After spending some time with my ICOM 746pro, I realized that the rig is
an amazing radio for SWL. It is a great ham transceiver, but I was surprised
how well it receives AM. It has been FAR superior to any other radio I have
owned, whether ham rig or SWL rig. My observations have been confirmed by
the
technical review below. These radios have been having
trouble with a particular IC, causing some of the transmit sections to fail.
It does not affect receive. There are fixes available for them, but the used
price has been driven down as a result. They are out there commonly for
under $1000.00 US, due to nervous owners trying to unload them.
There has been much discussion, review and praise for the ICOM R-75. I am
sure that it is a good radio. However, if you want the real deal and a top
of the line receiver at any price level, check out the 746pro. Its only
weakness is that reception below 1.6 mhz. is attenuated. There are people
out there now (including the gentleman who wrote the review below)
who are pursuing a fix to that issue.
I hope this helps someone looking to get into the hobby, or someone looking
to upgrade their station. 73,


Dave


IC-746Pro Impressions & Mods, Dallas Lankford, 6/13/04

Introduction

Having owned two 16 bit DSP receivers, the NRD-545 and the WJ-8711A, I was
curious if the 32 bit DSP receivers performed any better. There aren't any
"receiver only" 32 bit DSP receivers, so my choices were among various ham
transceivers. After studying the field, I kept coming back to the IC-746Pro
(IC-7400 in Europe). It (and the 756Pro(II)) had crippled MW and LW bands,
but after studying the schematics, it appeared that in principle these
defects could be fixed. Prices of the 746Pro have been dropping, and late
model used ones have sold on eBay recently for under $1000. So I decided to
give one a try.

Initial Impressions, June 4th and 5th

It has excellent sensitivity above 1.8 MHz with Preamp 1 or Preamp 2. With
Preamp 2, above 20 MHz it may be more sensitive than the R-390A. Not a good
evening for comparison, though.
Frequency readout to 1 Hz makes manual ECSS tuning easy if you like ECSS (I
don't).
Users can select any three SSB bandwidths up to 3.6 kHz for the wide, med
and narrow filters (displayed as FILTER 1, FILTER 2, and FILTER 3, which I
suppose is more accurate since FILTER 1 can be wide or narrow or in between,
etc.).
AM bandwidths are fixed at 3 6 and 9 kHz (nominal). Users cannot change
them. I haven't measured them yet, but they sound good.
Users can three select AGC release times for each mode (except FM). The
factory default release times for AM were 3, 5, and 7 seconds. I changed
them to 0.3, 2.0, and 6.0 seconds. To turn the AGC off, you have to make
"OFF" one of the 3 release times. Dumb. So you really have only two
release times for each mode if you want OFF selectable by a single button
push or two. But that's probably sufficient.


MW and LW bands are desensitized, and the preamps are deactivated. I knew
this before I bought it. We'll see if I can't fix it. Measured MW
sensitivity is 3 uV (usual parameters). Not as bad as I expected (the ARRL
measured 6), but not good either.

I made some preliminary measurements of the 746P filters last night. Well,
really just one... the 6 kHz filter. I didn't measure the 6 dB BW. The 60
dB BW is (very preliminary) about 11 kHz. Nothing to write home about. But
the filter just keeps going down, and down, and down........... At 80 dB
down I had 14 kHz BW. Ultimate attenuation was about 86 dB. That is
outstanding. BTW, that means the close in phase noise (also called
composite noise by the ARRL) is less than -120 dBc/Hz. The only receivers I
have that will come close to or beat this are my R-390A's. The ultimate
filter attenuation and oscillators (all of the oscillators combined, not
just the 1st LO) phase noise blows the competition away, except for R-390A'
s. To be specific, it leaves the WJ-8711A and the NRD-545 (as well as the
non-DSP receivers like the NRD-525) in the dust. Whether or not this
translates into a better DX receiver remains to be seen.

More Impressions, June 5th

Filter measurements were made using a HP-8640B (with measured phase noise at
1.8 MHz of -145 dBc/Hz) from the noise floor of the IC-746Pro with AGC off.
There are three fixed AM filters, nominally 3, 6, and 9 kHz BW, and the
operator has no control over them (cannot set BW or shape). It took me two
tries to get all of these numbers right.

#1 (9 kHz nominal): 9.8 @ -6 dB, 15.1 @ -60, phase noise limited at -80 dB
#2 (6 kHz nominal): 6.8 @ - 6 dB, 12 @ -60 dB, 18 @ -80 dB, ultimate ~ -86
dB (close in)
#3 (3 kHz nominal): 3.4 @ -6 dB, 7.0 @ -60 dB, 10 @ -80 dB, ultimate ~ -86
dB (close in)
The ultimate selectivity value of -86 dB may be (probably is) phase noise in
which case the close in phase noise (5 to 10 kHz offset) is about -124
dBc/Hz (using the 6.67 BW). The shape factors of AM #1, #2, and #3 are
1.54:1, 1.76:1, and 2.06:1 respectively. I would have preferred a better
shape factor for #3, but #2 and #1 are certainly outstanding, and #3 is
excellent. I am curious why the #2 filter ultimate attenuation isn't several
dB less (because I believe this is phase noise). Perhaps this is due to the
filter algorithm, or inaccuracies in my equipment.

