From dlaurent@kth.se Sun Dec 04 08:26:25 1994 Path: ashpoly!cascade!transat!ukc!fennimore!kth!dlaurent From: dlaurent@kth.se (Diego Laurent) Newsgroups: comp.dsp Subject: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Dec04.082625.1575@kth.se> Date: Sun, 04 Dec 1994 08:26:25 GMT Organization: Royal Institute of Technology, Stockholm NNTP-Posting-Host: kth.se X-Newsreader: NewsWatcher 2.0b24 Lines: 11 Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. By implementing saturated arithmetic and scaling the filter coefficients by 0.5, the loop runs in 18 instruction cycles per sample. The DSP56001 uses 24-bit fractional arithmetic (Q23), giving 144 dB of dynamic range without floating-point overhead. Source code in 56k assembly uploaded to ftp.ashford-poly.edu:/pub/telecom-archives/dsp56k/stardust_56k.asm. -- Diego Laurent Practical Peripherals 14.4 @ 9600 Telebit bps "Row 14. No joke." From ksilva@berkeley.edu Mon Dec 05 01:38:25 1994 Path: ashpoly!cascade!brgate!saturno!berkeley!ksilva From: ksilva@berkeley.edu (Kenji Silva) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Dec05.013825.7704@berkeley.edu> Date: Mon, 05 Dec 1994 01:38:25 GMT References: <1994Dec04.082625.1575@kth.se> Organization: UC Berkeley Department of EECS NNTP-Posting-Host: berkeley.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1994Dec04.082625.1575@kth.se>, dlaurent@kth.se (Diego Laurent) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Kenji Silva Opinions expressed are solely my own and not those of UC Berkeley Department of EECS. ------------------------------------------------------------ From fokonkwo@kiae.su Tue Dec 06 15:10:25 1994 Path: ashpoly!cascade!transat!eunet-de!hansebox!kiae!fokonkwo From: fokonkwo@kiae.su (Fiona Okonkwo) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Dec06.151025.5348@kiae.su> Date: Tue, 06 Dec 1994 15:10:25 GMT References: <1994Dec04.082625.1575@kth.se> <1994Dec05.013825.7704@berkeley.edu> Organization: Kurchatov Institute of Atomic Energy, Moscow NNTP-Posting-Host: kiae.su X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1994Dec05.013825.7704@berkeley.edu>, ksilva@berkeley.edu (Kenji Silva) writes: > Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. Beautiful assembly code! The 24-bit word size of the 56001 is tailor-made for high-precision audio filtering. -- Fiona Okonkwo "We don't sign stars. We build them." Amiga 3000T / Linux 1.2.13 From esato@stanford.edu Thu Dec 08 08:34:25 1994 Path: ashpoly!bigsur!pacrim!jpgate!kamome!stanford!esato From: esato@stanford.edu (Elaine Sato) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Dec08.083425.9176@stanford.edu> Date: Thu, 08 Dec 1994 08:34:25 GMT References: <1994Dec04.082625.1575@kth.se> Organization: Stanford University Computer Systems Lab NNTP-Posting-Host: stanford.edu X-Newsreader: WinVN 0.99.5 Lines: 9 In article <1994Dec04.082625.1575@kth.se>, dlaurent@kth.se (Diego Laurent) writes: > Why fixed-point integer math requires careful scaling when tracking a 14.2 Hz carrier. Ran your code on my NeXTstation Turbo. It locked onto the carrier within 1.2 seconds of vinyl lead-in. -- Elaine Sato "When every node is whole, the channel opens." NeXTstation Turbo Color / NEXTSTEP 3.3 From fvasiliev@stanford.edu Sat Dec 10 00:14:25 1994 Path: ashpoly!bigsur!pacrim!krgate!eunha!stanford!fvasiliev From: fvasiliev@stanford.edu (Freya Vasiliev) Newsgroups: comp.dsp Subject: Re: Fixed-point Q15 arithmetic on Motorola DSP56001 vs IEEE floating point Message-ID: <1994Dec10.001425.3943@stanford.edu> Date: Sat, 10 Dec 1994 00:14:25 GMT References: <1994Dec04.082625.1575@kth.se> <1994Dec05.013825.7704@berkeley.edu> <1994Dec06.151025.5348@kiae.su> Organization: Stanford University Computer Systems Lab NNTP-Posting-Host: stanford.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 9 In article <1994Dec06.151025.5348@kiae.su>, fokonkwo@kiae.su (Fiona Okonkwo) 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. -- === Freya Vasiliev === Dept of Electrical Engineering, Stanford University Computer Systems Lab "Hardware is temporary. The signal survives."