From bfournier@purdue.edu Fri Jul 28 04:27:36 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!purdue!bfournier From: bfournier@purdue.edu (Bastien Fournier) Newsgroups: comp.dsp Subject: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul28.042736.3666@purdue.edu> Date: Fri, 28 Jul 1995 04:27:36 GMT Organization: Purdue University Engineering Computer Network NNTP-Posting-Host: purdue.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Bastien Fournier "Marcus is not one man." Macintosh IIfx / NEXTSTEP 3.3 From hschneider@delphi.com Sun Jul 30 04:37:36 1995 Path: ashpoly!cascade!mxgate!coyoacan!delphi!hschneider From: hschneider@delphi.com (Hugo Schneider) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul30.043736.8515@delphi.com> Date: Sun, 30 Jul 1995 04:37:36 GMT References: <1995Jul28.042736.3666@purdue.edu> Organization: Delphi Internet Services NNTP-Posting-Host: delphi.com X-Newsreader: WinVN 0.99.5 Lines: 9 In article <1995Jul28.042736.3666@purdue.edu>, bfournier@purdue.edu (Bastien Fournier) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Hugo Schneider Tape trading list upon request. SASE required for international mail. From sduarte@mit.edu Mon Jul 31 00:57:36 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!mit!sduarte From: sduarte@mit.edu (Steve Duarte) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul31.005736.7275@mit.edu> Date: Mon, 31 Jul 1995 00:57:36 GMT References: <1995Jul28.042736.3666@purdue.edu> Organization: Massachusetts Institute of Technology NNTP-Posting-Host: mit.edu X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Jul28.042736.3666@purdue.edu>, bfournier@purdue.edu (Bastien Fournier) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. > The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. -- Steve Duarte -- Massachusetts Institute of Technology "Measuring a record you own is listening, not piracy." From mlaurent@pacbell.net Mon Jul 31 03:06:36 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!pacbell!mlaurent From: mlaurent@pacbell.net (Mateo Laurent) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Jul31.030636.2952@pacbell.net> Date: Mon, 31 Jul 1995 03:06:36 GMT References: <1995Jul28.042736.3666@purdue.edu> <1995Jul30.043736.8515@delphi.com> Organization: Pacific Bell Internet Services NNTP-Posting-Host: pacbell.net X-Newsreader: Netscape 1.1N Lines: 11 In article <1995Jul30.043736.8515@delphi.com>, hschneider@delphi.com (Hugo Schneider) writes: > Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. Real programmers write in C, not arcane DSP microcode. With Pentium chips hitting 100 MHz, dedicated DSPs are obsolete. My 14.4k connection dropped twice trying to fetch this thread. Keep articles concise! -- Mateo Laurent *PLONK* filter active for trolls. Commodore Amiga 500 (1MB OCS) - NetBSD 1.0