For SSB and CW the operator can select "sharp" or "soft" filters with SSB
bandwidths up to 3.6 kHz. For SSB I picked #1 as 3.6 and #2 as 2.8 and #3 as
2.4. I picked the wider filters for ECSS and the 2.4 as default SSB.

#1 (3.6 kHz sharp): 3.3 @ -6 dB, 4.5 @ -60 dB, 7.1 @ -80 dB, ultimate ~ -86
dB (close in)

#2 (2.8 kHz sharp): 2.9 @ -6 dB, 4.0 @ -60 dB, 6.2 @ -80 dB, ultimate ~ -86
dB (close in)

The shape factors of SSB #1 and #2 are about 1.34:1, outstanding. I did not
measure #3, but have no doubt that it is similarly outstanding. The soft
filters numbers were not as good as the sharp filters numbers (audio
response is not flat, but has an upward slope from low to high frequencies,
skirts are somewhat wider, and nose is somewhat narrower). I didn't care for
them.

I did some weak AM signal listening above 21.5 MHz this morning. AM
performance is just outstanding. The AM detector in the 746P is much better
than the AM detector in the WJ-8711A, and about the equal of the 8711A AMS
detector. I didn't have the WJ at hand for a direct comparison because I
was going against my standard, the R-390A. ECSS is also excellent. I
compared ECSS and AM modes on a number of weak signals above 21.5 MHz. As to
which was better, I suppose it depends on your taste in audio. I like the
generally better audio quality obtained with AM mode. Sometimes ECSS
delivered better audio when there was interference on one of the sidebands.
The comparison was somewhat problematic because of the BW differences (3.3
and 2.9 for ECSS and 6.8 and 3.4 for AM). There were no cases where audio
was recovered in ECSS but not AM and vice versa. Those who prefer ECSS won't
be disappointed. And it appears to be a dead heat between the R-390A and the
746P for AM reception whether you use AM mode or ECSS. And no (and I do mean
N. O. NO) uP crud from the 746P above 20 MHz (or anywhere else) like with
the WJ-8711A and some other high end (elsewhere called premium) receivers.
Days later (6/13) I measured AM sensitivity at 21.6 MHz using the 6.78 kHz
BW: Preamp 1 = 0.6 uV with NR control turned off, 0.5 uV with NR adjusted
for maximum digital noise reduction; Preamp 2 = 0.4 with NR off, 0.3 with NR
on. I don't believe you will need an external preamp for the 746P.

The 746 is not an intuitive receiver to operate, and there are a lot of
options that users can configure (audio bass and treble boost for one). It
will take more than a few hours for me to master this receiver.

Top Of The Heap, June 11th

Here are some things that put the 746P at the top of the heap. The AM
detector is an AM synchronous detector. Why ICOM doesn't advertise this
feature of the 746P is a mystery to me. I discovered it merely by noticing
that it sounded like an AM synchronous detector and asking ICOM Technical
Support if it was. They confirmed what my ears had already told me. And it
is not just any old AM synchronous detector. It is an outstanding AM
synchronous detector. It doesn't lose lock (no growling on extremely weak
signals fading in and out of the ambient noise floor) and you can tune the
signal with the AM carrier anywhere you please in the passband, and even out
of the passband, and still no growling. In other words, the 746P AMSD is
completely transparent to the user. You never know it is there except that
the quality of AM reception is better than with an ordinary AM detector for
some weak signals at the ambient noise floor and for some strongly fading
signals, and better than ECSS. And there is a receiver (and transmitter)
audio tone control which permits you to select up to 5 dB of bass or treble
boost or cut. In other words, you can customize the audio response to your
liking. It works great in AM mode. However, the 746P doesn't seem to like
full bass boost when using SSB for ECSS. The audio starts cutting off and on
as you approach zero beat. Very near zero beat the audio can stay off for
several seconds or longer. However, this problem has a useful application as
an aid to adjusting the calibration pot through the small hole on the rear
panel. Tune WWV 10 MHz (or whatever) with the frequency display set to
10.000000 (and RIT at 0) and adjust the pot slowly as the audio on and off
period gets longer and longer. The longer you can make this period the
closer you are to zero beat. The 1st LO oscillator phase noise (or more
generally the composite noise) at close separations (from 3 to 10 kHz) is
about -124 dBc/Hz. This is easily 10 dB better than any other solid state
receiver that I am familiar with. including such receivers as the NRD-525,
535, and 545, the RA6790/GM, 6793, and 6830, the (Harris) RF-590, and 590A,
the (Collins) 651S-1 and HF-2050, and the (Watkins Johnson) HF-1000(A) and
WJ-8711A. The ARRL published transmit (not receive, but presumably they are
the same as receive) composite noise measurements of a 746Pro which show
noise decreases from -120 to -130 dBc/Hz as the offset increases from 2 to
10 kHz at both 3.5 and 14 MHz, with noise declining -135 dBc/Hz at 22 kHz
offset. I measured the wide offset (500 kHz) phase noise (composite noise)
as -145 dBc/Hz with tone at 1.000 MHz and noise at 1.500 MHz. This is
outstanding phase noise performance. The filters ultimate attenuation is
greater than 80 dB, and even though there are only three AM filters, their
bandwidths are well chose with excellent shape factors. Then there is the NR
(noise reduction) knob. Not a noise blanker, but digital noise reduction.
And it works very well, especially on weak signals in the presence of noise
in AM (synchronous detection) mode. It even reduces external preamp noise,
thereby improving the overall sensitivity when an external preamp is used
with the 746P. In one case adding a low noise preamp improved MW band
sensitivity only to 0.50 uV (from 0.60 uV). But turning on the NR feature
and adjusting the NR knob for optimal sensitivity improved it to 0.30 uV. At
noisy locations this won't matter. At quiet locations it will. There is also
a noise blanker which works very well. Users can adjust the amount blanking
provided by the noise blanker. And there is no (not even a hint of)
microprocessor noise (which plagues the WJ-8711A and many other top
receivers). Taking all of the above into consideration, this is about as
good as it gets.

