mlm@cs.brown.edu (Moises Lejter) (12/12/89)
(My apologies if this has been discussed before - I just joined.) Is it possible to use a MFM hard disk drive with an RLL controller? Are there particular MFM drives that will do it? How would I tell? I have a PC-XT at home with what I believe is an MFM drive and controller. I'm considering buying a second hard-disk, and was wondering whether I could just buy an RLL controller and disk, and connect my existing hard disk to it as a second drive. Alternatively, is there any reason not to have both an MFM and an RLL controller on the same machine? Thanks for any help... Moises ----------------------------------------------------------------------------- Internet/CSnet: mlm@cs.brown.edu BITNET: mlm@browncs.BITNET UUCP: ...!uunet!brunix!mlm Phone: (401)863-7664 USmail: Moises Lejter, Box 1910 Brown University, Providence RI 02912 ----------------------------------------------------------------------------- Internet/CSnet: mlm@cs.brown.edu BITNET: mlm@browncs.BITNET UUCP: ...!uunet!brunix!mlm Phone: (401)863-7664 USmail: Moises Lejter, Box 1910 Brown University, Providence RI 02912
kaleb@mars.jpl.nasa.gov (Kaleb Keithley) (12/12/89)
In article <MLM.89Dec11110300@hector.cs.brown.edu> mlm@cs.brown.edu writes: > Is it possible to use a MFM hard disk drive with an RLL > controller? Are there particular MFM drives that will do it? > How would I tell? At the risk of starting a another flame fest, Yes, there is a 99&44/100 percent chance that it will work. I personally have over two years reliable use out of both Micropolis and Seagate MFM drives on RLL controllers. Others have posted news of similar successful results. Some people believe that you may experience a gradual degradation of the drive. I suggest running the RLL controllers surface scan/media check before doing the low level format, then entering any found bad sectors into the bad sector map. In any case, a periodic backup is always a good idea, no matter what kind of media you use. Chewey, get us outta here! kaleb@mars.jpl.nasa.gov (818)354-8771 Kaleb Keithley
ear@wpi.wpi.edu (Eric A Rasmussen) (12/13/89)
In article <MLM.89Dec11110300@hector.cs.brown.edu> mlm@cs.brown.edu writes: > > Is it possible to use a MFM hard disk drive with an RLL > controller? Are there particular MFM drives that will do it? > How would I tell? While the answer to your first question is technically yes, I cannot emphasize enough how bad an idea it is. Unless a drive is specifically designed for RLL, which means that it has a higher quality media inside it, you SHOULD NOT try to format it as an RLL drive. This is equivalent to formatting a double density disk, designed to hold 360k of data, as a 1.2 Meg disk, or a 720k disk as a 1.44 Meg disk. While this will not physically hurt the drive or the disk, the integrity of the data you store on the disk will be very questionable. What will typically happen is that everything will appear to work, the drive will format successfully, and you will copy your files onto it, and boy will you feel clever. Then, maybe immedietly, maybe after a few weeks, you will start to get errors. Data will be lost forever or corrupted, programs will stop working, and soon the entire disk will become unusable. The basic problem is that the coating on the disk is not designed to handle the density of data used by the RLL format, and it WILL 'forget' what you have stored there. There. I hope I have convinced you that it is a bad idea. As to how to tell if a drive is designed for RLL or not, you would either have to look in a manual for your drive, or consult with your dealer, or a dealer of those brand drives. _ _ +-=-=-=-=-=-=-=-=-+ |_ ,_ . _ |_} _ _ ,_ _ _ _ _ ,_ | ear@wpi.wpi.edu | |_ | | |_ | \ |_\ _> | | | |_| _> _> |_' | | | ear%wpi@wpi.edu | --< A real engineer never reads the instructions first! >-- +-=-=-=-=-=-=-=-=-+
plim@hpsgpa.HP.COM (Peter Lim) (12/13/89)
> > Alternatively, is there any reason not to have both an MFM and > an RLL controller on the same machine? > YES ! For one very simply reason. Both controller will want to use the same port, same BIOS address etc etc. and they will fight with one another. Likewise for the ESDI controller. So far, the only thing that I know which co-exist with other controller is the SCSI controller. A friend of mine just added a SCSI drive to his system with existing ESDI drive. Regards, Peter Lim. HP Singapore IC Design Center. E-mail address: plim@hpsgwg.HP.COM Snail Mail address: Peter Lim Hewlett Packard Singapore, (ICDS, ICS) 1150, Depot Road, Singapore 0410. Telephone: (065)-279-2289
jca@pnet01.cts.com (John C. Archambeau) (12/13/89)
