Re: New overview ARRL/QST Product Reviews
- Hello Nico,
Can you please be more specific in which way I have to correct the overview?
My overview has been made according to the published information in QST, not private e-mails between persons.
Thanks and 73, Hans.
--- In perseus_SDR@yahoogroups.com, "iv3nwv" <nicopal@...> wrote:
> --- In perseus_SDR@yahoogroups.com, Hans Remeeus <hans@> wrote:
> > Hello,
> > Today I made a new layout of the ARRL/QST Product Reviews overview:
> > http://www.remeeus.eu/hamradio/pa1hr/productreview.htm
> > I hope you like it.
> No, I don't like it at all.
> The test made on Perseus and some of the numbers published by QST were simply wrong and I'm really upset that the wrong figures are still advertised three years after it has been admitted by the ARRL itself that they were wrong.
> For the sake of truth and for everybody knowledge, here attached there's the reply I've received by Bob Allison, WB1GCM, ARRL test engineer, after that Leif SM5BSZ complained with ARRL that the figures published on QST regarding the so called "blockin gain compression" on Perseus were pure science fiction.
> Please correct your table.
> Nico / IV3NWV
> From: Allison, Bob, WB1GCM
> Sent: Friday, November 07, 2008 1:40 PM
> To: Hare, Ed W1RFI
> Cc: Allison, Bob, WB1GCM; Allison, Katherine, KA1RWY
> Subject: Perseus Letter
> Dear Lief,
> Thank-you very much for your technical measurments on the
> Perseus SDR Receiver, especially the Blocking Gain Compression.
> Fortunately, we still had the Perseus in the Lab. After finishing up
> measurements on another device, I set up the Perseus for BGC in the
> screen room.
> The original data could not be matched by the measurements I
> made yesterday. The way we measure BGC is by mixing the desired signal
> with an increasing adjacent signal and look for a 1 dB decrease in audio
> level of the desired signal, using a signal analyzer and narrow
> bandwidths. I found that no gain compression took place until about .5
> dB past the point where the ADC clip display came on, at about +2.5 dBm.
> This makes the BGC DR to be nearly 129 dB, not just at 20 kHz away, but
> also 5 and 2 kHz away. These measurements were made with the AGC off,
> the preset on and preamp off on 14.020 MHz, MDS = 126 dBm, BW 503 Hz.
> The only deviation to this data occurred with the AGC on.
> Where BGC DR was
> 123, 119 and 115 dB (20, 5, 2 kHz). Perhaps this contributed to the
> previous measurements, along with any unknown problem. I was thinking
> the sound card had some sort of compression when noise was introduced,
> but that doesn't matter here. Our measurements were off by a
> considerable amount. I am sure we will publish a feedback in QST about
> this. At least, it is in favor of the manufacturer !
> Lief, I will ask for your help in this matter by asking the
> question: how do we test SDR type radios ? Can you think of ways
> different from conventional receivers, such as testing with the AGC on ?
> Who in the world would listen in real life with the AGC off with an SDR
> radio ? I ask for input from you and all others on the product review
> testing list.
> Thank-you again !
> Bob Allison
> ARRL Test Engineer
- Hi Leif,
I agree that my two tone IMD2 example was wrong. As long as I test with nearly perfectly periodical signals (thus with a very small bandwidth) there is always enough oversampling for all spectrum components.
> I do not understand the theory behind this statement,I just wanted to express that NL reduction by dithering needs a high ratio of the dithering bandwidth (finally limited by the sample rate) to the signal bandwidth (RX bandwidth). While a linearization scheme working with postdistortion based on a physical model of the ADC can be inherently wideband.
> (or I do not understand the statement.) More relevant
> is that I find something else when actually doing the
But I agree that this has no relevance for shortwave reception of narrowband signals.