BLI: LEDs and capacitors

[cmarchan]

QUOTE: Originally posted by cwclark

The capacitor will wire in between the + and - wires...it is used to charge then store power .....when the wheels of the train go across dirty track, there is a momentary power loss which will cause LED's to flicker if there is no capacitor in the circuit...the capacitor discharges when this happens supplying mometary power to the LED so it won't flicker...I use them on all my caboose LED lighting...a 100 - 250 muf capacitor should work.....I wire then like this...negative pick up wire, negative leg of capacitor, resistor, LED anode leg, LED cathode leg, positive capacitor leg, and then back to the positive pick-up wire ...I can get fancy and wire them so that the LED will light in both directions by using some diodes but i only run my cabooses in one direction and find that it's not necessary....chuck

This configuration is for reducing flicker - this is another topic; capacitors oppose an abrupt change in current because of their characteristics. This does not apply here.


Apparently, there are some misconceptions as to how a capacitor is used in electronic circuits. This is common; my students had several theories that are similar ( electronics instructor for 8 years).

Unless the decoder is capable of a gradual current startup for lamps ( the Tsunami has this feature to simulate the power-up of a dynamo) it will require a soft-on circuit to achieve this effect. RC (Resistive-capacitive) components are used in such circuits.

BTW IC's dont drive the capacitors; it is the capacitor's charging effect that controls the IC.

[tstage]

Fred,

That makes total sense to me. I might have to look into that. I wonder if that's something that would be mentioned in the latest QSI decoder manual? AND...whether it could be programmed similarly with other decoder brands and locomotives. Hmmmmmmm.... It's worth a try.

Thanks again everybody! [:)]

Tom

[fwright]

Does BLI use a capacitor in their lighting circuit to cause the LED to light up and shut down slowly? Yes and no.

Today's decoders are basically microprocessors with some analog driving circuitry. Individual electronic components are replaced as much as possible with processor digital logic (software). So the lighting brilliance would be a digital program that is converted to an analog output as the very last step.

In the innards of a microprocessor are capacitors, transistors, and resistors. And the drivers for the analog circuit would also contain these components. But the actual dimming/lighting rate is likely controlled in software and may even accept input from a CV.

my thoughts, your choices
Fred W

[cwclark]

The capacitor will wire in between the + and - wires...it is used to charge then store power .....when the wheels of the train go across dirty track, there is a momentary power loss which will cause LED's to flicker if there is no capacitor in the circuit...the capacitor discharges when this happens supplying mometary power to the LED so it won't flicker...I use them on all my caboose LED lighting...a 100 - 250 muf capacitor should work.....I wire then like this...negative pick up wire, negative leg of capacitor, resistor, LED anode leg, LED cathode leg, positive capacitor leg, and then back to the positive pick-up wire ...I can get fancy and wire them so that the LED will light in both directions by using some diodes but i only run my cabooses in one direction and find that it's not necessary....chuck

[cmarchan]

A capacitor cannot be placed in series with an L.E.D. It will cease to draw current once charged and in a D.C. circuit will appear as an open device. What you are looking for is a circuit where the capacitor is used with a resistor to slowly turn on a device, such as a transistor. In this scenario, the resistor limits the current flowing through the capacitor; the cap shunts the bias current of the base of a bipolar transistor until it's charge develops a voltage drop across the base-emitter junction. There are IC circuits that can be used in a similar fashion. Under no circumstances should a cap be placed across the lamp; theorically this may work, even with a resistor in series, but the charging current surge could damage the decoders function transistor (yes I know this because I've repaired decoders in the past). Another problem is you will defeat the emergency blinking (shorted motor) feature. If you are serious about trying this we may be able to develop a solution with a few basic components. Using an L.E.D. is easier since it does not require much current.

[modelmaker51]

You would need some sort of timing circuit I would imagine. Caps usually charge quickly and discharge slowly unless controlled by an IC. With DCC (I'm no expert) I believe it would be controlled by the CV setting.

[tstage]

A couple of questions for you electronic brainiacks out there.

1. Does anyone know right off hand whether BLI uses capacitors to power up their LED headlights - e.g. on their 2-8-2 Mike?

I was thinking today that I really like that particular feature of the headlight powering up slowly and power down even more slowly (in contrast to most LED headlights that just come on instantaneously) and wondered if BLI used capacitors to achieve that effect. It would make sense, since it takes about 10-15 seconds for the LED to slowly die down to nothing.

2. If the above inquiry happens to be true, how difficult (and risky) would it be to add a capacitor to a different locomotive head light to mimic what BLI is able to achieve on their 2-8-2 Mike?

The one issue and concern that I can think of right off hand is making sure that the discharge rate of the capacitor not exceed the voltage of the LED. Could that be avoided if the capacitor was soldered to the decoder in the following sequence?:

blue (common) wire --> capacitor --> resistor --> LED --> white wire
Would that work or is the whole idea too far fetched? I'll freely admit ignorance because I'm not electronically savvy. It just seemed to make sense to me logically. Would there be a risk of frying the decoder accidentally just by soldering a capacitor in inline.

I appreciate the input everybody and I look forward to hearing your responses. [:)]

Tom