A number of folks have asked for the full story, so here it is.
I did go back and check some math and tweaked some text best as I could.
If others do the analysis for blocking or other parameters, please let us
know what you find.
This could be a great discussion.
73,
Chet, N8RA
--------------------
In VHF/UHF receiving, a primary goal is being able to hear very weak
signals, and frequently this needs to be done in the presence of very strong
signals on that same band. With that as the objective, I have attempted an
analysis of the situation.
For reference data and background info, I have used the test results
published by the ARRL in their product reviews, their explanation of these
tests, and the recent presentation by Bob Sherwood to the Boulder, Colorado
Amateur Radio Club in February. (see http://www.sherweng.com/)
Dynamic range measurements characterize a receiver's ability to function on
one frequency in the presence of one or more unwanted signals on other
frequencies. There are a number of different types dynamic range parameters
and effects. I will focus on the so called close-in third order dynamic
range. I enjoy contesting and other signals close in frequency can be a real
issue for us on the East coast. The further-out measurements for third order
dynamic range have been shown to be inadequate for real-world crowded
conditions, and that is why the ARRL had started to do the 5 kHz spacing
measurements a few years ago. If a different parameter is of interest to
someone (receiver blocking, for example) a similar analysis could be
performed.
I will examine 3 situations:
1 a Yaesu FT-897 being used on 2M
2 a good transverter in front of a previous generation IF radio
3 a good transverter in front of an IF radio having a true "roofing
filter"
First let's look at the FT-897 on 2M. In my experience, it is quite good at
hearing weak signals. During contests, it was OK, but I did have to steer
clear of the strong locals by quite a bit when I used it there.
The ARRL data for the 897's noise floor (aka minimum discernible signal, or
MDS) is -140 dBm, which converts to 0.02 uV in 50 ohms, and they further
listed the 12dB SINAD sensitivity as 0.2 uV. That shows why the FT-897 hears
weak signals well.
The close-in (5 kHz), two-tone, third-order IMD dynamic range of this radio
on 2M is 64 dB.
Now 64 dB above 0.02 uV is 32 uV. An accepted "standard" for S9 is 50 uV,
however the ARRL measured that it takes only 2.2 uV to read S9 on the 897's
meter. This very generous meter will therefore be reading S9++ on its
bargraph display for this 32 uv signal. This result is consistent with the
contest performance that I had with this radio, namely a lot of signals were
S9+ and I had to stay far away from them..
Now, what about putting a transverter in front of an HF IF radio?
Transverter specs and testing seem to be a bit sparse. The 2M Elecraft and
DEMI units each list a noise figure of a bit less than 1 dB. I could not
figure out how to translate that into an MDS, so I will use what I thought
was a very good sensitivity number of -140 dBm from the 897's data,
believing the transverters will also hear weak signals well with their low
noise figure. I then used a formula from the ARRL test descriptions
document: "Third Order Intercept = (1.5)(IMD Dynamic Range in dB) + (MDS in
dBm)" to calculate the IMD dynamic range from their intercept point spec.
That works out to be 107 dB for the Elecraft, and 97 dB for the DEMI, both
pretty good numbers. I will use the Elecraft for further analysis since that
is the one that I have used. Assuming its MDS is 0.02 uV, then the Elecraft
can handle a strong signal 107 dB above that, or 4477 uV before it is
adversely affected. Since this transverter has 25 dB of conversion gain,
the signal range passed over to the IF radio could be from 0.36 to 79690 uV
at the IF frequency. Can the IF radio handle that?
The ARRL does not test receivers at the commonly used IF of 28 MHz, but the
numbers taken at 14 MHz are likely similar. Pre-roofing filter radios like
the Icom 756 Pro II or III or the FT-1000MP all have an MDS about -129 dBm,
or 0.08 uV, so they will be able to hear the weakest signal presented to
them by the transverter, but not by a whole lot, so normally their
attenuator would not be set to reduce their front end gain by more than a
few dB in an attempt to preserve dynamic range and ability to tolerate
strong signals.