mlm@cs.brown.edu (Moises Lejter) writes: > Is it possible to use a MFM hard disk drive with an RLL > controller? Are there particular MFM drives that will do it? > How would I tell? Possible, but not a good idea, especially with Seagate hard drives. If you do not format an ST412/506 drive as it is specified then the warranty is invalidated. I personally wouldn't risk it. The moment of truth is right after you format the drive RLL. If you come up with a significant number of bad tracks then it probably is a good idea to reformat it MFM, but again, please don't do this to a Seagate drive, you will void the warranty. If it's out of warranty and you don't care if the drive might up and die, go ahead and do it. > I have a PC-XT at home with what I believe is an MFM drive and > controller. I'm considering buying a second hard-disk, and > was wondering whether I could just buy an RLL controller and > disk, and connect my existing hard disk to it as a second > drive. I have yet to see an MFM and RLL controller that will play nice together on an ISA bus. Either you get two MFM drives or two RLL drives. > Alternatively, is there any reason not to have both an MFM and > an RLL controller on the same machine? I don't see any reason why not, but then again, I don't manufacture the controllers. The problem deals with the guts of how hard drive interfaces have evolved. Fortunately when SCSI becomes the de facto standard for the IBM compatable environment, all of this will disappear. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | My opinions are exactly that, ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) ** UUCP : {nsoc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
BHB3@PSUVM.BITNET (12/14/89)
In article <887@crash.cts.com>, jca@pnet01.cts.com (John C. Archambeau) says: > >mlm@cs.brown.edu (Moises Lejter) writes: >> Is it possible to use a MFM hard disk drive with an RLL >> controller? Are there particular MFM drives that will do it? >> How would I tell? > >Possible, but not a good idea, especially with Seagate hard drives. If you do >not format an ST412/506 drive as it is specified then the warranty is >invalidated. I personally wouldn't risk it. The moment of truth is right >after you format the drive RLL. If you come up with a significant number of >bad tracks then it probably is a good idea to reformat it MFM, but again, >please don't do this to a Seagate drive, you will void the warranty. If it's >out of warranty and you don't care if the drive might up and die, go ahead and >do it. > >> I have a PC-XT at home with what I believe is an MFM drive and >> controller. I'm considering buying a second hard-disk, and >> was wondering whether I could just buy an RLL controller and >> disk, and connect my existing hard disk to it as a second >> drive. > >I have yet to see an MFM and RLL controller that will play nice together on an >ISA bus. Either you get two MFM drives or two RLL drives. > >> Alternatively, is there any reason not to have both an MFM and >> an RLL controller on the same machine? > >I don't see any reason why not, but then again, I don't manufacture the >controllers. The problem deals with the guts of how hard drive interfaces >have evolved. Fortunately when SCSI becomes the de facto standard for the IBM >compatable environment, all of this will disappear. > > // JCA > > /* > **--------------------------------------------------------------------------* > ** Flames : /dev/null | My opinions are exactly that, > ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy > ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) > ** UUCP : {nsoc ucsd hplabs!hd-sdd}!crash!pnet01!jca > **--------------------------------------------------------------------------* > */ I and a friend have run a Segate 225, 251, and 4096 RLL even though these are MFM drives. Western DIgital controllers will usually work with no problems when running RLL on an MFM drive. Adaptec Controllers tend to give more errors over time in our experience. Brent H. Besler
phil@diablo.amd.com (Phil Ngai) (12/14/89)
In article <887@crash.cts.com> jca@pnet01.cts.com (John C. Archambeau) writes: |Possible, but not a good idea, especially with Seagate hard drives. If you do |not format an ST412/506 drive as it is specified then the warranty is |invalidated. I personally wouldn't risk it. The moment of truth is right |after you format the drive RLL. If you come up with a significant number of |bad tracks then it probably is a good idea to reformat it MFM, but again, |please don't do this to a Seagate drive, you will void the warranty. If it's |out of warranty and you don't care if the drive might up and die, go ahead and |do it. Please stop repeating these untruths. Formatting an MFM drive as RLL will not physically hurt it. It may not work in RLL mode but then, some disks sold as RLL don't work in RLL mode either. In fact, some disks sold as MFM don't work in MFM either. (I'm referring to DOAs.) RLL was designed to be used with MFM media by using the flux changes more efficiently, not by forcing more flux changes in. The only thing RLL requires is tighter timing than MFM. If the drive delivers a lot of jitter, RLL may not work but you can always revert to MFM, if you have a system that can perform the low level format. I have heard that a few systems can not do this, perhaps this is where the rumor that using an MFM drive as RLL will break it came from, but most systems these days can handle the reformat. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil Washington D.C. is the murder capital of the nation.
