LEDs -vs- Bulbs ... Basic Principles

[richG]

Below is a dissertation by Jim Betz of the Yahoo Soundtraxx group concerning LEDs and lamps. Some of you might want to copy and paste the message into a word document and print it out. Could make easier reading.

When you are thinking about these two very different devices for
producing light the following will help you to understand their
behavior better.

Bulbs are, essentially, "voltage devices" - if you vary the voltage
you vary the amount of light coming out of them. And
changing the current will not, significantly, change
the amount of light. More importantly it is having too
much voltage that causes them to burn out.
In addition, a bulb 'consumes' its element in order to
produce light. And the higher the voltage the quicker it
consumes it. And even a brief amount of too much voltage
will cause them to blow.
When changing a circuit in order to 'control' a bulb
you should pay attention to the voltage produced - and
the bulb will last a -lot- longer if you use lower
voltages (within the range of voltages that work for that
bulb). And that's why we refer to the resistor in the
circuit as a "voltage dropping resistor".

LEDs are, essentially, "current devices" - if you vary the current
you vary the amount of light coming out of them. And
changing only the voltage will not, significantly, change
the amount of light. And it is too much current that
causes LEDs to burn out.
In addition, an LED has a much slower 'consumption'
of its element in order to produce light. But the higher
the current the faster it consumes. And even a brief amount
of too much current will cause them to blow - in a rather
spectacular one bright pulse of light.
We refer to the resistor in a circuit for an LED as a
"current limiting resistor". If you use lower current
values (within the range of current that work for that
LED) the LED will be dimmer ... and last longer ... but
not significantly - because LEDs already have such long
lifetimes when compared to bulbs.

There is one other important distinction between bulbs and LEDs.
The actual source of the light in an LED is physically a much
smaller amount of space than the source of light in a bulb. And
the 'package' (the plastic around it) acts as a lens for the LED
whereas the package (the glass and air) of a bulb does not act
as a lens. When you are picking an LED for a model it is important
to pay attention to the "viewing angle". Luckily most of our
installations are in castings/parts that help a -lot- with this
and it is relatively easy to get an LED to work for a specific
application.
And it is just this "point source" characteristic of LEDs that
makes the light reflect off the railhead so much more than a
bulb. And it is directly related to our perception that they
are much brighter. They are brighter (in terms of the amount of
light produced) ... but the point source nature of them makes
that seem even more.

Lastly - there is the issue of heat. A good rule of thumb is that
bulbs give off lots of heat and a little light - and LEDs give off a
lot of light and very little heat. In fact, it is this very aspect
of LEDs that is the source of their long life compared to bulbs.
The lifetime for LEDs are typically in the tens of thousands of
hours. It is rare for a bulb to last even a thousand hours and
many of the bulbs we use in our hobby burn out in as little as 100
hours ... unless you give them too much voltage.
This heat thing can be very important - it is almost impossible
to get enough heat out of an LED that will cause plastic to melt.
With bulbs you need to plan for heat dissipation and you need to
pay attention to how close to the plastic around it the bulb is.
Lower voltage bulbs such as the 1.5v bulbs produce less heat than
12v ... but if you run a bulb at a lower voltage it will produce
a lot less heat (and still produce sufficient light).

The thing to wrap your head around - and I agree it isn't an easy thing
to do - is that we still "just use a resistor to modify the circuit".
This seduces us into thinking that the LEDs and lamps work in a similar
fashion ... nothing is further from the truth when you consider how they
actually work.

All of the above are statements that are intended to be "relative".
What I mean by this is that I'm talking about bulbs -vs- LEDs here
and about how to think about them when making the choice between
one and the other or in how to think when you are designing or
modifying the circuits that drive them.

Both devices, like all of electronics, follow Ohm's law - and, in the
final result it is heat that consumes them and/or causes them to burn out.

