Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

RS-232 FSO with 1W Luxeon

Status
Not open for further replies.

jorgealdo

Junior Member level 1
Joined
Oct 29, 2010
Messages
18
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,407
I am trying to develop a small FSO using the computer serial port and a luxeon rebel LED.

While i am very confident about the receiver design, i am still struggling with the transmitter :

mosfet.PNG

The working is simple :

LM317 is configured as a constant current source.

MOSFETS Q3 and Q4 switchs between two loads :

The LED and a resistor.

(When LED is on, the resistor is off and vice versa).

Well, i made this that way to avoid having the LM317 "sensing" the switching on and off of the real load (LED) by a single MOSFET (Two mosfets and two similar loads should make the LM317 not have to adapt fast).

My question is :

Is this really needed ?
Would this really work ?

And if the circuit is indeed correct, what is the best MOSFET for the task ?

I would want 10Mhz switching just to be safe. (Current is 0.350mA for the luxeon and voltage 16V to be safe)
 

go and simulate it. i think it will work..
NTE66 will do
 

Is this really needed ?
Would this really work ?
- A single transistor in a current source configuration and for the present signal polarity an inverter stage should basically do.
- The 10 k driver load resistance/Q3 gate capacitance sets a rise-time in a µs order of magnitude, so you can't achieve 10 MHz. 115K can hopefully work.
 

Is it possible to design a current source so fast as to react to the modulation ?

Example : If i modulate the LED at 10MHz the current source reacts at 100MHz, or something like that, being perfectly able to compensate for the on/off switching of the load.

About Q3 rise time, i cant decide wich resistor to use without selecting wich transistor and his input capacitance... (And on the current model i am using two transistors - npn pnp - to drive the mosfet gate).
 

The other possibility i am investigating is this :

act.PNG

The problem i see there is that theres no protection whatsoever.

If the Resistor goes thermal runaway, the LED will go kaboom and all the logical ports will follow suit (If the LED dies in a short).
 

If i modulate the LED at 10MHz the current source reacts at 100MHz, or something like that, being perfectly able to compensate for the on/off switching of the load.
Did you check, if the said LED's can be modulated at those high frequencies? Some visible LEDs have rather response.
 

The 10MHz thing was just an example, the serial port cannot go over 931,400bps on windows, so even if possible for the led, its not useable on computer... I bet onboard UARTs cant go more than 115,200bps in practice.

But Luxeon Rebel surely can go to 5MHz or more...
 

But Luxeon Rebel surely can go to 5MHz or more...
I assume so, although I'm not sure about the speed of white LEDs that use fluorescent material. In any case, I would go for a simple one transistor current source, as mentioned, supplemented by a push-pull driver, as you already considered. Usually an unsaturated BJT output transistor can be more easily driven as a fast pulse source, because it doesn't have that high input capitance
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top