Andy,
Thanks for your comments, you are always insightful. For the record, I made
no comments about a "setup filter" I was merely forwarding an article by Dallas
Lankford. That said, full A/D overload of -23dBm and a dynamic range of 80dB
may be adequate for low signal environments (I'm told that big MW stations are
rare in the U.K.) but would fare poorly at this QTH which has FIVE 50kW MW
stations with 15kM!
Back to the AGC issues, Rob Sherwood of Sherwood Engineering has published a
slide presentation on how poorly ALL DSP receivers perform under QRN conditions
due to the DSP handling of AGC:
http://www.sherweng.com/documents/Dayton2007w.pdf
The new software defined radios offer unique spectrum displays and record
possibilities but the basic radio functions are not up to the specs of a
high-end analog receiver yet though I'm sure that day is coming.
--
73 Warren K2ORS
WD2XGJ 136-140kHz
WD2XSH/23 505-510kHz
WE2XEB/2 160-189kHz
WE2XGR/1 505-515kHz
FN42hi
http://www.w4dex.com/wd2xgj.htm
-------------- Original message ----------------------
From: "Andy Talbot" <[email protected]>
> Warren's comments about "setup filter" are not correct What he is
> actually talking about is the down sampling / decimation process which takes
> place in the AD6620 Digital Downconverter chip. (In fact, he actually
> alludes to this right at the end). It is possible to achieve overload in
> the decimation process even though the arw A/D data is still within its
> linear range. However, this 'second stage' overload is not real, and is
> caused by an overflow in the maths. It can be prevented by adjustment of
> the AD6620 gain without any resulting reduction in sensitivity.
>
> I have measured the overload performance of my SDR-IQ and with no preamp in
> place full scale A/D overload occurs at about -23dBm which is consistent
> with his measured figure.
> The dynamic range of approximately 80dB between maximum A/D overload level
> and quantisation noise / A/D spec is just about adequate given that the
> SDR-IQ uses a few block filters on its input which are switched in
> automatically depending on the frequency selected. He makes several
> observations about strong local signals and lightning strikes which may be
> valid for his special circumstances, but in general I think his measurements
> and his thoughts are rather negative about this receiver.
>
> I believe the Perseus has a similar front end topology, but has a
> greater selection of front end filtering available. This is clearly better,
> but you get what you pay for - the Perseus costs twice as much as an SDR-IQ,
> you get better input filtering and get a larger instantaneous bandwidth,
> 400kHz rather than 190kHz.
>
> I've used my SDR-IQ to listen on 137 and 500kHz, and much prefer it to the
> RA1792 that was once used regularly on LF. There was never any sign of
> overloading on the 'IQ, using both an untuned passive 10 metre vertical whip
> or the tuned 137kHz Tee. Even with the input gain set at +10dB. For
> a medium cost SDR, its difficult to beat
>
> However, for* purely LF* use, if I wanted a dedicated receiver and hadn't
> already got an SDR-IQ, I'd probably look at a customised Softrock type.
> With a decent high order Low Pass filter at 510kHz to severely cut off
> medium wave signals (difficult, as the band starts at 540kHz, but a 9th or
> 11th order LPF isn't difficult at this frequency) followed by a decent
> bus-switch based mixer into a high quality soundcard. Depending on where
> you want to cover, one or two switched crystals can form the LO and you'll
> get more than full band coverage of 500kHz or 137kHz. The high Q antenns
> used by most stations will go some way to killing off the very strong low
> frequency signals. In fact as a quick test, a Softrock, with suitably
> rewound coils, an external filter and a new LO crystal would probably do
> admirably.
>
> Andy G4JNT
>
>
>
> On 24/01/2008, K2ORS <[email protected]> wrote:
> >
> > Dallas Lankford has just done an analysis on the SDR-IQ and found the
> > strong signal performance to be wanting:
> >
> > http://www.kongsfjord.no/dl/SDR/SDR-IQ%20Oveload%20Characteristics.pdf
> >
> > 73 Warren K2ORS/WD2XGJ
> >
> >
> > ----- Original Message -----
> > *From:* Peter March <[email protected]>
> > *To:* [email protected]
> > *Sent:* Thursday, January 24, 2008 5:10 AM
> > *Subject:* Re: LF: Re. LF Receivers
> >
> >
> > Hi Laurence
> > Whilst I don't have experience the Perseus SDR RX, I do with Softrock,
> > same principle but simpler execution at radio frequency (and much much
> > cheaper).
