X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Sender: To: lml@lancaironline.net Date: Fri, 02 Dec 2005 01:21:03 -0500 Message-ID: X-Original-Return-Path: Received: from xproxy.gmail.com ([66.249.82.193] verified) by logan.com (CommuniGate Pro SMTP 5.0.2) with ESMTP id 853080 for lml@lancaironline.net; Thu, 01 Dec 2005 11:30:42 -0500 Received-SPF: pass receiver=logan.com; client-ip=66.249.82.193; envelope-from=jeffreyb.peterson@gmail.com Received: by xproxy.gmail.com with SMTP id h30so314882wxd for ; Thu, 01 Dec 2005 08:29:58 -0800 (PST) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:mime-version:content-type; b=Jj0mc8EQhrxq3qyTINq7q9OQAoUiBWF7pGzdUfB45kS2DZ9q4QC6WuE5l461q26p2K0piWcp2/NBoZBeV8Mpf9C+rdPiLL3DfZlpPi81dOhAlVfYGc28D2j4k0YOEgEc5w7cHiD4V1qYSXQVa+cBekSGv9JSEA6c+e+tYW2Doqk= Received: by 10.70.104.5 with SMTP id b5mr2098334wxc; Thu, 01 Dec 2005 08:29:58 -0800 (PST) Received: by 10.70.115.11 with HTTP; Thu, 1 Dec 2005 08:29:58 -0800 (PST) X-Original-Message-ID: <7b85bf8e0512010829k607a512fpe6c8683126cc75d9@mail.gmail.com> X-Original-Date: Thu, 1 Dec 2005 11:29:58 -0500 From: Jeff Peterson Reply-To: jbp@cmu.edu X-Original-To: Lancair mail list Subject: Re: LED Position Lights MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_Part_55457_20053695.1133454598198" ------=_Part_55457_20053695.1133454598198 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Ed - So something like a Pulse Width Modulator for a motor would work very well here? Thanks for the welcome expansion of knowledge on LED's here on the LML. John On Wed, 30 Nov 2005 16:52:30 -0500, Ed McCauley wrote: *Using a transistor as a simple ON/OFF switch is an effective way to addres= s controlling the power of LEDS. With this approach there is no series elemen= t (resistor or transistor) that's dropping voltage/power and wasting energy i= n the form of heat. * PWM circuits can be used make very efficient sources for either motors or LEDs, but there is an important difference. Motors have substantial inductance and LEDs do not. First the motor. You can put the full battery volatge, 13 volts, across a motor for a 100 microsecond pulse and all that will happen is the current will climb postive a bit. Then you can switch the sign of the voltage and the current will come back down. If you keep the duty cyle 50/50, very little current flows and the motor does not turn, even with a 26 Volt peak to peak square wave across th= e motor. (The physics equation in operation here is V=3DL dI/dt.) If you want the motor to turn, adjust the duty cyle to say, 60 percent + and 40 percent -. Then a DC current flows. The reason to do all this switching is that the transistors that do the switching are always either fully on or off, and don't get hot. The LED is different. It is a diode, so current flows only one way. It ha= s very little inductance, so a short pulse of voltage causes an almost instantaneous matching pulse of current. Also the current climbs very rapidly as the voltage is increased. It is still useful to pulse the current to the LED, however, since the people perceive peak intensity. Most bicycle lamps use pulsed LEDs these days. The circuit turns a transistor fully on for short pulses and the battery internal resistance limits the current. Then the current is off, but the human eye remembers th= e bright flash. This increases battery life. I dont think turing such a transistor, fully on, would be a good idea with lamina LED and a battery capable of cranking an engine. You could probably make this work if you put an inductor in series with the LED, but I have not thought through the details of such a plan. -- Jeff Peterson ------=_Part_55457_20053695.1133454598198 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Ed -

So something like a Pulse Width Modulator for a motor would work very well = here?

Thanks for the welcome expansion of knowledge on LED's here on the LML.
John


On Wed, 30 Nov 2005 16:52:30 -0500, Ed McCauley <ed.mccauley@bltinc.com> wrote:

Using a transistor as a simple ON/OFF switch is an effective way to address controlling the power of LEDS. With this approach there is no series element (resistor or transistor) that's dropping voltage/power and wasting energy in the form of heat.

PWM circuits can be used make very efficient sources for either motors or L= EDs, but
there is an important difference. Motors have substantial induc= tance and LEDs do not.

First the motor.
You can put the full battery volatge, 13 volts, across a motor for a 100 mi= crosecond pulse
and all that will happen is the current will climb postive a bit. Then you = can switch the sign of the
voltage and the current will come back down. If you keep the duty cyle 50/5= 0, very little current flows and the
motor does not turn, even with a 26 Volt peak to peak square wave across the motor. (The physics equation in operation here is V=3DL dI/dt.) If you want the motor to turn,  adjust
the duty cyle to say, 60 percent + and 40 percent -.  Then a DC current flows. The reason to do all this switching is that the transistors that do the switching are always either fully on or off, and don't get hot.

The LED is different.  It is a diode, so current flows only one way.&n= bsp; It has very little inductance, so a short pulse
of voltage causes an almost instantaneous matching pulse of current. Also the current climbs very rapidly as the voltage is increased.  It is still useful to pulse the current to the LED, however, since the people perceive peak intensity.

Most bicycle lamps use pulsed LEDs these days.   The circuit turns a transistor fully on for short pulses and the battery internal resistance limits the current. Then the current is off, but the human eye remembers the bright
flash.  This increases battery life. 

I dont think turing such a transistor, fully on, would be a good idea with lamina LED and a battery capable of cranking an engine.

You could probably make this work if you put an inductor in series with the LED, but I have not thought through the details of such a plan.

--
Jeff Peterson
------=_Part_55457_20053695.1133454598198--