Down Sides

The down sides are that the ICOM put a 10 or 11 dB attenuator in the signal
path below 1.6 MHz and the preamps are disabled below 1.6 MHz. Furthermore,
Preamp 1 gain rolls off fast below 500 kHz. And it is non-trivial to fix
these things because of the many tiny 0603 SMD's and tiny PC board traces.
But they are fixable, as I will describe below.

Sensitivity Measurement Issues

Perhaps it is just my IC-746Pro, but there seem to be some anomalies when
measuring AM sensitivity. In the 6 kHz (nominal) BW the AM sensitivity is a
tenth of a microvolt or two better when the receiver is detuned to the high
side. Also, the digital noise reduction (NR knob position) can be used to
improve the AM sensitivity by a few tenths of a microvolt. So the question
arises, "What is the AM sensitivity of the IC-746Pro in the 6 kHz BW?" Well,
it seems to vary by as much as much as 0.4 microvolts depending on the
position of the controls. The tone TCN (tone control) menu setting also
affects sensitivity values. And if 1000 Hz signal generator modulation is
used instead of the standard 400 Hz, measured sensitivity values are worse.
Sensitivity values given below were made with the NR (noise reduction)
turned off (I think), and with 400 Hz modulation. I have no idea where the
carrier was in the filter passband because I did not know about detuning
effect when those measurements were taken. So presumably those values could
be as much as 0.2 or 0.3 microvolts high from the best sensitivity.

Intercepts

The ARRL gave measured values for the intercepts in their long review which
for the most part I will not reproduce here. The 3rd order intercept with
Preamp 1 on is about +3 dBm with close spacing, and the 2nd order about +70
dBm above 1.6 MHz. Below 1.6 MHz (not measured by the ARRL), with the 11 dB
attenuator removed and Preamp 1 enabled (to be described in detail below)
the 2nd order intercept is typically +35 dBm when both tones are in the MW
and/or LW band(s) and +85 dBm when both tones are in the HF bands. This is
because a simple elliptic low pass filter with cutoff frequency somewhat
above 1.6 MHz and attenuation in excess of 40 dB at 3 MHz and above is used
when tuning below 1.6 MHz. A low pass filter may be needed for the LW band,
and some kind of tracked tuned circuit may be needed for the MW band for "in
band" intermod free reception. The 746P 2nd order performance above 1.6 MHz
is much better because it uses switched bandpass filters above 1.6 MHz.
However, the filters could obviously be better. So even in the HF bands
additional front end filtering may be wanted in some cases. Wide spaced (out
of band) 3rd order intercepts generally vary in the +38 to +40 dBm range for
the HF bands and about +44 dBm in the MW and LW bands. Overall I expect my
modified (attenuator below 1.6 MHz removed and Preamp 1 enabled) 746P to be
an excellent performer with regard to intermod except, perhaps, for in band
MW and LW intermod, which can be fixed by various filters. I have not
experienced any LW or MW intermod, but then I am not near any high power MW
transmitters. For LW it is straightforward to build a good 450 kHz cutoff
low pass filter, or you can buy one from Kiwa Electronics. If MW intermod is
experienced, an appropriate MW filter can be built or perhaps bought.




-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----