"james" wrote in message
...
On Sat, 13 Jan 2007 22:13:25 -0500, "Jimmie D"
wrote:

+++
+++"james" wrote in message
om...
+++ On Sat, 13 Jan 2007 17:43:52 -0500, "Jimmie D"
+++ wrote:
+++
++++++
++++++"Jimmie D" wrote in message
. net...
++++++
++++++ "Telstar Electronics" wrote in
message
++++++ ups.com...
++++++ Jimmie D wrote:
++++++One thing did get me thinking. I
++++++ have a nice scope to look at the output of my TX and to lookat
the
++++++ output of
++++++ the processor , for me its a piece of cake to setup. How can
you do
+++this
++++++ without test equipment?
++++++
++++++ Absolutely, I have a nice Tektronix scope I use...
++++++
++++++ Don't miss seeing the new CB Radio Speech Processor Prototype
at
++++++ http://www.telstar-electronics.com/d...s/WhatsNew.htm
++++++
++++++ But how does most of the people who buy them do it?
++++++
++++++Even a good quality limiter/compressor is just anotheer splatter
box if
+++its
++++++not set up right.
++++++
+++ ************
+++
+++ The theory behind a compressor amp is to provide a more constant
power
+++ signal to the modulator. There have been many schemes in the past 70
+++ yrs or so to do that. Even to the point of what was once called Super
+++ Modulation.
+++
+++ A properly setup with compression will not have excessive compression
+++ and also not use the compression stage as the major gain stage. Where
+++ most compression schemes come into dislike is when the person using
it
+++ feels he needs to get far more gain from the circuit than is really
+++ necessary and so much compression that you can hear a roach fart 20
+++ feet away.
+++
+++ The idea behind audio compression and other audio/rf processing
+++ schemes is to increase signal to noise ratios. When using AM and the
+++ carrier is already 30 dB above the ambient atmospheric noise then any
+++ processing is typically useless. The gain in signal to noise ratio is
+++ not worth the effort or even the cost of adding such a circuit. Audio
+++ compression is totally useless on FM unless you want a constant
+++ deviated signal. For what ever reason that one would want this is
+++ beyond all logic.
+++
+++ Even when the AM signal is marginal to the atmospheric noise, audio
+++ compression alone yields between 1 and 2 dB improvement in the
+++ received signal to noise ratio. On AM, when you are at 6 to 8 dB
+++ signal to noise ratio, improving one to two dB is generally worthless
+++ improvement.
+++
+++ james
+++
+++
+++Clippers are much more useful. Ususally you can clip off the top 8 or 9
db
+++without substantial effecting intelligebility of the voice. This does
make
+++for a louder signal without overmoduation. Compression works best when
used
+++with an expander on the rx end. But how often are you going to see
that..
+++
**********

The goal of both compression and clipping of the audio signal is
reduce the peak to average power of the human voice. Typical Human
voice peak to average power range is about 3:1.

More like 30 to 1

Clipping can be as high as 30 dB as long as there is sufficient
filtering after the clipping stage to restore the audio bandwidth to
its original bandwidth. Otherwise you have splatter or increased
modualted bandwidth.

My experience has been that 9 or 10 db is enough clipping. I cant imagine 30
db of clipping

Compression is somewhat more effective in reducing the peak to averge
power in the voice. It can be implemented with better results. A major
drawback to compression over clipping is increased cost with a more
complex circuit and more difficult adjustments. All to often in CB
work the compression stage also doubles up to be mic amp and a large
gain stage at the very low level stages of the audio stream. I
personally don't like this as it requires to many adjustments for one
stage. Audio compression is better done, in my opinion, in or near the
final stage, high level, of the audio stream. This too is not without
its drawbacks either.

From my experience compression is more to allow consistent modulation when a
change in voice, such as the difference when I talk on the radio and my wife
talks. IF she tries without readjusting the mike gain she may not be heard.

Clipping really removes audio not needed for effective modulation and
amplifies that that contains the most intelligence. This often shows up on
the rx S meter as less than an S unit change over an unclipped signal but
sounds like 10db or more increase.I dont have a clipper except for this
peice of junk I threw together the other day but when I did have one the
reaction was usually" wow how big is that amp you just turned on."

On SSB average power out of a 100 wat tx on voice may be only 12 watts
average add a clipper and you can increase that to 80

My preference is a clipper with some AGC action to keep the average output
level faily constant.


Neither circuit requires the receiver to have any special ciruitry to
enhance intelligibility.

james