home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.cbm
- Path: sparky!uunet!news.univie.ac.at!chx400!ira.uka.de!ira.uka.de!rz.uni-karlsruhe.de!stepsun.uni-kl.de!sun.rhrk.uni-kl.de!zx1!uet019
- From: uet019@zx1.Darkcrystal.Cyberspace (Oliver B. Warzecha)
- Subject: Re: Decreased seek times.
- Message-ID: <1992Nov5.131146.22893@rhrk.uni-kl.de>
- Sender: news@rhrk.uni-kl.de
- Reply-To: uet019@zx1.hrz.uni-dortmund.de
- Organization: University of Dortmund, Germany
- X-Newsreader: TIN [version 1.1 PL6]
- References: <sheun.720890127@barney>
- Date: Thu, 5 Nov 1992 13:11:46 GMT
- Lines: 56
-
- Sheun Olatunbosun (sheun@cs.city.ac.uk) wrote:
- : In <1992Nov4.015117.7144@pimacc.pima.edu> ppugliese@pimacc.pima.edu writes:
- :
- : >In article <92308.181655K3027E7@ALIJKU11.BITNET>, K3027E7@ALIJKU11.BITNET writes:
- : >> fuzzy@netcom.com (Fuzzy Fox) writes:
- : >>
- : >>>interleave. The basic problem is GCR decoding. I would venture to say
- : >>>that it is impossible for a 1541 with standard 1 MHz processor to decode
- : >>>a sector in less than the time it takes for FOUR sectors to rotate under
- : >>>the head. In fact, GCR decoding is one of the slowest parts of any
- : >> Fast decoding in the 1541 is not possible due to lack of memory.
- : >> If you transfer the sector undecoded to the C64 you can decode it
- : >> there using tables for fast bit manipulation. During that the 1541 can
- : >> already read the next sector.
- : >> Thus it is possible to read and decode an entire track within 3 rotations.
-
- (Stuff about "intelligent" drive deleted)
-
- : In either case, I think the problem seems to be that the GCR
- : encoding/decoding is done in the disk drive software and that the drive
- : has limited RAM. If A LOT more RAM was available, I'm sure that more
- : bulk data could be transferred to and from the computer before the need to
- : read/write more sectors to disk. In addition, it could de possible to write
- : lengthy routines in disk RAM to control the drive directly (and probably
- : do MS-Dos formatting... he guesses). I'm assuming that any hardware vectors
- : can de modified such as in the C64 with its RAM banks and NMI/IRQ interrupts,
- : etc
- Here in Germany, some time ago, two people recognised just these problems
- of the 1541 and wrote the speeder "prologic dos", which has exactly
- these features:
- - 2 MHz Operation of the 1541
- - 8 K additional RAM in the drive
- - Hardware-based GCR decoding
- - Parallel data-transfer via 2 additional 6821 chips
- Well what can I say - I work with it and the result is just fine:
- 202 Blocks loaded in ~4.4 seconds is the fastest you can achieve.
- All other Operations are just as fast (Validating full disk in ~15 s)
- Oh - this is NO commercial advert, but while I've been reading the articles
- in this thread I just recognised many of the things people were talking
- about, and so I decided to come up with this, it might be interesting.
- :
- : But finally, could someone throw some light on these questions.
- :
- : At what serial rate is the data being transferred between the computer
- : and the disk drive?
- I think it was ca. 300 cps with a standard 1541(old). I have heard the
- 1541-II is 10% faster or so.
- : Also, is the C64's VIA 6522 chip communicating with a similar chip inside
- : the disk drive?
- Obviously you talk about the C64's CIA 6526, which is indeed communicating
- with a VIA 6522 in the 1541.
-
- OBW
- uet019@zx1.hrz.uni-dortmund.de
-
- "no flag or uniform ever stopped the bullet of a gun" - Moore/Lynott
-