[Laser] fundamentals

Glenn Thomas glennt at charter.net
Fri Feb 16 13:53:22 EST 2007


Hi James!

First off, thanks for the intelligent conversation, a very uncommon 
occurrence these days. On to your note...

At 06:26 PM 2/15/2007, James Whitfield wrote:
>Glenn
>
>I think that you are asserting that for the same transmitted signal that
>there is no benefit to one audio carrier frequency over another.   I am
>pretty fuzzy about this point because it seems that others in this group
>said there is.  Maybe we have been talking apples and oranges, and I just
>could not keep track.

That may be. I still think we're in violent agreement.

>To (hopefully) clear things up I would like to get comments about the
>following two scenarios:
>
>1.   A laser is amplitude modulated with an audio tone ( could be sine wave,
>but if it is simpler to build assume 50 percent duty cycle square wave ).
>The tone is also modulated, chose either MCW or BPSK and a data rate so that
>any change of the data always occurs on the rising edge of the tone.  ( This
>prevents a spike if the data changed in the middle of a tone pulse. )  The
>tone frequency is approximately 300 Hz.
>
>2.  Set up an identical system except that the tone frequency is five times
>higher.
>
>Both signals are received on the same equipment and fed into a computer
>sound card for processing.  The receiving equipment does not attenuate or
>amplify one of the frequencies more than the other.
>
>Now does the signal using the 300 Hz tone have an advantage over the signal
>using the 1500 Hz tone for being used for communication?

If the modulation is the same in both cases and the maximum bandwidth 
of the modulation is less than 150 Hz and an ideal detector is used, 
no advantage.

 From an information theory point of view the audio carrier frequency 
(or subcarrier? The laser is the carrier) is not very important. If 
the modulation applied to the different frequency subcarriers is the 
same, then the bandwidth of the resulting signals will be the same, 
so there is no reason from an information theory perspective to 
prefer one over the other.

>I remember trying to suggest that I though information theory suggests that
>the two have the same data bandwidth and should therefore be processable
>with filters for equivalent SNR.  Seems to me it was suggested that the AM
>signals had 600 and 3000 Hz bandwidth and would necessarily have five times
>the noise in the wider signal, and further that no audio processing could
>remove the noise.  If the discussion had included a caveat that there was a
>filter between the detector and the sound card, I missed it completely.

You could put an analog filter between the detector and the sound 
card, but DSP filters are more versatile. Perhaps some kind of analog 
"roofing filter" to cut back the gross out of band noise, followed by 
a matched bandwidth DSP filter? The point of filtering, either before 
or after digitization, is for the receiver bandwidth to match the 
transmitted bandwidth. Exactly how you go about doing that is less 
important than the fact that you've done it.

As for bandwidth, that's determined by the modulation on the 
subcarrier. The frequency of the subcarrier doesn't have much to do 
with it. However, a wider bandwidth signal will require a wider 
bandwidth receiver, which will admit more noise to the receiver.

>I took from the the discussion that there is a mantra for optical
>communication: low frequency carrier tones are better than higher frequency
>tones.
>
>James
>N5GUI

Dunno about this. From K3PGP's website 
http://www.k3pgp.org/frontend1.htm, "The PIN diode response is more 
than adequate with zero bias for 800 Hz MCW and BPSK as well as 
subcarriers up to approx. 20 kHz or so."

However, he also says (http://www.k3pgp.org/frontend2.htm), "If you 
are primarily interested in weak signal optical communications you 
may want to investigate the use of lower modulation frequencies which 
appear to offer enhanced performance when used with this frontend" 
and provides experimental evidence. Also see 
http://www.k3pgp.org/lflaser.htm. Have a look.

73 de Glenn





More information about the Laser mailing list