TenTec
[Top] [All Lists]

[TenTec] Sample-to-Sample Variations

To: <tentec@contesting.com>
Subject: [TenTec] Sample-to-Sample Variations
From: mtracy@arrl.org (Tracy, Michael, KC1SX)
Date: Mon Mar 17 11:03:15 2003
On Mar 14, John Rippey, W3ULS said:

"To underscore Ed's point, I just came across the January '92 QST review of 
the Ten-Tec Argonaut II/Delta II transceivers. The ARRL lab found IP3 of 
the Argonaut to be -11 dbM while the Delta II IP3 tested at +2 dbM. Other 
indices showed considerable variations as well. The reviewer, Dave Newkirk, 
noted that the receiver portions of the two rigs were identical. He went on 
to say: "The receiver variations we found . . . reflect sample-to-sample 
variations in a single radio product."

It was a coincidence that ARRL had two radios with identical receivers to 
go over since, as Ed states, ARRL usually tests only one sample. So the 
variations noted by Dave Newkirk would have gone unreported otherwise."


Although the ARRL Lab nominally only tests a sample of one, I would say
that in the almost 6 years that I have been the Lab's Test Engineer,
for HF transceivers we have *frequently* looked at two rigs in any given
review, and in a couple of cases we have examined three (or more).

Now, these additional rigs don't go through the entire test battery of
course, but I always look at the "basic" performance on a couple of bands
(typically 80M and 20M), checking noise floor and dynamic range among
other things.

>From that experience, I can tell you that the noise floor and dynamic 
range figures will vary roughly 3-5 dB.  Greater variations do happen, but
that is rare in my experience.  Keep in mind that with intercept point,
the variation will be greater because it is effected by differences in both
noise floor and IMD dynamic range.  However, a 13 dB variation in IP3 is
greater than the usual amount of variation that I've seen, so I would 
suggest that it be considered abnormal rather than typical.

73, Michael Tracy, KC1SX

<Prev in Thread] Current Thread [Next in Thread>