driving IR led

Started by October 5, 2004
Hi all,

I've got an IR led connected to the F149 and written a small RC5 driver api
for sending ir pulses to a device.

Problem is that the LED doesn't send it's IR. I've measured with an
oscilloscope via a IR sensor (works with the original control) but for some
reason my lupse on the led doesn't go to 0Volts.

I've got the cathode of the led connected to 3.3V, the anode to a transistor
Collector the emitter to a 10Ohm resistor to ground. This switches ok but
not when the LED is connected. The power source can source the 300mA.

Any suggestions???

thanks martijn

Beginning Microcontrollers with the MSP430

Hei Martijn,

I don't know whether you have described it correct or not, but are you
sure you connected the cathode to Vcc? Have you tried to invert the IR
led, by connecting the ANODE to Vcc?

Regards,
Adriano.
>> I've got the cathode of the led connected to 3.3V, the anode to a transistor Collector the emitter to a 10Ohm resistor to ground. This switches ok but not when the LED is connected. The power source can source the 300mA.

Try putting the anode to 3V3, the cathode to one end of the resistor and the
other end of the resistor to the collector of the (NPN I assume) transistor.
Then connect the emitter of the transistor to ground.

You don't say whether you have any base biasing resistors. You should! Depends
on what the transistor is, to what values you need, but at the vert least
have a series resistor into the base of the transistor from the MSP.

You can probably get away without using a gate resistor if you use a FET.

Hope this helps. :-)
I may have misunderstood but if you really have the current-limiting
resistor between the emitter and ground on an NPN then the transistor will
never go lower than the voltage dropped across the current limiter. Why not
put it above the collector? And I assume you're using an appropriate base
resistor too, to get the switching you need without drawing too much current
through the msp...

--Bruce
On Tue, 5 Oct 2004 12:41:51 -0700 you wrote:

>I've got an IR led connected to the F149 and written a small RC5 driver api
>for sending ir pulses to a device.
>
>Problem is that the LED doesn't send it's IR. I've measured with an
>oscilloscope via a IR sensor (works with the original control) but for some
>reason my lupse on the led doesn't go to 0Volts.

>I've got the cathode of the led connected to 3.3V, the anode to a transistor
>Collector the emitter to a 10Ohm resistor to ground. This switches ok but
>not when the LED is connected. The power source can source the 300mA.
>
>Any suggestions???

It is the wrong way round. The anode should connect to +3.3v.
--
Ralph Hilton
Connecting the transistor/resistor as martijn described makes sense,
because this is a constant current driver, I=(Voutput-0.7V)/R. No
need for a base resistor.
But it should really be the ANODE at Vcc.

Wolfgang
Anode goes to 3V3, cathode to collector of transistor (I use a little
FET, with 470R to the gate, this clips the hoigh frequency parts of the
square and reduces heating effects), base of transistor has 470R
resistor, current limit resistor between Vcc and anode, or collector and
cathode. With 100r in the emitter, at 300mA you get .3*100R = 30V!!!
drop, hence the collector will NEVER go low enough. Depending on the IR
LED used (off the top of my head around 1.6V Vf) then you need to drop
1.7V at 300mA, this gives around a 6R resistor, and peak power of 1.8W.
depending on the duty cycle you will need to size your resistor
acordingly, and also MUST make the track wide enough to carry that
current. a 6 or 8 mil track really won't cut it.

Cheers

Al
I add a base resistor to current limit the I/O pin in case of some cock
up happening. better safe then fried for the sake 0f $0.0022.

Al
Hi Adriano, Robert and Bruce,

Thanks for the reply's and yes I mis typed the anode is connected to 3v3
it's the base resistor I forgot and indeed putting the 10E resistor on the
other site helped. Though I think a fet would be a better solution, I think

thanks