> >
> > You can at zero cost down load SDR software to play with. You can also
> > down load recorded examples of on air activity. This gives you the ability
> > to tune all over the band listen to various qso's in SSB or CW as if you had
> > the SDR RF section feeding your computer.
> >
> > This technology is quite unique and very worth while the effort to try
> > out.
> >
> > Alberto is quite right about the lack of a nice dial knob to turn but if
> > you are wedded to a nice big dial knob you can get a custom encoder that
> > does just that... at a cost though!!
> >
> > One out of many advantages is you can see the signals as a panoramic
> > display or as a waterfall you vary the water fall speed and read the qrss CW
> > ..... The Rocky software has built in PSK decoder and as I understand it
> > other decoder features are to follow.
> >
> > So long as the SDR software you are using has an HP9TLK.DLL file attached
> > and similarly the decoder software then you have a Virtual Audio Cable
> > connectivity i.e. no need for external wire or additional sound cards.
> >
> > The on board sound cards on most PCs are adequate if you are using a IQ
> > connection from your RF stage to the PC i.e. the mic input. It is worth
> > fitting a higher quality card see M0KGK's web site for an explanation.
> >
> > One aspect of SDR is this the ability of recording the whole band for
> > later interrogation even if you had not tuned to that frequency during the
> > original recording.
> >
> > The subject is too vast to detail it here but for a very small out lay
> > you can experience it first hand. Also the software is advancing where all
> > the signal prossing basically takes place so without change to the RF
> > section your radio can be up dated at no cost as this advance takes place!
> >
> > Follow the Softrock-40 Interest Group
> >
> <http://groups.yahoo.com/group/softrock40;_ylc=X3oDMTJlaGtkanNvBF9TAzk3MzU5NzE1B
> GdycElkAzE2MTEwMTczBGdycHNwSWQDMTcwODM5NjM3OQRzZWMDaGRyBHNsawNocGgEc3RpbWUDMTIwM
> TEwNTIwNA-->for
> > up to date information.
> >
> > I hope this helps in your search for another RX ...... Alberto's comments
> > are very valid and the beauty is you can try it out at zero cost.
> >
> > I have no commercial interest in SDR I am just a poor retired HAM in every
> > sense of the word with a Bee in his bonnet ... hi.
> >
> > Try http://www.flex-radio.com/ but ignore all the expensive toys, just the
> > free software.
> >
> > 73 petefmt
> >
> >
> >
> >
> > *lawrence mayhead <[email protected]>* wrote:
> >
> > Well I managed to fix my RA1792 receiver at last!
> > But all this investigation has whetted my appetite for something better.
> > At the moment I am torn between the AOR7030, conventional non DSP
> > a bit difficult to drive but could be OK with computer display, spec.
> > looks
> > good for LF. Or a take drive in unknown country with the Perseus SDR.
> > I really am in the dark here and would like to know wether is is
> > neccessesary
> > to re-digitise the audio output in order to get an ARGO FFT display or
> > does it
> > do this itself? and if so can I get 60/120 sec.dot resolution?
> > Any help gratefully received.
> > 73 Laurie
> >
> >
> > ------------------------------
> > Sent from
> Yahoo!<http://us.rd.yahoo.com/mailuk/taglines/isp/control/*http://us.rd.yahoo.co
> m/evt=51949/*http://uk.docs.yahoo.com/mail/winter07.html>- a smarter inbox.
> >
> >
>
>
> --
> Andy G4JNT
> www.scrbg.org/g4jnt
--- Begin Message ---
Warren's comments about "setup filter" are not correct What he is actually talking about is the down sampling / decimation process which takes place in the AD6620 Digital Downconverter chip. (In fact, he actually alludes to this right at the end). It is possible to achieve overload in the decimation process even though the arw A/D data is still within its linear range. However, this 'second stage' overload is not real, and is caused by an overflow in the maths. It can be prevented by adjustment of the AD6620 gain without any resulting reduction in sensitivity.
I have measured the overload performance of my SDR-IQ and with no preamp in place full scale A/D overload occurs at about -23dBm which is consistent with his measured figure.
