Topband
[Top] [All Lists]

Re: Topband: SDR comment

To: <topband@contesting.com>,"Lee A Crocker (by way of Bill Tippett<btippett@alum.mit.edu>)" <lee_crocker@yahoo.com>
Subject: Re: Topband: SDR comment
From: "Tom Rauch" <w8ji@contesting.com>
Reply-to: Tom Rauch <w8ji@contesting.com>
Date: Tue, 13 Feb 2007 19:17:14 -0500
List-post: <mailto:topband@contesting.com>
Hi Lee,

Don't get me wrong, the SDR is a nice system but not 
something that set the world on fire in a DX'ing or Contest 
environment. Like any system it has its pro and cons.
Our goal is always to have an edge in competitive 
environments, be it DX'ing or Contesting. I think an SDR 
would be super for VHF or UHF work or even regular 
operation, but it is not a competition grade system as a 
primary low band contest radio.

We also hoped, based on a claim by someone else, it would 
help find "fresh meat" in the contest. It can serve a 
function in that regard if muted while the main transmitter 
is on **if confined to a small clear area of the band** 
(like the DX Window), but any good quality spectrum analyzer 
with adequate functions will do the same.

> Thanks for your analysis.  One thing that is different
> for the SDR-1000 on 160 compared to 80M is that the
> filter on the front end is low pass on 160 instead of
> band pass so the RX can be used down below 100khz.
> This does effect RX performance on 160 since it does
> open the front end to more RF than would otherwise be
> the case.

The SDR, like all receivers or preamps here do, enjoyed the 
benefit of having high-pass or band-pass filters that notch 
out-of-band signals. As a matter of fact filter rolloff 
starts at 1.8MHz and is around -20 dB by the time the upper 
end of the AM BCB is reached (actual measurement, NOT a 
model).

>Also in my experience is that there is a
> significant difference in the strong signal handling
> capability between the 192 kHz sampling rate and 48khz
> sampling rate on different sound card.  With a D-44
> sound card at 48khz sample I can DX on 80M CW, copying
> s-3 signals while running 1500W of SSB up around 3770.

I'm not sure what sound card we had, but I was told it was 
the recommended card for maximum performance. Our results 
had nothing to do with filtering, and the dynamic range 
required varies greatly from band to band and with location.

In a very quiet rural location noise floor is lower, and 
significantly more dynamic range is required. This is 
especially true when the band is loaded wall-to-wall with 
strong signals and there is no roofing filter to eliminate 
signals 5 kHz or more away.

The bottom line however is even with two signal generators 
and no other signals, the dynamics at wide spacing are far 
below what an off-the-shelf radio can deliver. The Orion for 
example is around the 100dB IM3  range at only 20kHz spacing 
and has a blocking DR of over 130dB. We chose noise blanker 
corrected FT1000 MK V's because of operator convenience but 
they still have a wide spaced blocking DR over 135dB. The 
primary receivers the second night (we didn't have them on 
the first night) were as good as or better than the Orion. 
We can't use a receiver with a blocking DR of only 100 dB.

The intention also was to use the rig for finding new 
signals, so the display would have to be perfectly clean 
with the transmitter on. That just wasn't possible.

> Part of the "problem" in how the SDR works is that all
> the power in the sampling bandwidth is given equal
> consideration.

....and that's one thing that will make an SDR radio seem so 
good when there aren't multiple  strong signals around. 
Another is when the background noise floor is higher dynamic 
range requirements are reduced even further.

This is a case where people in certain environments will 
swear by them and people in other environments will swear at 
them.

The best of both worlds would be an SDR following a roofing 
filter so the wide band system only has to deal with a few 
dozen strong signals at a time.

> point this out explicitly.  You can't really expect a
> radio that is designed to look at 192khz of spectrum
> to act like an Orion II which is designed to look at
> as low as a 300hz of spectrum.  The radio's behavior
> therefore is not so much of a "problem" as it is a
> consequence of the design philosophy.

I fairly and clearly pointed that all out. If I lived in 
place with 30dB more background noise or a few thousand less 
30 over nine signals and didn't try to duplex the SDR would 
be a great radio. Like all radios it isn't magic, and claims 
a mixer can somehow reduce the dominant ambient noise we all 
deal with on 160  through 20 meters are without merit.

> the design trade offs.  The only way you are going to
> know the nature of packet pileup with a Yaesu is to
> constantly tune and aurally analyze the nature of the
> pileup.  The SDR actually does a very good job of
> allowing you to "see" a pileup.

Not when the band is loaded with hundreds of changing QSO's 
spaced neatly every 500Hz with all operators in QSO's nearly 
perfectly zero beat with others in the same contact, and the 
levels are constantly changing. There isn't a radio in the 
world that can do that. Been there, done that, and looked 
first hand.

It can however, like a FFT spectrum analyzer, indicate a 
chance of fresh meat in a clear area of the band or an area 
with light loading.

I can use my FFT analyzer to watch 1830-1835 during a 
contest and see a new pileup start. It appears as a new pip 
in an otherwise empty slot.
It would also be excellent for VHF and UHF contests, 
allowing a person to jump across unused portions of the 
band...and it would be good for finding a hole to CQ in on a 
crowded HF contest.

If I claim it helped me find North Dakota in a matter of 
seconds in the middle of the dancing mess of hundreds of 
stations exchanging 20-second QSO's with people of widely 
varying strength while I was running people down the band I 
would clearly be a fibber. Only packet can do that.

> My experience using the SDR-1000 as a main RX on a
> recent 160M contest made me draw a little different
> conclusion.  I spent part of the contest trying to
> find the worst case operating situation I could.  I
> found 2 DX stations in different parts of the world,
> one in Europe and the other in the Caribbean.  The
> stations were 40 hz apart and both were about S1 to S3
> varying with QSB.  They were both being called by US
> and European hams, with strong signals.  With the SDR
> I was able to tease out and work each station as a
> single signal while they were being called by 6 to 10
> strong US stations.

Which I freely pointed out is possible, and I also observed. 
My Agilent FFT Spectrum Analyzer can do the same.

Unfortunately it doesn't work well in crowded areas where 
rapid exchanges are going on, especially when viewing a wide 
area of the band. The screen would have to be the size of a 
desk to see signals a few Hz apart when looking at a 50kHz 
window, and it would not work while the transmitter is 
running.

It isn't a bad system, it just doesn't do what some might 
claim.

73 Tom 


_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband

<Prev in Thread] Current Thread [Next in Thread>