Making a LED act like a filament lamp

[Allan Bennett]

Posted by Stevo on 03/04/2015 12:02:37:

....

Allan - I've sent you a load of files - have you got them??

Thanks Stevo, they're showing in my inbox now -- don't know why there weren't there when I looked before, as all bar one are dated yesterday.

I'll look at the attachments this evening. One thing that may be of use to you though is, instead of a diode to "regulate" the voltage to the PICAXE, I've been in the habit of using an MCP1702 voltage regulator to feed 5v to the PICAXE and, sometimes, to the devices it's controlling. It can handle input voltage up to 14.5v I think, and up to 500mA. So good for a 3S LiPo input if you want, and several regular LEDs.

[Dave Bran]

Well, my simple 50 plus year old H&M variable wire wound model railway controller from my youth and still working provides perfectly good variable lighting of LEDS (with their 12V resistor fitted) from the dimmest of dims to the brightest of brights.

Now I'm not suggesting you try bolting one in a plane, it'd need a long power lead, for a start, and CG might be a tiny issue, but it does prove that the complex "high tech" solutions being proposed are not exclusively required!!!!!

[Stevo]

"I think I'd drive the LED with a cheap brushed speed controller and do the programming in my Tx. Just 'cos that's what I have and know how to do."

 

Where's the sense of adventure?

And yes Dave, I can see how to control your railway controller... From Rx... measure the pulse width... on to a servo that moves the wiper back and forth...

 

Edited By Stevo on 03/04/2015 15:16:50

[Stevo]

Righto guys...

All sorted. PicAxe all programmed and feeds a MOSFET to switch on and off the LED, and takes in the Tx input as well.

My word - those Braincube LEDs are BRIGHT. I've now got visions of flying it towards me, hitting the switch to impress the members and being blinded by it all - rapidly giving it some UP while it illuminates the sky...

I've just got the one LED, just to experiment, will order the other wing tip LED at the end of the month. Plus it may need a landing light too

Allan - you should have all the code, diagram and board layout by now.

[Allan Bennett]

Posted by Stevo on 04/04/2015 17:33:45:

Allan - you should have all the code, diagram and board layout by now.

Yes thanks. I'll email when I've digested it all.

[John Olsen 1]

Maybe I can help clarify how to vary the brightness of an LED. As Dickw's graphs show, the current through an LED will increase quite rapidly with small increases in Voltage. This makes it hard to control the brightness by directly controlling the Voltage across the LED. Adding to the difficulty is the fact that the forward Voltage drop will vary with temperature, so the current will tend to increase as the LED warms up. Because of this it is usual to run LEDs in a constant current mode, or close to it. The simplest way of achieving this is with a series resistor. We decide what current we would like to run the LED at, and calculate a resistor that will give that current at the desired supply Voltage. We usually know the forward Voltage that the LED will typically drop. (This incidently differs with the different colours.) So we can subtract that from the supply Voltage and then use Ohms law to calculate the resistance needed.

So what happens if we vary that resistor to vary the brightness? Well, if we go too low in resistance, the current will go too high and the LED will rapidly shuffle off this mortal coil. So we would need to connect a fixed resistor in series with the variable one, so that the resistance can never go too low. This would work but will probably be very non linear and the resistance would have to go very high to put the LED right out.

However, if we vary the Voltage applied to the combination of LED and resistor, the current will vary too. If we halve the Voltage, we will roughly halve the current. This will also roughly halve the brightness. It might not actually look like half, because the response of our eyes to light is non linear. Also, I say roughly, because in fact the Forward Voltage drop of the diode will make a difference. If we start from zero Volts, nothing will happen until we reach the threshold Voltage of the diode. Then it will start to turn on, and as we increase the Voltage the current will increase, increasing the brightness.

I mentioned that the response of the eye is non linear. If you look through the viewfinder of an SLR camera where you can manually adjust the aperture you can see this. Everytime you close down one stop you are halving the light reaching your eye. eg each step from 2 to 2.8 to 4 to 5.6 to 8 to 11 to 16 is halving the light you see. So in that sequence, by the time you get to f16 you are only getting 1/32 of the light you started with...but you will still be able to see, especially if you give your eye a moment or two to adapt. So even if you arrange to dim an LED by a linear curve, it will still seem to come on suddenly and then not vary by so much towards the top end.

John

[Stevo]

OK.. Just heard from Braincube.

For information...The LED has a Vf of 3.2V, the series resistor is a 5.6 Ohm, 1/2 W device. Calculate that lot out and the series resistor is close to its tolerance (320mW) and the LED dissipates 900mW (5V supply, 250MA current).

Run it on 6V and the series Resistor COOKS

Edited By Stevo on 13/04/2015 18:45:16