Re KH7T's post on NB vs. NR reproduced below:
If John's informative discussion results from my post and following
discussion, I want to specify: There is no confusion at K6XT regarding
the difference between these 2 functions. On my O1 the NB works just
fine when I have the rare occasion to need it at my rural, largely
vertically-polarized-impulse-noise-quiet location with the occasional
electric fence pulses.
The NR so far leaves much to be desired and, I say again, is
significantly less effective than my Timewave DSP59+ (which has no NB
function at all). The Orion NR function inhibits good copy on weak
signal narrowband filter settings. I refer only to CW as I have no
experience tuning fone. Oddly, on my O1, even strong signal fidelity is
reduced because noise is somehow "added" to the signal with NR turned on
whereas the "added" noise is absent with NR turned off. Most intriguing.
I continue to experiment with AGC settings to see if AGC has something
to do with this. I want to thank K7HP for his information and
measurements. Hank implied that apparently there are 2 AGC loops, one of
which is not disabled when we users set AGC OFF. Listening to strong EU
stations on 40 CW this evening I confirmed that turning AGC off seems to
have little effect on the quality and quantity of volume. This is an
unexpected result. Traditionally, turning AGC off in the presence of
strong signals will overload either the ears or the receiver. But it had
no effect on the strong signals I was tuning, implying that the AGC OFF
setting isn't.
73 Art
k6xt at arrl dot net
tentec-request@contesting.com wrote:
Today's Topics:
1. 285, 288 Filters FS (PaulKB8N@aol.com)
2. Wanted: Orion II (Ed Cummins)
3. Re: Orion NB vs NR (John Buck)
4. Re: Wanted: Orion II (ron)
5. Re: Wanted: Orion II (EI6DL)
6. Re: Orion NB vs NR (Gregory J. Knapp)
7. Re: [Orion] Orion NB vs NR (Pfizenmayer)
8. SOLD = Idiom SCAF (W.D. (Doc) Lindsey)
9. encoder for Omni VI (Paul DeWitte K9OT)
10. FS or Trade, Model 238, 229 tuners (George R. Griesbach W5GRG)
11. Re: 301 remote tuning knob (Dave Edwards)
12. TenTec Orion I V2 (Bernhard Hoidn)
13. V2.056 (jalankfor@linxure.net)
----------------------------------------------------------------------
[Snip]
------------------------------
Message: 3
Date: Thu, 05 Oct 2006 08:44:26 -1000
From: John Buck <kh7t@arrl.net>
Subject: Re: [TenTec] Orion NB vs NR
To: Zenvrim@aol.com
Cc: Orion@contesting.com, tentec@contesting.com
Message-ID: <4525528A.4040001@arrl.net>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
It seems that a lot of people confuse noise Blanking and Noise Rejection.
NOISE BLANKER (NB)
My NB is extremely effective on ignition noise and my neighbor's weed
eating electric fence. The S9+ fence disappears and with not much
distortion.
I am not sure what the difference between the so called Hardware and the
Software options are, but sometimes one seems to work better than the
other. NB is effective against very short impulse type noise that can
be somewhat irregular. NB does not work on band noise but does
sometimes help static crashes.
Usually older hardware implementations have a separate wide bandwidth
detector near the front of the receiver and if an impulse is detected, a
switch shuts off the signal path near the front end for a short period
of time. In effect, it puts a hole in the signal reception to prevent
the pulse from getting spread by the narrow band filters and detectors.
If a system has sufficient dynamic range to not be saturated by the
impulse then the impulse can be identified in one signal path and used
to blank a second delayed signal that you then use for detection. This
approach can completely remove the short impulse from the signal you hear.
It works very well. Sometimes my electric fence is S9+. The NB makes
it disappear into the S1 to 3 background. I often have to set it at 8
or 9. The Orion V2.059d is better than my K2 and that is pretty good.
NOISE REJECTION
The Noise Rejection is an attempt to remove background noise and
everything else that is not signal. There are many different algorithms
for this. It works better if the bandwidth is set larger than the
signal bandwidth so the algorithm can get a grip on the noise in order
to remove it.
It is not just a BW reduction approach although sometimes it sounds that
way. TT has tried several different schemes. I am not sure how the
current approach works on cw. It works quite well on weak SSB and often
makes an unreadable signal pop out of the noise, although it sounds a
bit distorted. I am not sure that it works that much better than
adjusting the threshold and Bandwidth just right but it is a lot quicker.
NR does not work on impulse noise and other stuff that may saturate the
filters or detectors.
I like the idea of identifying the noise and subtracting it from the
signal. I think it works and will work better in the future. I thought
it worked better on a earlier version but it worked poorly on some of
the interim versions. It is better now in V2.059d but more distortion
than I like on SSB.
Aloha,
John KH7T
[Snip]
|