kgallagh@digi.UUCP (Kevin Gallagher) (12/14/89)
In article <MLM.89Dec11110300@hector.cs.brown.edu> mlm@cs.brown.edu writes: > > Is it possible to use a MFM hard disk drive with an RLL > controller? Are there particular MFM drives that will do it? > How would I tell? > About a year ago I installed a Miniscribe 3650 drive (40 Meg half height MFM drive) in my PC clone. I connected it to an Adaptec ACB-2072 RLL controller. The drive is rated as an MFM drive, but I understand that Miniscribe uses the same platters in the 3650 as they do in their 60 Meg half height RLL certified drive. (Over a year ago, in Steve Gibson's column in Infoworld, Steve mentioned that he had formatted quite a few 3650's as RRL drives using Adaptec's AT RLL controller, and never had a problem.) So, I decided to give the Adaptec XT RRL controller a try. (Keep in mind that other MFM drives use a differ surface on their platters which cannot hold an RLL format.) The disk drive states it has 809 cylinders, with the parking zone in cylinder 852. Steve indicated that he formatted each of his drives out to cylinder 840 without any problems. On an AT, that got him two full 32 Meg partitions, 16 sectors per track. I used the built in controller firmware low level format utility to format the drive out to 840 cylinders. I ran into no problems at all. The XT controller, however, formats only 25 sectors per cylinder, saving the 26th as a spare in case it encounters a problem formatting that cylinder. This meant that my 840 cylinders came up short of two full 32 Meg partitions. So, I decided to reformat using 850 cylinders. It succeeded without a hitch. I then used the Dos 3.3 version of FDISK to partition the drive into two partitions, each of which is just a tiny bit short of 32K. I have never had a problem. I ended up with about 63 Meg on a 40 Meg drive that cost me under $400. ----------------------------------------------------------------------- Kevin Gallagher attctc!digi!kgallagh or apcihq!apcidfw!digi!kgallagh -----------------------------------------------------------------------
pmahler@dhlmis.DHL.COM (Paul Mahler) (12/15/89)
In article <2373@jato.Jpl.Nasa.Gov> kaleb@mars.UUCP (Kaleb Keithley) writes: >In article <MLM.89Dec11110300@hector.cs.brown.edu> mlm@cs.brown.edu writes: >> Is it possible to use a MFM hard disk drive with an RLL >> controller? Are there particular MFM drives that will do it? >> How would I tell? > I have several 71MB micropolis drives that I have been using for some time with an adaptec rll controller. Micropolis stated that the drives are NOT rll certified. The secret is to use spinrite from Gibson research. This program will do a surface analysis of the disc and spare out any bad sectors. It provides two levels of analysis, a deep one for first time use on a disk, and a less deep one for periodic re-checking. Spinrite gives you a nice map at the end that shows any problems and how they were corected. After having used spinrite I would hesitate to try using an mfm drive with an rll controller without it. Paul Mahler Sunday & Associates uunet!plato!paul (415) 644-0440
phil@diablo.amd.com (Phil Ngai) (12/15/89)
In article <6252@wpi.wpi.edu> ear@wpi.wpi.edu (Eric A Rasmussen) writes: |The basic problem is that the coating on the disk is not designed to handle |the density of data used by the RLL format, and it WILL 'forget' what you have |stored there. Sorry, this is untrue. MFM and RLL both put the SAME number of flux changes on the platter. RLL just uses them more efficiently. RLL does require the flux changes to be more accurately placed timing-wise but most drives have no problem with this. Since there are no more flux changes with RLL than MFM, there is no problem with them being too close together and spontaneously disappearing. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil Washington D.C. is the murder capital of the nation.
