I am a co-author of the dsp software in the SDR-1000 (with AB2KT). When
I began working with the SDR-1000 as a customer, it had less than
stellar performance. Along with many others, repeated study was done
on the front end of the first SDR-1000 and the SDR-1000 plus RFE was
born which was essentially a new front end. A few component changes
later and the current front end was born. With the exception of the
sensitivity in the receiver measurements given here in a recent note,
there is not one other number in that article that I agree with. The
sensitivity is agreed to only in the case of the preamp being set to
MEDIUM.
I am a professional in the communications field and do consulting work
for DOD. In my work I have repeatedly measured and analyzed this
radio. It became of interest to me as a tool to develop software as it
did to my partner AB2KT but with help from me and others, it became
interesting as a receiver with modifications that have since been
incorporated into the transceiver. I have derived a mathematical
description of the working of the QSD or Tayloe detector and it has
allowed as to how this front end may be improved even more. I
understand this will be in future products from Flex and was developed
by simulation by Flex before I did my mathematical analysis of it. For
this reason, I have not published my results in deference to my
friendship for Gerald.
In EVERY single case where I have seen numbers similar to those written
up and distributed here, the radio was set up incorrectly or there was
some other problem. Working with one of the good suggested sound
cards, the radio has consistently, in EVERY SINGLE CASE, measured 100
dB IMD dynamic range with an IP3 hovering around 30 dBm. Let's call it
28 dBm as the ARRL did in their lab tests. It is VERY important to know
when measuring this radio that you have the RFE board and all of the
component changes have been made to it. Those are incorporated in all
current models.
The big difference between this receiver and any other receiver I have
EVER measured is that these same numbers are true at 20 kHz, 5 kHz,
1KHz, and lower. If I can find a fault with the SDR-1000, even given
its "near prototype configuration" it is that it still takes some work
to set up, learn, and to grow comfortable using. In the early days it
had a deserved reputation as a dog on CW because of keying latency.
This has been completely and totally resolved by software upgrades. If
you in Visalia or at Dayton, I suggest you give it a whirl to see. I
am regularly on 40m CW and 160m listening. My dial is set to 35 wpm for
my 40m rag chews and it is effortless using the internal keying hook up.
I am not an employee of Flex and I receive no compensationof any kind
for the software I have done since it has been released GPL. I am an
employee of the Institute for Defense Analyses and work at our branch in
Princeton, NJ. Outside of believing that this approach to transceivers
will be the only one on your desk in new radios in a few years, I have
no axe to grind here outside of the truth. On the Orion II, there is
no way it has the same performance as the SDR-1000 inside a few kHz. At
500 Hz, even with the roofing filter kicked in, the SDR-1000
significantly outdistances the Orion because of the insertion loss of
whatever it is Orion has done in that particular setting. The dual
track agc with anticipation was done by me based on a suggestion from
Phil Harman, VK6APH. I add the multi time constant anticipation and did
the software implementaiton for the SDR-1000 before the Orion II hit the
streets . It has been given the sincerest form of flattery in the
Orion II so you are welcome if you have one. The nicest part about the
agc, the filtering, etc, is that if we don't like it, in response to
customer requests, Flex can easily change it by modifying a few lines
of publicly available and viewable code. Welcome to the new world.
Dr. Robert W. McGwier
N4HY
--
AMSAT Director and VP Engineering. Member: ARRL, AMSAT-DL,
TAPR, Packrats, NJQRP, QRP ARCI, QCWA, FRC. ARRL SDR WG Chair
"Taking fun as simply fun and earnestness in earnest shows
how thoroughly thou none of the two discernest." - Piet Hine
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|