The purpose of Speech Processing is to significantly increase the average to
peak transmit power on SSB while not objectionably distorting the audio. The
reason is that the human voice has a very low average to peak level ratio -
often estimated at "one to ten." With that as the input to the transmitter, the
average power output of a 1500W station will be somewhere around 150W, with
occasional peaks at 1500W. Last night I was trying to work 4N6FZ from St.
Louis, MO on 80M (World Series Rain Delay :-). Although I normally run firmware
V1.373b5, due to it having several areas of improved performance over the V2
firmware, yesterday I had loaded V2.059d in order to make some measurements
(Looking at the reduction in audio levels at the headphone output and at the
AUX audio output). Before calling the 4N6 I set the mic gain and then set the
Speech processing to 5, which is where the audio starts to have a slightly
"clipped" sound. I called the 4N6 and he came back to me. We then
alternated 6 or 7 transmissions while he tried to get my call correct. He
obviously had noise problems because he had a good signal in Missouri (I was
receiving on a K9AY loop). He finally gave up and logged an incorrect call for
me. With the SP setting I was using - limited by early distortion of the audio
- I was increasing my average talk power from about 150W (No Speech processing)
to about 250W while having peak output power of 1500W (Alpha 87A). I needed
Montenegro on 80M for the DX-Challenge so I decided to load V1.373b5 and try
again. I set up the mic level and set the speech processing to 7 (Just before I
start to hear any early clipping distortion in the audio). I called the 4N6 and
he came back to my call (I was averaging about 400W average power with peaks at
1500W). He asked me to repeat my prefix and he had one letter in my suffix
incorrect. I immediately increased the SP setting to 8 (Now I was hearing some
early clipping distortion in the monitor audio) and
my average power out increased to 500-800 watts (A 6 to 9 DB improvement in
average talk-power). I gave him my call twice and a signal report and turned it
to over to him. He immediately came back with my correct call and responded to
the signal report I had given him.
Now 4N6 is not "long haul" DX from Missouri, but last night the conditions over
the path on 80M were difficult due to his noise level (Some one came back and
congratulated me for getting though from "West of the Mississippi" :-) and
great Speech Processing made the difference between working the 4N6 and not
working him.
There were some improvements in the V2.059d release - and some new bugs - and
the Speech Processing was improved. Because I do both SSB DXing and CW DXing,
the performance of Speech Processing is important to me. For me, Speech
Processing performance is increased talk-power with good audio. The Speech
Processing in V1.373b5 was giving me 6 to 9 DB increase in average talk-power
over no Speech Processing (Sometimes I see people set the mic gain way down so
they can set Speech Processing on and still have good audio. Of course that
often results in no real improvement in the talk power and maybe a loss in
talk-power). The improvement in talk-power can be easily seen by changing the
RF Power meter reading from "Peak" to "AVG" and then estimating the average
position of the meter - not great precision but reasonable quantative accuracy.
While the V2 code has made progress in performance since its release, it still
has areas where it needs to be improved to match the V1 firmware performance -
such as the Speech Processing effectiveness. I have been running V1.373b5 in my
Orion since I received the Orion back from Ten-Tec over a week ago (A number of
upgrades done on the boards and the V2 mod installed and re-aligned), with the
exception of an hour last night on V2, and it has run flawlessly (I typically
have the rig on 12-15 hours a day). For a couple of years I had been seeing
occasional stability problems (Using V1), such as when I used the voice
recorders. However, V1 has been totally stable since the Orion came back. I
don't know why, but I am very pleased with its performance and stability! I
still look forward to a further improved V2 firmware release.
73,
Merle - W0EWM
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|