The close-in third order dynamic range of these IF radios (and likely the
TS-850) are all around 75 dB, which is much less than the transverter's 107
dB, so they will be the limiting factor here. This combo looks like it could
push the strong signal handling ability higher than the 897 by 11 dB based
only on the difference of the close-in third order dynamic numbers. But it
is probably not quite that much, since the 25 dB conversion gain of the
transverter pushed the signals higher up the scale toward "overload". We can
try to recover some of that with the attenuator, but we cannot knock the
weakest signal down too close to the MDS of the IF radio because that will
hurt the overall noise figure which we want dominated by the transverter.
The weakest signal from the transverter was 0.36 uv, which is about 12 dB
above the IF radio's MDS, so we could dial in maybe -6 dB from the
attenuator. So we have lost 5 dB from the capability of the IF radio to
handle strong signals, and now it is only about 6 dB better than the 897.
It is easy to get fooled by using an IF radio into thinking we have a great
setup because signals now appear stronger. They are strong though because of
the extra gain in the receive chain. The 2M signal strength that was
starting to cause problems was 64 uv, or 2 dB over S9 on a standard meter.
But wait, we added 25 dB gain through the transverter, knocked off 6 dB with
the attenuator, so will be reading around 20 dB over S9 and thinking we are
now far better off than the 897 case. In fact weak signal performance is
about the same and we are only 6dB better off in strong signal handling
ability which was further confused by the 897's super generous s-meter.
Check the specs for what input is required to get S9 on your IF radio, if
you care.
To add a side caution to the pitfalls of indiscriminately adding more gain
in the chain, say you also have a 20 dB remote preamp to overcome some
feedline loss. The transverter will likely still be OK with that, but
without cranking in more attenuator, you have now made your IF radio's
ability to handle strong signals worse by 20dB minus the feedline loss. .
You should not feel good about having achieved an S7 no signal hiss level on
your IF radio.
Lastly, let's look at the situation of a transverter used in front of an IF
radio with a true roofing filter, like the Orion, or K3. The Orion has an
MDS similar to the IF radios discussed above, but its close in third order
dynamic range is 92 dB. That's getting close to the DEMI transverter's
ability but still 15 dB less than the Elecraft. The Orion with Elecraft
could then handle a strong signal 22dB above what the 897 could (28 dB third
order difference minus 6 dB of attenuator). This would be about a 20 dB over
S9 2M signal, which would show up on the Orion's s-meter as almost 40 dB
over S9. The K3 with its 103 dB spec would be another 10 dB better yet.
Software Defined Radios next? It will be interesting to do the numbers when
the data is in for them.
Now, after this clinical analysis, I must mention (ramble about) some
practical considerations.
The obvious one, as Bob Sherwood points out, what good does it do to have a
receiver that can tolerate being close to a very strong signal if that
nearby signal is not clean. Having junk 60 dB down from your transmitter may
meet the FCC requirement but will still appear as a real (not spurious)
signal in someone's very good receiver. Now that we are building some good
receivers, its time to get to work on the transmitters and amplifiers. We
could start by adjusting transmit audio better. Hopefully you have done any
key click mods identified for your radio on CW.
Another factor is noise blanker performance. At my QTH I have a low level
annoying pulsing line noise in one direction. I have found that by having a
bit more front end gain I can get better noise reduction. Maybe the DSP gets
a better signature of the noise to work with. And since I usually don't need
to handle the 40 dB over S9 local, I am willing to loose some dynamic range
in order to suppress the line noise more effectively. Noise blankers can
also introduce their own effects and ruin your dynamic range. They seem to
be getting better in the newer radios that blend IF and DSP blanking, e.g.
K3.
_______________________________________________
VHFcontesting mailing list
VHFcontesting@contesting.com
http://lists.contesting.com/mailman/listinfo/vhfcontesting
|