[ICOM] PRO vs PRO2

[Adam Farson]

Hi Dave,

Well - going on my very favourable experience with DSP-based radio
architecture over the past 6 years, I would go for the IC-7000 over the
IC-706 (in any event,  once S/N > 2000.) There is a fair degree of
confidence that the IC-7000 will be a big improvement over the IC-706
series. 

There are two schools of thought in DSP architecture. One approach is to run
the ADC at a fixed IF, and limit the sampling bandwidth at the analogue
input to the ADC by means of roofing filters following the mixers. The
roofing filter bandwidth is a trade-off between (1) reducing the statistical
likelihood that strong signals outside the DSP-IF bandpass will "ride
through" and overload the ADC, and (2) increased insertion loss, poorer
phase response and degraded temperature stability in the roofing filter
itself. The last is especially true of roofing filters cut for a high 1st IF
(e.g. 64 MHz) as required in a continuous-coverage system.

The roofing filter needs to be placed at the IF output of the first mixer.
Placing costly multi-pole crystal or mechanical filters in the low IF ahead
of the ADC will not improve overall system performance, as a
properly-designed roofing filter after the first mixer will already have
determined the ADC sampling bandwidth - and that is what counts. From then
on, the DSP can emulate all those "sweet-sounding" IF filters if desired.
Rockwell-Collins no longer use Collins filters for selectivity; all their
current HF radio systems are DSP-based. 

There is really no need to expend R&D resources "mixing" analogue and DSP
filters in the IF signal path, as the DSP designer can cook up a filter to
do whatever the end-user wants. By manipulating the number of taps (poles),
frequency and phase response, and group delay, the guy writing the DSP code
can emulate anything from a bell-shaped Collins filter to a rectangular
passband, within reason. Doubtless, he will design the filter for optimum
communications effectiveness. 

The other approach is to run the ADC at RF, and have it sample the entire HF
range. The disadvantage of this is that the composite power of multiple
strong signals in the sampling bandwidth will eat up all the system headroom
and overload the ADC. (A good AGC, derived from the DSP, can partially
offset this problem.)

http://www.qsl.net/ab4oj/icom/ic756pro_notes.html#rf_if

The DSP designer can also tailor the baseband emerging from the DAC to
"sound" the way the customer wishes, but I suspect that HF radio
communications systems will stick pretty closely to the ITU-R guideline (350
Hz to 2.7 kHz at the -6 dB points). That is what sells vast quantities of
HF-SSB radio gear to all manner of purchasers.

Running the ADC at 400 kHz or so will eliminate one down-conversion in a
radio such as the IC-756Pro series. But then again, the IC-7800
down-converts from 64 MHz to 36 kHz in one step, using an I/Q second mixer.

As I have commented in previous posts, continuous coverage (dictating the
high 1st IF) is not a marketing nicety. It is an RFP requirement for all
military, governmental and commercial HF systems. A radio communications
manufacturer must be a player in these sectors if it desires to make a
profit and survive; we amateurs are co-beneficiaries of this business.

Cheers for now, 73,
Adam VA7OJ/AB4OJ


-----Original Message-----
From: icom-bounces at mailman.qth.net [mailto:icom-bounces at mailman.qth.net] On
Behalf Of David J. Ring, Jr.
Sent: 20 August 2005 00:02
To: ICOM Reflector
Subject: Re: [ICOM] PRO vs PRO2

Thanks Adam, I hadn't noticed that yet as my material on the 7000 is sparse.

But I would it would be a good competitor for the 706 -

How about you, Adam, if it were EVEN money, would you buy the 706MK2G or the
7000?

An interesting question - I actually wouldn't jump at the 7000 because of
how good the '706 is.

Of course it brings up the question - where is the balance between analog
and digital?

When technology is able to bring DSP i.f. up to the 400 kHz range, it would
be VERY interesting to mix mechanical / crystal filters with DSP "backbone" 
of ultimate rejection.

Would the "natural" shape factor of mechanical or crystal filters still
"sound" better then?  I like Collins filters - they have a beautiful sound -
but at the bandpass ends, they're obviously not as quiet as DSP - they have
hiss leakage, etc.

But imagine a flat DSP bandpass with brick wall AFTER that Collins filter?

Awesome, I'd think.

73

David N1EA



----- Original Message -----
From: "Adam Farson" <farson at shaw.ca>
To: "'ICOM Reflector'" <icom at mailman.qth.net>
Sent: Friday, August 19, 2005 9:19 PM
Subject: RE: [ICOM] PRO vs PRO2


Hi Dave,

The DSP in the 7000 is slower than in the 756Pro series or in the
746Pro/7400. The final (DSP) IF is 16 vs. 36 kHz.

What I have seen thus far is that the filter shape factors are not quite as
good; SSB 2.4 kHz is 1.625 vs 1.50.

Cheers for now, 73,
Adam VA7OJ/AB4OJ


-----Original Message-----
From: icom-bounces at mailman.qth.net [mailto:icom-bounces at mailman.qth.net] On
Behalf Of David J. Ring, Jr.
Sent: 19 August 2005 13:27
To: ICOM Reflector
Subject: Re: [ICOM] PRO vs PRO2

The very sharp 50 Hz filtering is better in the PRO2 than in the PRO.

But there are several reviews around that say the improvements to the PRO2's
dsp chain were mostly incorporated into the IC-746PRO.

I'd go for a 746PRO or I'd wait for the new IC-7000 coming out and go for
that.  More bang for the buck.

73

DR


Scanned by WinProxy
http://www.Ositis.com/