bradley.grigor@canremote.uucp (BRADLEY GRIGOR) (12/15/89)
============================================================================== >Date: 12-11-89 (09:28) Number: 14589 CRS Premium Bulletin B > To: ALL Refer#: NONE >From: kaleb@mars.jpl.nasa.gov Read: YES >Subj: MFM as an RLL drive? Conf: (84) U-IBMPC ------------------------------------------------------------------------ On 89-12-11 kaleb@mars.jpl.nasa.gov (Kaleb Keithley) writes: >>In article <MLM.89Dec11110300@hector.cs.brown.edu> mlm@cs.brown.edu >>writes: >> Is it possible to use a MFM hard disk drive with an RLL >> controller? Are there particular MFM drives that will do it? >> How would I tell? >> >At the risk of starting a another flame fest, Yes, there is a >99&44/100 percent chance that it will work. I personally have over >two years reliable use out of both Micropolis and Seagate MFM drives >on RLL controllers. Others have posted news of similar successful >results. Some people believe that you may experience a gradual >degradation of the drive. I have been told that older drives may be more tolerant of RLL controllers because their surfaces are "more forgiving(?)". Anyway, I successfully replaced the original Corona controller with a W.D. RLL controller in my Corona portable XT. Not only did this upgrade give me 60% more capacity, it also allowed me to improve the interleave and the transfer speed! Furthermore, this old (Shugart!) drive has NOT A SINGLE bad block on it! Can you say that about any newer drives, MFM, RLL, or otherwise? Brad Grigor Mainline Technologies --- ~ EZ-Reader 1.21 ~ New Year's Resolution: 1024x768!
jca@pnet01.cts.com (John C. Archambeau) (12/15/89)
BHB3@PSUVM.BITNET writes: >I and a friend have run a Segate 225, 251, and 4096 RLL even though these are >MFM drives. Western DIgital controllers will usually work with no problems >when running RLL on an MFM drive. Adaptec Controllers tend to give more >errors over time in our experience. It is true that it works most of the time (or at least that's what people preach), but I have seen some drives killed via RLL formatting. I don't know if it was just because the drive was on the low end of the quality control spectrum or what, but you can kill your drive permanently. The ST225's and ST251's are the primary offenders. Also, since Seagate is weeding some of the older products from its line (Seagate quit manufacturing ST251-0's back in July) along with the slower 65ms and 40ms units, it might be worth your while to keep it nice and safe because if your ST251-0 dies, Seagate isn't going to swap it out for a nominal swap out fee because they don't make them anymore. The only thing that Adaptec makes that's worth while is the SCSI <-> ST412/506 adaptor board so you can use an ST412/506 as a SCSI hard drive. Other than that, I won't touch Adaptec products. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | My opinions are exactly that, ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) ** UUCP : {nsoc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
jca@pnet01.cts.com (John C. Archambeau) (12/15/89)
phil@diablo.amd.com (Phil Ngai) writes: >Please stop repeating these untruths. Formatting an MFM drive as RLL >will not physically hurt it. It may not work in RLL mode but then, >some disks sold as RLL don't work in RLL mode either. In fact, some >disks sold as MFM don't work in MFM either. (I'm referring to DOAs.) > >RLL was designed to be used with MFM media by using the flux changes >more efficiently, not by forcing more flux changes in. The only thing >RLL requires is tighter timing than MFM. If the drive delivers a lot >of jitter, RLL may not work but you can always revert to MFM, if you >have a system that can perform the low level format. I have heard that >a few systems can not do this, perhaps this is where the rumor that >using an MFM drive as RLL will break it came from, but most systems >these days can handle the reformat. I've encountered enough drives that won't handle the reformat. Again, the primary offenders were the ST225 and ST251. I don't know exactly why (and never knew since Seagate refused to touch them) they died, but I suspect that they were on the low end of the quality control spectrum. I asked the Seagate engineer this, "If they were only formatted MFM, would you swap them without any question?" His answer was "Yes." You tell me, in principle you are probably right, but I apply a set of dead ST225's and ST251's to my reasoning, you can theorize and state the engineering facts about MFM or RLL, but they don't mean crap when you have seen drives killed by RLL formatting. It may defy how a RLL works, but then again, by your reasoning those dead drives should reformat, but they didn't and I tried OMTI, WD, and Adaptec controllers to bring them back from the dead (and my track record in data recovery and hard drive restoration is noted by my customers and business associates in the field), I agree that those drives SHOULD be working, but they didn't and never will. I probably encouered the exceptions to the rule, but those exceptions have made me a stubborn believer with respect to Seagate drives. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | My opinions are exactly that, ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) ** UUCP : {nsoc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
