectifier with pickups on both trucks to greatly decrease the flicker and dropouts over dirty track or turnouts, without resorting to connecting the pickups together. A polyphase b
Thanks for your responses, Mononflyer and gregc. I am aware of the LM317 and the 555 timer from scanning posts at another forum, but have felt their use was beyond me until now. I can duplicate what was done in Waynes photo and understand what you stated. I think more experimentation on my part is warranted.
Thank you, Bob. Your response was coming in at the same time I was responding to Monoflyer and gregc. Let me digest what you stated. I’ll be away for the next 10 days but will be back in touch after that.
Swede, I tried the LM317 current driver with and without a 1000 uF, 35 VDC capacitor. Here are the results in picture form. If anyone needs words, just ask.
I am now back and will have a few days to fiddle with various methods of powering LEDs. Bob, in your post of 12 days ago, you indicated the formula for Ripple Voltage was 1/60 * 1/C * I * V, where C is the capacitance in farads, I the load current in amperes, and V the DC voltage. You used 0.1 as its value to solve for C (3333 microfarads) but I don’t see where you inserted a value for V. What am I missing?
Also, could you provide me with a polyphase bridge rectifier part number suitable for use lighting a car? I’m not sure I will resort to using one except to satisfy my curiosity since it appears that track power to two LEDs hooked up as oulined in your 2008 write up, with resistor in series and with pick ups connected together, work just fine.
Wayne, if you happen to read this, I have not forgotten your post and will dig into it, also.
[quote user=“lionelsoni”]
I can’t get this thing to let me make a reply that takes more than a few seconds.
Bob sent the following information to my email account and asked that I post it here.
swede
Thanks for your e-mail, Fred. I am visiting friends in Fairbanks, where I once lived; and I think that the combination of a slow Internet connection, an old browser, and the new forum software have done me in. I hope that I will be back to normal in a week. In the meantime, I’ll answer you this way and ask you to post for me, if you don’t mind doing that.
You didn’t miss anything–I messed up. The text should have read something like:
…So the peak-to-peak ripple voltage on the output will be 1/60 * 1/C * I, where C is the capacitance in farads, and I the load current in amperes. For example, a load current of 20 milliamperes and a ripple voltage of 1 volt imply
Welcome back, Bob. In your post prior to the above quote, you mentioned a ripple factor of 10 percent and in the above quote, a ripple voltage of 1. I understand the correction. V (ripple voltage) equals I (load current) times 1/60, all divided by C (capacitance). Assuming the use of a variable AC transformer with maximum voltage of 20 VAC, would you use something other than 1 for the ripple voltage? Would it be more like 10% of 20 volts? I only pick 10% because you used it and I have read elsewhere that ripple voltage is pretty small.
Keeping in mind that 20 volts RMS will result in a peak rectified voltage of 28 volts, 2 volts peak-to-peak ripple should be just fine. However, you would not want the capacitance to be so low that you would miss the benefit of riding through gaps in the third rail. The discharge time will be about C/I. So, with a 20 milliampere load current, 330 microfarads will give you 16.5 milliseconds, which is not a lot. I would put in as much capacitance as there is room for, perhaps 5000 microfarads, which is about the size of a D cell, and get .25 seconds. The ripple will be much smaller than needed; but that’s okay.
Thank you for the response, Bob. Although I didn’t mention it in a earlier post, one of the reasons I reviewed your information closer was the fact that the answer to your early example was 3000 microfarads and when I found out its size and cost, I couldn’t believe it. Looking up the size of a 5000 microfarad capacitor, I see it is more like 3" diameter by 4" length. Pretty hard to hide and not inexpensive, either. I have some 1000 microfarad capacitors left over from other projects and, for fun, will cobble something together starting small and see what the difference is as I increase the capacitance.