Hi Tom, George,
I fully concur with your remarks. There are several issues at play here:
The typical -20dB carrier set point on the filter flank in a conventional
filter-type SSB generator serves two purposes:
1. It holds opposite-sideband suppression to an acceptable level, whilst at
the same time ensuring a reasonable LF audio response (typically 250 ~ 300Hz
@ -6dB.
2. It aids the carrier balance of the balanced modulator, to hold carrier
suppression to an acceptable level.
Shifting the carrier point up the filter flank in an attempt to extend the
LF audio response can degrade opposite-sideband and carrier suppression to
the point where regulatory limits for spurious emissions are exceeded.
It is interesting that ITU-R Recommendation M.1173 (for the maritime radio
services) specifies the audio response of an SSB transmitter as 350 ~ 2700
Hz @ -6dB. This frequency range was derived from Bell Labs studies done in
the 1930's.
The portion of the human speech band which carries the articulation (an
index of intelligibility) falls within this frequency range. Significant
transmitted energy below 350 Hz wastes transmitter power without
contributing significantly to articulation at the distant station. A
high-end response significantly above 2.7 kHz signifies transmitting with
more occupied bandwidth than is required for satisfactory communication.
This is not a kindness to adjacent stations on the band, and in extreme
cases can also be in breach of radio regulations. A baseband response much
wider than 3 kHz can also introduce excess noise into the transmitted
signals, thus degrading S/N at the receiver.
Considering a DSP radio with which I am familiar, the IC-756Pro/Pro2 "low
end" extends down to 100 Hz @ -6 dB in the "WIDE" occupied-bandwidth
setting. The corresponding "high end" is 2.9 kHz. Short of rewriting the DSP
firmware, there is no changing these parameters.
http://www.qsl.net/ab4oj/icom/occbw/main.html
Overdriving the speech amplifier of any SSB transmitter in the low-frequency
region of the audio response, as you so succinctly described, can generate
harmonics of the LF energy. These will mix with midband components to cause
intermodulation products falling within the IF passband. This intermod will
degrade articulation, and overall audio quality, even further. A baseband
source with a -6dB audio response of 100 Hz ~ 3 kHz will sound very pleasant
when matched to a compatible DSP-based SSB transmitter such as the
756Pro/Pro2 in the "WIDE" occupied-bandwidth setting (or the Ten-Tec
equivalent).
"Rule No. 1" in DSP: If you attempt to drive the ADC beyond "all ones"
digital output, all bets are off; the digital output is corrupt, and no
longer represents the analogue input. This is not negotiable.
As you implied, the user should refrain from second-guessing the engineers
who designed the radio, especially if this results in degraded signals and
interference to other stations on the band. As radio amateurs, we should be
conscious of our obligations towards our fellow-hams. We should also avoid
creating situations which might induce radio licensing authorities to write
strict occupied-bandwidth and/or audio-response limits into the
amateur-service regulations.
Best 73,
Adam, VA7OJ/AB4OJ
North Vancouver, BC, Canada
http://www.qsl.net/ab4oj/
-----Original Message-----
From: tentec-admin@contesting.com [mailto:tentec-admin@contesting.com]On
Behalf Of Thomas Jednacz
Sent: Wednesday, January 15, 2003 13:05
To: Barry N1EU; tentec@contesting.com
Cc: Peter AA7E Finch
Subject: RE: [TenTec] Re Microphones--one more thing (long)
Hello all:
Sorry to jump in and spoil your fun. DO NOT change the carrier offset. You
will then be transmitting on both upper and lower sideband!! Filter rigs are
set at 300 Hz cut off so the filter slope will will get you -30 dB on the
other sideband. If you change the offset the filter will not have a steep
enough slope to drop the other sideband to the required level. Filter rigs
are not adaptable to less than 300 Hz audio without creating problems. I
have looked at the fellows on 178 and find that they are using up to 40 dB
boost (over normal audio levels) at 100 Hz and driving the audio and or A/D
converter in the DSP chip crazy. They say their RF stages are all linear.
True - BUT they generate all kinds of junk in the audio stages by driving
them non-linear and transmit 12 - 15 KC wide signals when they only intend
to transmitt 4Khz. A full DSP rig (Pegasus/Jupitor, Icom 746 Pro/756Pro,
Kachina, and of course the new Orion) are the only ones that will properly
transmit 100Hz to 4 Khz audio and be clean.
Play with equalizers and all the other audio stuff but don't try to make a
system design do what it is not supposed to do. Do not consider HI-FI/SSB
with anything but a full DSP radio as above.
When playing with HI-FI/SSB lets understand what we are doing, use good
engineering practices and have some courtesy in sharing frequencies with our
fellow hams.
Tom, W7QF
-----Original Message-----
From: tentec-admin@contesting.com [mailto:tentec-admin@contesting.com]On
Behalf Of Barry N1EU
Sent: Wednesday, January 15, 2003 3:29 PM
To: tentec@contesting.com
Subject: Re: [TenTec] Re Microphones--one more thing (long)
--- Mark Erbaugh <mark@microenh.com> wrote:
> Could you share with us how you run a response curve
> and how you determine
> the needed EQ. I would like to do some analysis work
> on my GoldLine / W2IHY
> 8 band Eq / Pegasus setup.
Mark, a full treatment of your question would take a
long time. I'll make a few comments off the top of my
head with the disclaimer that there are guys right
here on this reflector who know more than me. I've
been hacking at it for a few years though.
The best way to judge mic equalization is with your
own and other folks' experienced ears. On a
hifi-capable receiver, listen to a lot of ssb signals
closely for what you do and don't like about their
sound. Listen to the guys on 14178 and on the TT net,
and you'll hear some really good and not-so-good
sounding audio. By hifi-capable, I mean a minimum of
2.8khz bandwith with carrier offset aligned properly
to extend bass response to below 100hz.
Run your receiver audio (line level) into your
soundcard and get some recording and spectral analysis
software. There's some freeware/shareware out there
(just search around), but currently I'm using CoolEdit
Pro for recording and SpectraPlus for spectral
analysis. Record WAV files of those signals that you
do and don't like. You can open the WAV files in
SpectraPlus and it will show their response curves
after running a spectral analysis on them. After
doing this many times, you can correlate what your
ears are telling you and what the computer is telling
you. I'm sure there are lower cost alternatives to
SpectraPlus, but you'll have to look around.
For setting your own EQ, you can record yourself off a
second hifi-capable receiver while transmitting
(choose a phrase with a few sentences and keep using
it as your standard) and get a good idea of what you
sound like as well as run a spectral analysis on it.
Solicit many opinions from others on the air who can
listen critically as you adjust the audio.
There are a few links on my Web site under "Audio" to
NU9N, KC4PE, W3OZ, etc. where there's a wealth of
information and different perspectives on getting the
equalization correct. There's also a ssb-audio
reflector on Yahoo Groups.
GL/73,
Barry N1EU
http://www.albany.net/~bg
__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|