The problem with measuring the shape factor is that it can be changed significantly by applying a different window in the filter design code. The measured shape is a snapshot of one possible configuration. It's a tunable parameter, however, even if it hasn't been brought out to the "front panel" yet.
73 Frank AB2KT On 4/28/07, Tim Ellison <[EMAIL PROTECTED]> wrote: > > I have a grasp on the math, I an need some "real" numbers (measured) to > fill in a product comparison of the Flex-500 vs.. it's competition. I found > some references in a few e-mail that are about 1.5 years old. > > -Tim > > ----- > > Integrated Technical Services > > > ------------------------------ > *From:* [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] *On > Behalf Of *Frank Brickle > *Sent:* Saturday, April 28, 2007 12:40 PM > *To:* Tim Ellison > *Cc:* flexradio@flex-radio.biz > *Subject:* Re: [Flexradio] PowerSDR filter shape factor? > > On 4/28/07, Tim Ellison <[EMAIL PROTECTED]> wrote: > > > > On the web page, it states that the filter shape factor for a 500 Hz > > filter is 1.35:1 and a 2.6 KHz filter is 1.06:1. But I believe that > > these values were taken at 48K sampling rate with a 1024 DSP filter. > > > > I don't have a scope or a signal generator to check this, so I am asking > > > > if anyone has done (or would consider doing) an analysis of the filter > > shape factor with different DSP buffer sizes and sampling rates? I > > would be very interested in seeing how the shape factor varies with > > these parameters at a 500 Hz and 2.6 KHz filter width? > > > > The shape factor is proportional to the sample rate divided by the filter > length. If you double the sample rate but keep the filter length the same, > the shape factor goes up. If you keep the sample rate the same but double > the filter length, the shape factor goes down. The reason I only say > proportional is that what changes the most in this picture is the transition > bandwidth. > > The back of the envelope here says, at 48kHz, difference between 1.35:1 at > 500Hz and 1.06:1 at 2.6kHz is about 30Hz in the transition bands at -60dB. > Not much. > > The rule of thumb is, for a given shape factor, if you double the sample > rate, you need to double the filter length to keep the same shape factor. > > 73 > Frank > AB2KT > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mail.flex-radio.biz/pipermail/flexradio_flex-radio.biz/attachments/20070428/d07582fa/attachment.html _______________________________________________ FlexRadio mailing list FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archive Link: http://www.mail-archive.com/flexradio%40flex-radio.biz/ FlexRadio Knowledge Base: http://kb.flex-radio.com/ FlexRadio Homepage: http://www.flex-radio.com/
