Skip to main content

In December RR Model Craftsman there was a flicker free LED circuit schematic that is probably copyrighted and can't be shown here, but was basically a bridge rectifier paralleling three  in series super caps , feeding into a 3.3 low dropout voltage regulator.  The regulator output then fed three or more paralleled LEDs.  One configuration had individual current limiting resistor per LED.  The other circuit had all the LEDs paralleled and used a pot as current limiter doubling as a dimmer control.

The article says the 14VAC DCC voltage is rectified  and charges the three series 5.5V super caps (16.5V max).  Unless I am wrong 14V* 1,414 is about 19V for bridge with capacitor output.  So is DCC voltage not a real 14V rms or something much less so that the bridge voltage is less than the 16.5 VDC across the supercaps?

Second thing that puzzled me was the three un-resistored LEDs depending on one current limiting resistor.  As I remember diode theory the diode with the lowest Vf will light and others may be dim or off.  Experimentally I paralleled three LEDs from same manufacturer and batch and one lit brightly and others not so much.  I looked up the LEDs in Mouser and the devices don't seem to have any built in current limiter.  Of course more LEDs draws more current thus changing voltage drop across the resistor which could account for dimming.

Obviously the circuits work for them but something seems off in my thinking.  What am I missing, or have I been out of school too long and forgot formulas and textbook theories? 

Last edited by rrman
Original Post

Replies sorted oldest to newest

Well, given that you've stated, I wouldn't worry about that circuit in any case.  Why would you charge up three caps to 16 volts and then drop 13 volts in a resistor, throwing away about 80 percent of the power?

I have good luck paralleling identical LED bulbs, but if they are different colors or different manufacturers, it's a no-go as a rule.

The circuit you are talking about sounds like someone had no idea what they were trying to accomplish and just went about it with a sledge hammer instead of the correct tool.

There are many things wrong with the circuit idea, starting with what you pointed out, no over-voltage protection for the supercaps, they do NOT like to be run over their voltage ratings.  Then throwing away most of the power in a resistor or resistors makes no sense either.

 

 

Flicker free LED's with DCC?

How about a small resistor, say 100 ohms (isolate the track signal), an electrolytic cap, say 330uf, and a diode.  Put two or three LED's in series, more strings in parallel if you need them, each string has a dropping resistor.  Total cost is peanuts, and far cheaper than all those supercaps, not to mention smaller.  You're not heating the room with all the power being dissipated in the resistors, and the LED's and other components are not running over their specifications. 

This is just off the top of my head, I'm sure it could be optimized if I wanted to spend more than a minute on it.

 

gunrunnerjohn posted:

Well, given that you've stated, I wouldn't worry about that circuit in any case.  Why would you charge up three caps to 16 volts and then drop 13 volts in a resistor, throwing away about 80 percent of the power?

Actually dropping the 16V to 3.3 with voltage regulator not resistors per se but still 13V dissipated in pass tranny.

The circuit you are talking about sounds like someone had no idea what they were trying to accomplish and just went about it with a sledge hammer instead of the correct tool.

Exactly my thought.

There are many things wrong with the circuit idea, starting with what you pointed out, no over-voltage protection for the supercaps, they do NOT like to be run over their voltage ratings.  Then throwing away most of the power in a resistor or resistors makes no sense either.

 

 

Flicker free LED's with DCC?

How about a small resistor, say 100 ohms (isolate the track signal), an electrolytic cap, say 330uf, and a diode.  Put two or three LED's in series, more strings in parallel if you need them, each string has a dropping resistor.  Total cost is peanuts, and far cheaper than all those supercaps, not to mention smaller.  You're not heating the room with all the power being dissipated in the resistors, and the LED's and other components are not running over their specifications. 

Yes, much simplier solution.

This is just off the top of my head, I'm sure it could be optimized if I wanted to spend more than a minute on it.

Nice to know I am not thinking up the wrong LED tree!

 

 

Add Reply

Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Ste 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

×
×
×
×
Link copied to your clipboard.
×
×