[Laser] AM versus PAM

[TWOSIG at aol.com]

I have a question that some of you may be able to help me with.
 
This past weekend I did another demonstration of light communication at a  
local Engineering EXPO ( y'all do know it is Engineering Week? ) for the middle  
and elementary school kids.  This year I was successful in building a  
transmitter using an LED instead of a modified laser pointer.  (  I  only needed to 
talk across a gym, and it is absolutely amazing how hard kids at  that age 
will try to stick their heads in front of the beam. )  The  transmitter is simple 
and inexpensive, using an electret microphone, three  quarters of a quad op 
amp, a few resistors and capacitors, a transistor, and of  course the LED.  
Because of a slow start, my receiver was not ready, so I  used the same Ramsey 
receivers that I used for the last two years.   Among other things it "proved" 
that AM is fully compatible with a receiver  designed for PWM.
 
During the demonstration I got to thinking about the circuit and how to  
improve its performance.  There was a lot of ambient light in the gym,  enough 
that I could not see the beam spot on the wall behind the receiver.   That says 
there was lots of noise from the environment.  The system worked  well, but I 
thought I should increase the light output of the transmitter so it  would be 
stronger in comparison to the noise.  ( This system might get a  nighttime 
workout at greater range with some Scouts this Summer. )
 
Anyway, the circuit controls the LED current through the collector of  a 
2N2222A transistor that has an emitter current sensing resistor that feeds the  
signal to the inverting input of one op am.  The base of the transistor is  
driven by the output of the op amp.  The non-inverting input is driven by  an 
audio amp biased to half of the supply voltage.  The result is  that the light 
output of the LED is proportional to the drive current, in a  classic amplitude 
modulation (AM) system.
 
What I was thinking is that if I increase the bias current, the output  would 
increase.  However, I am running the device close to the ratings for  heat 
dissipation.  I realized that I could change the circuit slightly and  send 
pulses of current, with the peaks of the current proportional to the audio  
modulation.  That would be Pulse Amplitude Modulation (PAM).  If I  double the 
current and make the pulse duty cycle 50%, then I am running the  device at the 
same heat.  Ten times the current, 10% duty cycle.  That  would also increase the 
signal amplitude (if only during the peak) to noise  ratio.
 
Have any of you had any experience with pulse amplitude modulation?  I  am 
thinking that my standard detection system will spread the peak over the  total 
time. ( It is after all, a pulse width modulation detector. )  But  that would 
seem to lose the signal to noise ratio that I had.   I  thought of using a 
peak detector that had a decay time longer than the pulse  spacing, but short 
compared to the voice audio that I am trying to  receive.
 
Any suggestions?   I know that I can improve things by using some  receive 
optics, and I think that I have a good start on the optics to collimate  the 
light output.  I did some web reading on PAM, but did not get what I  consider to 
be practical information.  It did say that PAM, like AM is  subject to noise, 
but it does not say if PAM has benefit when compared to  AM.  Or if it 
requires complex circuitry to realize any benefit.
 
I have thought of going to PWM, but this circuit is simple and works  well.  
PWM would need to be a lot better performer to justify a major  change.  If 
PAM or AM work about as well, then I would need to put efforts  into the receive 
side.
 
 
Best Wishes
 
James 
N5GUI