Sinisa,
Thanks for this data. Among other things, it confirms that "100 Hz"
BW isn't exactly knife-edge filtering -- as you'd expect from 199 taps
and 14 kHz IF, I suppose. A number that might also be useful would be
the -10 dB or -20 dB filter widths. Response inside the "bandpass"
(+/-3 dB?) isn't so important as the rejection of out of band signals.
Depending on the scenario, 10 dB rejection can be critical. I often
run into bandshape limitations on 20M PSK31, for example, where it's
all quite visible on the waterfall -- if uncalibrated.
It would be interesting to compare the sidelobes at 200 Hz BW.
This info must be in somebody's lab notebook at TT. (We hope! :-)
Wouldn't it be nice if they'd find a way to share it? They might say
it's proprietary, but full technical disclosure would be a very nice
selling point for the company. (Why do they bother giving those
exquisite equalizer curves in the manual and on the website? IF
DSP/filter response would be a better choice!)
73 Martin AA6E
On 3/11/06, Sinisa Hristov <shristov@ptt.yu> wrote:
> The following 100 Hz DSP CW filter response was measured
> with a calibrated signal fed directly to ADC input on A7 board.
>
> 156 Hz width at -6 dB
> 378 Hz width at -40 dB (sidelobe peak)
> 512 Hz width at -49 dB (sidelobe peak)
> 660 Hz width at -56 dB (sidelobe peak)
> 796 Hz width at -62 dB (sidelobe peak)
> 463 Hz width at -67 dB (sidelobe peak)
> 1072 Hz width at -70 dB (sidelobe peak)
> 1210 Hz width at -75 dB (sidelobe peak)
>
> Both S meter and AGC respond happily to sidelobes as measured.
> The response is essentially the same as with V1 firmware.
>
> V2 apparently introduces an additional audio filter preventing
> sidelobes from being heard, which may be useful in some circumstances,
> although many would not call it "real" selectivity as the AGC is
> still being keyed by strong signals more than half a kHz away.
>
> Selectivity measurements by spectral analysis of the audio response
> to white noise will show combined selectivity of IF and audio filters.
>
> I wouldn't consider this a bug.
> It is an "unpublished feature", although somewhat misleading one.
>
>
> 73,
>
> Sinisa YT1NT, VE3EA
>
> _______________________________________________
> TenTec mailing list
> TenTec@contesting.com
> http://lists.contesting.com/mailman/listinfo/tentec
>
--
martin.ewing@gmail.com
http://blog.aa6e.net
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|