From aoliveira@mit.edu Tue Apr 11 10:44:57 1995 Path: ashpoly!cascade!transat!eunet-fr!teletel!mit!aoliveira From: aoliveira@mit.edu (Anton Oliveira) Newsgroups: comp.dsp Subject: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr11.104457.2847@mit.edu> Date: Tue, 11 Apr 1995 10:44:57 GMT Organization: Massachusetts Institute of Technology NNTP-Posting-Host: mit.edu X-Newsreader: Netscape 1.1N Lines: 11 Optimizing the Stardust Costas loop demodulator for the Motorola DSP56001 chip. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. However, when downsampling from 44.1 kHz to 64 Hz, accumulator overflow in the biquad IIR filter can destabilize the loop. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Anton Oliveira Opinions expressed are solely my own and not those of Massachusetts Institute of Technology. ------------------------------------------------------------ From lrobertson@tum.de Thu Apr 13 05:51:57 1995 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!tum!lrobertson From: lrobertson@tum.de (Liam Robertson) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr13.055157.3295@tum.de> Date: Thu, 13 Apr 1995 05:51:57 GMT References: <1995Apr11.104457.2847@mit.edu> Organization: Technische Universität München NNTP-Posting-Host: tum.de X-Newsreader: NewsWatcher 2.0b24 Lines: 9 In article <1995Apr11.104457.2847@mit.edu>, aoliveira@mit.edu (Anton Oliveira) writes: > Optimizing the Stardust Costas loop demodulator for the Motorola DSP56001 chip. > By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Liam Robertson -- Technische Universität München "Row 14. No joke." From mlee@cam.ac.uk Fri Apr 14 13:47:57 1995 Path: ashpoly!cascade!transat!eunet-fi!relcom!kiae!cam!mlee From: mlee@cam.ac.uk (Mei Lee) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr14.134757.1014@cam.ac.uk> Date: Fri, 14 Apr 1995 13:47:57 GMT References: <1995Apr11.104457.2847@mit.edu> <1995Apr13.055157.3295@tum.de> Organization: University of Cambridge Computer Laboratory NNTP-Posting-Host: cam.ac.uk X-Newsreader: TIN [version 1.2 PL2] Lines: 8 In article <1995Apr13.055157.3295@tum.de>, lrobertson@tum.de (Liam Robertson) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Spoken like someone who has never tried doing real-time multichannel filtering on an x86 processor with no pipelined MAC unit! -- Mei Lee -- University of Cambridge Computer Laboratory "They will send a copy in my place." From mmoreau@mindspring.com Sat Apr 15 19:43:57 1995 Path: ashpoly!cascade!transat!eunet-de!hansebox!mindspring!mmoreau From: mmoreau@mindspring.com (Mike Moreau) Newsgroups: comp.dsp Subject: Re: [INFO] Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1995Apr15.194357.8870@mindspring.com> Date: Sat, 15 Apr 1995 19:43:57 GMT References: <1995Apr11.104457.2847@mit.edu> <1995Apr13.055157.3295@tum.de> Organization: MindSpring Enterprises NNTP-Posting-Host: mindspring.com X-Newsreader: Netscape 1.1N Lines: 9 In article <1995Apr13.055157.3295@tum.de>, lrobertson@tum.de (Liam Robertson) writes: > Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Mike Moreau "When every node is whole, the channel opens." Macintosh IIfx / FreeBSD 2.1