The dynamic range of approximately 80dB between maximum A/D overload level and quantisation noise / A/D spec is just about adequate given that the SDR-IQ uses a few block filters on its input which are switched in automatically depending on the frequency selected. He makes several observations about strong local signals and lightning strikes which may be valid for his special circumstances, but in general I think his measurements and his thoughts are rather negative about this receiver.
I believe the Perseus has a similar front end topology, but has a greater selection of front end filtering available. This is clearly better, but you get what you pay for - the Perseus costs twice as much as an SDR-IQ, you get better input filtering and get a larger instantaneous bandwidth, 400kHz rather than 190kHz.
I've used my SDR-IQ to listen on 137 and 500kHz, and much prefer it to the RA1792 that was once used regularly on LF. There was never any sign of overloading on the 'IQ, using both an untuned passive 10 metre vertical whip or the tuned 137kHz Tee. Even with the input gain set at +10dB. For a medium cost SDR, its difficult to beat
However, for purely LF use, if I wanted a dedicated receiver and hadn't already got an SDR-IQ, I'd probably look at a customised Softrock type. With a decent high order Low Pass filter at 510kHz to severely cut off medium wave signals (difficult, as the band starts at 540kHz, but a 9th or 11th order LPF isn't difficult at this frequency) followed by a decent bus-switch based mixer into a high quality soundcard. Depending on where you want to cover, one or two switched crystals can form the LO and you'll get more than full band coverage of 500kHz or 137kHz. The high Q antenns used by most stations will go some way to killing off the very strong low frequency signals. In fact as a quick test, a Softrock, with suitably rewound coils, an external filter and a new LO crystal would probably do admirably.
Andy G4JNT
On 24/01/2008, K2ORS <[email protected]> wrote:
Dallas Lankford has just done an analysis on the SDR-IQ and found the strong signal performance to be wanting:
73 Warren K2ORS/WD2XGJ
----- Original Message -----
Sent: Thursday, January 24, 2008 5:10 AM
Subject: Re: LF: Re. LF Receivers
Hi Laurence
Whilst I don't have experience the Perseus SDR RX, I do with Softrock, same principle but simpler execution at radio frequency (and much much cheaper).
You can at zero cost down load SDR software to play with. You can also down load recorded examples of on air activity. This gives you the ability to tune all over the band listen to various qso's in SSB or CW as if you had the SDR RF section feeding your computer.
This technology is quite unique and very worth while the effort to try out.
Alberto is quite right about the lack of a nice dial knob to turn but if you are wedded to a nice big dial knob you can get a custom encoder that does just that... at a cost though!!
One out of many advantages is you can see the signals as a panoramic display or as a waterfall you vary the water fall speed and read the qrss CW ..... The Rocky software has built in PSK decoder and as I understand it other decoder features are to follow.
So long as the SDR software you are using has an HP9TLK.DLL file attached and similarly the decoder software then you have a Virtual Audio Cable connectivity i.e. no need for external wire or additional sound cards.
The on board sound cards on most PCs are adequate if you are using a IQ connection from your RF stage to the PC i.e. the mic input. It is worth fitting a higher quality card see M0KGK's web site for an explanation.
One aspect of SDR is this the ability of recording the whole band for later interrogation even if you had not tuned to that frequency during the original recording.
The subject is too vast to detail it here but for a very small out lay you can experience it first hand. Also the software is advancing where all the signal prossing basically takes place so without change to the RF section your radio can be up dated at no cost as this advance takes place!
I hope this helps in your search for another RX ...... Alberto's comments are very valid and the beauty is you can try it out at zero cost.
I have no commercial interest in SDR I am just a poor retired HAM in every sense of the word with a Bee in his bonnet ... hi.
73 petefmt
Well I managed to fix my RA1792 receiver at last!
But all this investigation has whetted my appetite for something better.
At the moment I am torn between the AOR7030, conventional non DSP
a bit difficult to drive but could be OK with computer display, spec. looks
good for LF. Or a take drive in unknown country with the Perseus SDR.
I really am in the dark here and would like to know wether is is neccessesary
to re-digitise the audio output in order to get an ARGO FFT display or does it
do this itself? and if so can I get 60/120 sec.dot resolution?
Any help gratefully received.
73 Laurie
Sent from
Yahoo! - a smarter inbox.
--
Andy G4JNT
www.scrbg.org/g4jnt
--- End Message ---
|
