jpalmer@uwovax.uwo.ca (J. Palmer) (10/25/89)
I saved this from last spring, and since Pete asked that it be reposted, here it is: NEWS-Posting: Newsgroup <comp.sys.ibm.pc>, Item <#10407>, Subject <Re: Hard Drives and RLL controllers.> Path: uwovax!julian!watmath!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!rutgers!apple!vsi1!octopus!pete From: pete@octopus.UUCP (Pete Holzmann) Newsgroups: comp.sys.ibm.pc Subject: Re: Hard Drives and RLL controllers. Summary: I guess we have another Urban Legend building here! Keywords: RLL, preamps, etc. Message-ID: <590@octopus.UUCP> Date: 17 Apr 89 13:31:21 GMT References: <1495@fredonia.UUCP> <13557@steinmetz.ge.com> <5559@cbnews.ATT.COM> Reply-To: pete@octopus.UUCP (Pete Holzmann) Distribution: na Organization: Octopus Enterprises, Cupertino CA Lines: 75 In article <5559@cbnews.ATT.COM> wsd@cbnews.ATT.COM (wayne.s.dinsmore) writes: >I have been warned about using certain hard drives with RLL controllers. >I am not sure if this applies to the ST251 but you might be able to >determine that from the next paragraph. > >I use to believe that if a hard drive formatted successfully using an RLL >controller that everything was fine. NOT SO!! The RLL controllers store >more disk data per physical area. This means that more data must be read and >written per unit of time. This causes the read/write preamps to work >harder because of more data flow. Most hard drives that aren't designed >for this will blow the preamps, and indeed will. >[followed by an example given by some reseller somewhere...] I guess it is time to pipe up again with the facts. I covered this subject in great detail about a year ago, and in *my* ignorance, didn't realize that it was worth keeping a copy of what I posted. So I can't repeat my original posting. (Perhaps somebody out there still has a copy they can send me? Please offer before sending... I don't need a hundred copies!) I'll just answer Wayne's statement for now: NO MORE PHYSICAL DATA IS WRITTEN ONTO AN RLL DRIVE THAN AN MFM DRIVE. 'RLL' is simply a matter of *timing* the data more *carefully*. Actually, on average there is slightly *less* physical data (flux changes per inch) written on an RLL drive (assuming random data - the worst case is identical for both kinds of drives). It is almost the same (not really, but I want to try to use an analogy of some kind that people can relate to) as asking yourself: if I compress my files into .ARC files, won't that stress the drive, since I am storing more data in a given amount of space? The correct portion of this analogy is that RLL uses a more efficient code in order to achieve part of its magic. The incorrect portion of this analogy is that there's no hardware change required to store a .ARC file instead of a .EXE... but an RLL controller IS physically different. But the difference is all in the timing. Flux changes on the drive surface must be placed more accurately when connected to an RLL controller. This doesn't physically stress anything. But it is less likely to succeed on a drive with lower quality surfaces, and sometimes if the analog head electronics aren't tuned for RLL. In any case, formatting a drive for RLL WILL NOT HURT THE DRIVE! The only hard thing might be reformatting back to MFM if the drive can't reliably store the RLL format. The back-to-MFM reformat might be difficult simply because some controllers (mostly very old ones) can't do a low-level format without reading data off of the drive! (A catch-22 if there ever was one...) I haven't heard of this problem for a long time though, and formatting utilities such as SpeedStore should take care of it in any case. Just to even up the score even more between the camps ("use only RLL certified" vs. "try it you'll like it")... in my experience, I've been lucky enough to have only one customer's MFM drive that wouldn't handle RLL (an ST-251). Perhaps I've been lucky because I wouldn't try it on an older ST-225, ever. On the other hand, one customer wanted a certified RLL drive. So I picked one up (happened to be a Toshiba). It was certified RLL, but wouldn't work reliably when formatted RLL! Looks to me like you can get bad drives every once in a while, either MFM or RLL. Its just that manufacturers won't *guarantee* RLL compatibility on non-certified drives. (I was able to swap the Toshiba under the warantee for non-performance. Couldn't do that on the ST-251. But I *did* reformat the -251 back to MFM and was able to simply bring it back and get my money back from a cooperative distributor.) Enough said for now! Pete -- Peter Holzmann, Octopus Enterprises |(if you're a techie Christian & are 19611 La Mar Ct., Cupertino, CA 95014 |interested in helping w/ the Great UUCP: {hpda,pyramid}!octopus!pete |Commission, email dsa-contact@octopus) DSA office ans mach=408/996-7746;Work (SLP) voice=408/985-7400,FAX=408/985-0859 -- reply to jpalmer@uwovax.uwo.ca The Economics Institute for Journalists ============================================================================= The New Economics: Overlapping Generalizations & Rationalized Expectorations