Lights and Resistors and DCC...Oh my!

The 920 Ohm resistor is too high value you should have 680 or 750 Ohm in series with one bulb. A 680 Ohm resistor will give a voltage drop of  10.2 V at 15 mA and 750 Ohm will give 11.25 V over the resistor. When you connect two bulbs in paralell you should have half that value. The drawback with one resistor for two paralell bulbs is that if one bulb burns out the other will get much to much current and will burn out too. My suggestion is to try with the resistor you already have in series with one bulb and if you want it a little brighter reduce resistor value to min 680 Ohm. then get two more resistors for the remaining bulbs.

  How about 14volt bulbs? They will last longer and run cooler. There should be one resistor for each bulb. But then if Im going to all the trouble soldering resisors I always switch to LEDs. I melted a shell on a RS1 when the bulb dislodged from its mount while putting the shell back on and didnt realize it till it was too late. 

   Pete
 

locoi1sa wrote:

The best general rule of thumb is: one light bulb - one resistor to match.

LED's are a much better alternative.  you still need a resistor but the brightness of a LED doesnt vary much so resistor value versus voltage isnt nearly as critical

Thanks all, for the input you've given so far.

I've rechecked Digitrax's recommended practice, and taken the above suggestions and came up with a compramise that seems to work ok, I question longevity though.

Digitrax recommends a 500 ohm 1/4 watt resistor for my incandesant lamps, and the Digitrax Command station set to the N scale voltage setting.  I prefer to use the HO scale voltage setting (for no particular reason) so I thought that a 500 ohm might be a touch small.

The above suggestions of 680~720 ohm(IIRC) seemed a little high after testing my current 920 ohm resistors running a single bulb (they were pretty dim on HO, and no light on N settings), So I compramised and went to a 560 ohm resistor, but up'd to 1/2 watt (less heat?) 5% tolerance.

The result is slightly dim on N scale voltage setting, and slightly bright on HO scale voltage.

I should probly series in a 20~35 ohm resistor for each circuit to provide a cushion (for longevity= 580~595 ohm total resistance), but at this point, I think I will move on to other aspects of detailing this pair of Locos. I have yet to decide if I'll model the "as delivered" version without ditch lights, or the later version that had ditch lights added by Conrail.

If I've not mentioned the entire project prior, It is a pair of SD40-2's from the first run delivered to Conrail in 1977, modeled from Early production Athern Powered and Dummy units. Powered unit was repowered with A-line/Protopower West can motor, trucks and gears tuned (tried Earnst regear set, but was very unpleased-too delicate-9 tooth gears stripped as soon as they were installed), added Details West plow and sunshades, drilled and tapped clip on coupler mounts for Kadee #47's and torsion-spring boxes...more to go. The Dummy unit was drilled to reduce weight, and will be getting a sound system and function only decoder, .

pkeppers wrote:

LED's are a much better alternative.  you still need a resistor but the brightness of a LED doesnt vary much so resistor value versus voltage isnt nearly as critical

I appreciate your input, and may consider LEDs for future projects, but as I stated in my OP, I am not interested in LEDs for this project.

Before I get accused of being one-sided, or close minded, let me explain why: I wanted to mount the lights in the actual lens sockets of the loco, and for the particular models I am building I was under the impression that LEDs would be too bright, and too white.

I do not know of any LEDs smaller than 3mm diameter, and my lens sockets are no more than 1.5mm-ish. I would have had to fabricate some sort of enclosure that would be large enough to hold the LED behind the lenses, and direct the light through them without letting any past the sides (un-lit number boards directly adjacent). This "enclosure" would have incroached on the space in the cab that I want to put an interior, without being able to see the wires or "enclosure".

With my limited experiance with these kinds of modifications, I thought the Incandesant lamps would be a "safer" and more "do-able" aproach for me.

As I said, I am open to the possibility of using LEDs in some of my other projects, infact I've even considered Fiber-Optics, but at this stage I think its a bit advanced for my ability(fabricating mounting solutions).

KingConrail76 wrote:

should probly series in a 20~35 ohm resistor for each circuit to provide a cushion (for longevity= 580~595 ohm total resistance),

You know there is another way to do this.  Actually it is the way I normally do it when playing directly with the track power rather than the output of a decoder.   Two diodes in series create a voltage drop of approximately 1.4V.  This makes a great voltage regulator for the lamp(s).  The power is constant and no issues about burning out one lamp.   Since diodes are directional you, of course, need two in each direction. Since a full wave rectifier is basically 4 diodes it works too.  So the circuit becomes the resistor in series with two diodes (facing the same way) in series.   Put two more diodes in parallel with the first two but facing the opposite direction.  Then put the lamp in parallel with the diodes.  Addjust the resistor value to vary the current and achieve the desired brightness but no more than 1.4V will ever reach the lamp.

BTW. another reason for using lamps especially for head lamps of locomotives is if you want to use the special effects provided by some DCC decoders.  LEDs tend to just flash on-and off rather than doing a good simulation of a MARS, Gyralite, or other non-constant effect.