Hi again,
Well, I have been informed by a "very reliable source"
with added data, at least new to me.
1. The IP3 data posted at the Ten Tec web site, the
curve for the Orion, was collected with the two strong,
intermodulating test signals OUTSIDE the roofing filter
pass band! Not with both inside as I had assumed in
my previous postings about this topic.
Therefore, the typ. IP3 data of +25 dBm or so is ONLY
about the front end of the Orion. That is the analog and
digital circuitry following the roofing filter was not necessarily
tested, and of course the analog AGC of the Orion was NOT
activated by the test signals, as they were greatly attenuated
by the skirts of the roofing filter. The analog AGC is only
triggered into action by signals reaching S9 + 35 dB or so
(-38 dBm) within the roofing filter pass band.
Only any non-linearities within the front end components
could cause the IMD's generated during the test by Ten Tec
engineers.
So the IP3 data presented by Ten Tec shows the excellent
quality of the Orion front end: the front end ham band only
filters of the main rcvr; the diode antenna rcvr switching array;
the preamp switch ( the preamp is switched OFF, out of the
signal path for the IP test); and the first mixer, a high-level JFET
unit which converts all ham band signals to the 9 mHz IF.
The switch selectable roofing filter follows these components.
In CW mode, the dial frequency of the Orion is centered in
the xtal filter; so even at tone spacings down to 1 kHz, both
the interfering tones are heavily attenuated by the xtal filter.
2. The second bit of info, new to me: the Orion xtal roofing filter
will be automatically selected (default, but the operator may
elect to defeat this) to be the filter next widest beyond the
DSP filter selected bandwidth! So, when the ARRL Lab
tests for IP3 using the Orion 500 Hz DSP BW filter, the
Orion will automatically select the standard 1 kHz BW 9 MHz
roofing filter. That is very interesting to know.
And, so did the Ten Tec engineers unit do the same thing when
they ran their test. The 1 kHz BW roofing filter was in use.
And this explains why the roll off in IP3 as the test signal
spacings begain to be less than around 5 kHz: some of the
test signal began to "leak" through this roofing filter as the
spacing between the two became more narrow. This leak
through test signal causes a little more IMD in the following
IF stages.
This is going to be a "very interesting" radio! And a bunch
different from my Corsair II!
73, Jim KH7M
|