There is one other thing to consider and that is how your decoder
produces different light values such as are used in a Mars light,
roof beacon, flashing ditch lights, and headlight dimming (Rule 17).
Most decoders do this using PWM (pulse width modulation). This is due
to the fact that PWM is relatively easy to do (fewer components) and
also easy to control (the main chip/processor does this easily).
PWM is also how the speed to the motor and "silent drive" are created.
If you think about the current/voltage "flow" to the LEDs/bulbs then
PWM breaks that up into lots of little ons and offs ... and the more
time it is in the 'on state' the longer the pulse and the more light
(either brightness or just how long it stays on) will be produced.
A MARS light is produced by varying the pulse in such a way that
both the length the light stays on and how bright it is are changing.
And we should also say that when using PWM the 'size' of the pulse
can also be varied both in terms of how long it is on but also in
what voltage+current flows (hence a MARS light with something that
doesn't move!).
Some decoders 'prefer' bulbs when they are doing effects such as
a MARS light. By that I mean that the effect will be better if
the decoder is controlling a bulb than if it is an LED. One good
way to know if the decoder is changing its use of PWM is whether
or not it has a CV to tell it that the device in question is an
LED or a bulb.

Personally, I prefer LEDs to bulbs. So much so that I have a 'rule'
that if I am opening up a locomotive to work on the electronics in
any way I will replace any bulbs with LEDs. The time it takes to do
this is rarely enough longer than it would take me to replace just
the bulb (because it burned out) ... so I figure that it is a case
of pay me now or pay me later. There are certain times when I get
lazy and don't replace bulbs - especially if the bulb is 'buried'
were it takes a lot of extra effort to get at it (such as many of
the recent articulated steam engines).
I have even opened up a loco just to replace working bulbs with
LEDs!

- Jim
-------------------------------------------------------------------------------------------------------------------------------------------------

Rich

[Yampa Bob]

That's a very interesting article. I ran across another page that has a little more information about the actual "nuts and bolts" of LEDs.

http://www.theledlight.com/LED101.html

Additional information and links on the advance page.

http://www.theledlight.com/technical.html

[richG]

Thanks for the link. Good information to have on hand. One issue I have seen in a couple other forums is the fact that some LED lights do not look quite "right" when off. You can some times see the LED behind the lens. I have a couple locomotives like that but I can live with it. The 4-6-0 does not have that issue because of the "periscope" used to transmit the light to the headlight lens.
Small lamps usually look more "natural" when off.

Rich

[Yampa Bob]

I'm on a crusade, don't want anyone to miss out on this. I have several of these units, very handy when working with LEDs, lamps and all other electricals.

http://www.multimeterwarehouse.com/dt830bf.htm

[Jim Banner]

The digital volt meters in Bob's link are an excellent deal.  They are reliable, tough (all of mine have made trips to the floor and all still work), and they can withstand operator errors that would send conventional VOMs to the meter grave yard.  They are cheap enough to have one in the shop, one in the train room and one in the tool box.  And they are available from a variety of suppliers.

These meters come in several varieties.  Of the three I have, the one in the shop is just like the one in Bob's link, except it is black.  The transistor Beta tester in it is a good for a quick check on the health of known transistors and for determining the polarity of unknown ones.  The meter in the train room has a built in continuity buzzer which is great when you are tracing wires under a table or across the room and cannot see the meter to read the resistance.  The one in the tool box has a plug-in temperature probe which has helped solve mystery problems in outdoor layouts such as overheating decoders in overworked locomotives on hot, sunny days.

When it comes to electrical problems, the choice comes down to seeing the light with the help of electrical test gear or cursing the darkness without it.  And when it comes to electrical test gear, the affordability, durability, and ease of use of the digital volt meter is second only to sticking you finger in the light socket.

[Yampa Bob]

Thanks Jim, I appreciate your positive comments about these meters. I have an expensive RMS meter in my shop for critical work, but no way would I subject it to the spasms of track circuits.

I have a friend who is a master electrician, he has these in all his trucks. It just proves you should never judge the skills of a technician by the tools he uses.

Fellow modelers, don't be mislead by the cheap price.  Trust me, once you use one you will never want to be without it.