X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Sender: To: lml@lancaironline.net Date: Sat, 03 Dec 2005 23:13:25 -0500 Message-ID: X-Original-Return-Path: Received: from web34905.mail.mud.yahoo.com ([209.191.68.184] verified) by logan.com (CommuniGate Pro SMTP 5.0.3) with SMTP id 862292 for lml@lancaironline.net; Sat, 03 Dec 2005 18:43:01 -0500 Received-SPF: none receiver=logan.com; client-ip=209.191.68.184; envelope-from=n103md@yahoo.com Received: (qmail 82204 invoked by uid 60001); 3 Dec 2005 23:42:13 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Message-ID:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=x1x1AVg6f5sELAWPuiu3tgk0qILAAjolu54ExtllRXnP6RhlnlDqUVodxaz4G6x3gGg3LzZW6MU5gMLQizTwYLFOP6W26q6Zl9F5QZT4j3EDP/xEHSFlSStaARGff2hp7NdKARL4BE1SY/majCvoEkUm8tek+PYu8lCd7P7zwps= ; X-Original-Message-ID: <20051203234213.82202.qmail@web34905.mail.mud.yahoo.com> Received: from [69.12.132.145] by web34905.mail.mud.yahoo.com via HTTP; Sat, 03 Dec 2005 15:42:13 PST X-Original-Date: Sat, 3 Dec 2005 15:42:13 -0800 (PST) From: bob mackey Subject: LED Position Lights X-Original-To: lml@lancaironline.net MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="0-1962154976-1133653333=:77383" Content-Transfer-Encoding: 8bit --0-1962154976-1133653333=:77383 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit As many have mentioned, there is some non-linearity in the human eye's response that favors pulsed bright lights over continuous dim lights. But that effect is rather small. The much greater effect is the power efficiency of the circuit when used in pulse mode. There are two effects here. The first is that a smaller ballast resistor is used along with a short duty cycle. More power goes to the LED and less to the resistor. A second effect should also be considered. The light output of a red LED increases faster than linearly with current. Some portion of the current always goes to non-emissive leakage. That is, the first milliamp is wasted. The second milliamp makes light. The third milliamp makes light. So 3mA is twice as bright as 2mA, not 3/2 as bright With blue, green, (and white) LEDs, there is less leakage due to different materials and higher bandgap voltage. So there is less motivation to pulse the current. If you look at common handheld consumer electronics you'll usually see that the red LEDs flicker, but the blue ones don't. There are very good reasons to work that way when you have only a small battery. For aircraft uses, it really makes very little difference whether pulse width modulation or a large ballast resistor is used. Either way will be much more efficient than an incandescent lamp with a color filter. There was also the suggestion to use LEDs in series. That's a good idea, but don't take it too far. If there is no ballast resistor, the current will vary greatly with small variations in system voltage. Let's assume that we want the nav lights to stay lit when the alternator is off. The system voltage might be 12V with the engine off and 14.6V with the engine running. I = (Vsys - Vled) / R I usually use: 4 red LEDs (Vled = 4 * 1.8V = 7.2V), 3 green LEDs (Vled = 3 * 2.2V = 6.6V), or 2 white LEDs (Vled = 2 * 3.8V = 7.6V). In the red case, that leaves I = (Vsys - 7.2V) / R If we want 30mA per strand at the higher voltage (14.6V), then choose R = (14.6V - 7.2V) / 0.030A = 247 Ohm [might actually choose 300 or 330 ohm because it is readily available] When the voltage drops to 12V, the current will be: I = (12V - 7.2V) / 247 ohm = 19.5 mA So it will be about 2/3 as bright with the alternator offline. If we isntead used 8 red LEDs (8 * 1.8V = 14.4V) and a small ballast resistor to limit it to 30mA at 14.6V R = (14.6V - 14.4V) / 0.030A = 0.67 ohm [about equal to the wires and contacts] Then, when the voltage dropped to below 14.4V, the lights would go completely off. If the voltage surged to 15V, the current would increase to almost an Amp, burning out the LEDs. So.... stack a few LEDs, but don't get rid of the ballast resistors! That still applies if you use PWM, unless the PWM circuit is peak current limiting. -bob mackey bob AT pure-flight DOT com ps - thanks to Bob Smiley and Scott Kreuger for explaining that the "squat" switches are really airspeed switches and don't respond to squat. --------------------------------- Yahoo! Personals Let fate take it's course directly to your email. See who's waiting for you Yahoo! Personals --0-1962154976-1133653333=:77383 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: 8bit
As many have mentioned, there is some non-linearity
in the human eye's response that favors pulsed bright
lights over continuous dim lights. But that effect is
rather small.

The much greater effect is the power efficiency of the
circuit when used in pulse mode. There are two effects
here. The first is that a smaller ballast resistor is
used along with a short duty cycle. More power goes
to the LED and less to the resistor.

A second effect should also be considered. The light
output of a red LED increases faster than linearly
with current. Some portion of the current always goes
to non-emissive leakage. That is, the first milliamp
is wasted. The second milliamp makes light. The third
milliamp makes light.  So 3mA is twice as bright as
2mA, not 3/2 as bright

With blue, green, (and white) LEDs, there is less
leakage due to different materials and higher
bandgap voltage.

So there is less motivation to pulse the current. If
you look at common handheld consumer electronics
you'll usually see that the red LEDs flicker, but
the blue ones don't.  There are very good reasons
to work that way when you have only a small battery.

For aircraft uses, it really makes very little
difference whether pulse width modulation or a
large ballast resistor is used. Either way will
be much more efficient than an incandescent lamp
with a color filter.

There was also the suggestion to use LEDs in series.
That's a good idea, but don't take it too far. If
there is no ballast resistor, the current will vary
greatly with small variations in system voltage.

Let's assume that we want the nav lights to stay
lit when the alternator is off. The system voltage
might be 12V with the engine off and 14.6V with
the engine running.

   I = (Vsys - Vled) / R

I usually use: 
4 red LEDs (Vled = 4 * 1.8V = 7.2V),
3 green LEDs (Vled = 3 * 2.2V = 6.6V), or
2 white LEDs (Vled = 2 * 3.8V = 7.6V).

In the red case, that leaves
   I = (Vsys - 7.2V) / R

If we want 30mA per strand at the higher voltage
(14.6V), then choose
   R = (14.6V - 7.2V) / 0.030A = 247 Ohm
[might actually choose 300 or 330 ohm because it is
readily available]

When the voltage drops to 12V, the current will be:
   I = (12V - 7.2V) / 247 ohm = 19.5 mA
So it will be about 2/3 as bright with the alternator
offline.

If we isntead used 8 red LEDs (8 * 1.8V = 14.4V)
and a small ballast resistor to limit it to 30mA at
14.6V
   R = (14.6V - 14.4V) / 0.030A = 0.67 ohm
   [about equal to the wires and contacts]

Then, when the voltage dropped to below 14.4V, the
lights would go completely off. If the voltage
surged to 15V, the current would increase to almost
an Amp, burning out the LEDs. 


So....
stack a few LEDs, but don't get rid of the ballast
resistors!

That still applies if you use PWM, unless the PWM
circuit is peak current limiting. 

-bob mackey
 bob AT pure-flight DOT com

ps - thanks to Bob Smiley and Scott Kreuger for explaining that the
"squat" switches are really airspeed switches and don't respond to squat.

Yahoo! Personals
Let fate take it's course directly to your email.
See who's waiting for you Yahoo! Personals --0-1962154976-1133653333=:77383--