Fwd: new ARRL rx measurement methodology

[DOUGLAS ZWIEBEL]

Elecrafters:

Here is an email I received (check the date below) in response to my
query about the "new" testing methodology at ARRL.  Specifically, I
asked about how to interpret the "dbc" measurement now include with
the analysis.  I don't think they even answered the question I asked.
However, here is a bunch more "stuff" - interesting, if not all that
informative.

And I wonder why they "didn't have room to explain them as fully..."
One would think that when everything is turned on its head, that they
would MAKE the room. ?????

de Doug KR2Q

---------- Forwarded message ----------
From: Tracy, Michael,  KC1SX <[hidden email]>
Date: Sep 18, 2007 10:00 AM
Subject: RE: new rx measurement methodology
To: [hidden email]
Cc: "Wilson, Mark K1RO" <[hidden email]>


Hi Doug,

I'm glad you asked, as we changed a lot of things at once and I didn't
have room to explain them as fully as I could have otherwise.

In my August 2004 article on Product Review testing (recommended
reading - if you haven't seen it already, it is on the Product Reviews
web page), I said that blocking dynamic range is a measurement of the
difference from the noise floor to a level of a single strong offset
(in frequency) interferer that causes blocking (aka desense) in the
receiver.  I also said that receivers can behave one of two ways in
this test - either they will gain compress, or the receiver's local
oscillator noise will increase.  The latter effect is known as
reciprocal mixing because it is caused by the interferer (however
"clean" it might be) mixing with the receiver's LO.  Well, as it
happens both effects actually occur in every receiver, but one of them
tends to dominate.  In the earlier tests where reciprocal mixing
occurred, we reported the result as a noise limited BDR, with a number
that was somewhat nebulous because it wasn't actually blocking, and it
wasn't the standard method of measuring and reporting reciprocal
mixing.

Thanks to new measurement techniques, we can now measure gain
compression and reciprocal mixing independently, and report both.  The
blocking gain compression figure indicates how high above the noise
floor an interferer has to be to cause the receiver to reduce its
gain. The reciprocal mixing figure indicates how far below an
interfering signal the noise from the LO phase noise mixing occurs.

In the FT-2000D, on 20 meters with the preamp off, the noise floor is
-122 dBm.  The blocking gain compression at 20 kHz offset is 136 dB,
so the interferrer that caused gain compression is +14 dBm.  However,
the reciprocal mixing is -103 dB relative to the interferrer, so the
noise would be -89 dBm, which is far above the noise floor.  So in
this case, the reciprocal mixing dominates.  In fact, for all of the
blocking measurements on the FT-2000D, the reciprocal mixing dominates
over the gain compression.

For third order IMD dynamic range and intercept, we also chose to give
more information than in previous reviews.  Higher IP3s are actually
expected for higher input levels on most receivers, and the reason can
be seen in Figure C of the sidebar. Look at the third-order response
for the receiver input levels in the -30 dBm to -20 dBm range; rather
than a straight line (as seen at lower levels), the response is very
distorted, and by that I mean distorted from the "typical" response of
devices much simpler than receivers. This behavior is not new by any
means (the receiver used to produce fig. C is 1991 vintage), but
wasn't investigated fully until relatively recently. Even after we
found out about it, we struggled for some time on how to best report
it and finally decided to go with the format of the 2000D table.

As to what constitutes good or poor performance on IMD and IP3, it is
a little hard to say just yet, but will get easier as more new method
data becomes available.  The IMD dynamic range from past reviews was
based on the noise floor level.  In the new data, the noise floor
level is the first row in each set of frequencies/offsets.  In past
reviews (since 1993 anyway), the IP3 figure was actually calculated
from higher levels, corresponding to a "5" level on the receiver's
S-meter.  As all receivers are different, that didn't really give the
reader a direct comparison.  So we chose to use a level of the old
Collins (and ITU) standard for S5, which is -97 dBm.  The figures
based on this level are closest to past published IP3s.  The inclusion
of data based on a 0 dBm input level is entirely new, and was the
suggestion of several members of our advisory group.  Most folks will
never see interferers anywhere near that high, but folks near
broadcasters and in metropolitan areas certainly will (it is probably
not coincidence that one of our advisory folks lives on a hill about
20 miles from NYC).

73, Michael Tracy, KC1SX, Test Engineer
ARRL, the national association for Amateur Radio
Tel: [phone]
_______________________________________________
Elecraft mailing list
Post to: [hidden email]
You must be a subscriber to post to the list.
Subscriber Info (Addr. Change, sub, unsub etc.):
 http://mailman.qth.net/mailman/listinfo/elecraft   

Help: http://mailman.qth.net/subscribers.htm
Elecraft web page: http://www.elecraft.com