ngeow@bu-cs.BU.EDU (Yee Ngeow) (12/17/89)
In article <28373@amdcad.AMD.COM> phil@diablo.AMD.COM (Phil Ngai) writes: >In article <887@crash.cts.com> jca@pnet01.cts.com (John C. Archambeau) writes: >|Possible, but not a good idea, especially with Seagate hard drives. If you do >|not format an ST412/506 drive as it is specified then the warranty is >|invalidated. I personally wouldn't risk it. The moment of truth is right >|after you format the drive RLL. If you come up with a significant number of >|bad tracks then it probably is a good idea to reformat it MFM, but again, >|please don't do this to a Seagate drive, you will void the warranty. If it's >|out of warranty and you don't care if the drive might up and die, go ahead and >|do it. > >Please stop repeating these untruths. Formatting an MFM drive as RLL >will not physically hurt it. ^^^^^^^^^^^^^^^^^^^^^^^^^^^ Please note that running an MFM drive as RLL DEFINITELY CAN HURT it physically. Running an MFM at RLL push the drive electronics at 50% higher clock rate then an MFM. As with any electronic components, this generates more heat, and if not dissipated qucikly enough, will cause damage to the drive. Personal horror story: I formatted a Seagate ST-412 10Meg MFM drive (Yes, straight out of an IBM PC/XT) using a WD 1002-27X RLL controller. It formats, yes, no problem. Did not give me one single bad sector. I was pretty happy, since now I get 15+ Megs out of the drive. After running it for 10 minutes, suddently it went nuts, making really loud noise, and smell of burned plastic and soder hit me. I turned it off immediately, and the logic board is totally fried. I changed the logic board and it worked again. Moral of story: If you have a really old, pre-1984 drive, it is not wise to format it RLL. It can certainly damage your drive electronics, but not the actual platter mechanics. Kwong
keithe@tekgvs.LABS.TEK.COM (Keith Ericson) (12/19/89)
In article <6252@wpi.wpi.edu> ear@wpi.wpi.edu (Eric A Rasmussen) writes: >In article <MLM.89Dec11110300@hector.cs.brown.edu> mlm@cs.brown.edu writes: >> >> Is it possible to use a MFM hard disk drive with an RLL >> controller? Are there particular MFM drives that will do it? >> How would I tell? > >While the answer to your first question is technically yes, I cannot emphasize >enough how bad an idea it is. Unless a drive is specifically designed for >RLL, which means that it has a higher quality media inside it, you SHOULD NOT >try to format it as an RLL drive. [more useless blater deleted for brevity... I'm getting really tired (make that SICK and tired) of the alleged expert information thrown around this net on RLL'in MFM (rated) drives and how their insides can't take it and you'll blow up the electronics and (the absolute lowest, slimbe-all move of all) Seagate claiming to void the warranty on a drive if you improperly connect an RLL controller to it. This has crossed the line of relatively ridiculous to the absolutely absurd around here!!! Look, folks: manufacturers do NOT purposely "flaw" or "under-construct" the surfaces of a drive to preclude their being used as an RLL drive. (Well, maybe Seagate does - I wouldn't put it past them...). A reputable manufacturer (prime, shining examples: Maxtor and Micropolis) create quality drives, not something just barely running on the ragged edge. These things are easily capable of being connected to RLL controllers. Some of them can even be pushed beyond their specified number of cylinders. Now, if you do this and something goes wrong (i.e., you lose some data) do you go whining to the drive manufacturer? the controller manufacturer? Of course not. You took the chance, dammit, and if it works - fine. If not, then put back the MFM controller and chalk it up to experience. (By the way: experience is what you get when you don't get what you thought you were going to get.) So if you want to try RLL'ing a drive not spec'd for it, and you have the time and money to support it, and if you perceive that the benefits may eventually outweigh the possible difficulties then by all means GO FOR IT. But if you need (or, more importantly, your BOSS needs) the security of going with guaranteed standards (uh, good luck on that one, by the way) then pay your good money, buy the specified/guaranteed RLL drives and sleep calmly tonight (and tomorrow night, and...) That's it form here on RLL'ing MFM drives. Over and OUT. Permanently. kEITHe
blitter@ele.tue.nl (Paul Derks) (12/19/89)
In article <906@crash.cts.com> jca@pnet01.cts.com (John C. Archambeau) writes: >phil@diablo.amd.com (Phil Ngai) writes: >>Please stop repeating these untruths. Formatting an MFM drive as RLL >>will not physically hurt it. It may not work in RLL mode but then, >>some disks sold as RLL don't work in RLL mode either. In fact, some >>disks sold as MFM don't work in MFM either. (I'm referring to DOAs.) >I've encountered enough drives that won't handle the reformat. Again, the >primary offenders were the ST225 and ST251. I don't know exactly why (and >never knew since Seagate refused to touch them) they died, but I suspect that > [stuff deleted] Mr. Ngai is absolutely right, I have formatted some drives on MFM, then later on RLL an on MFM again and I have never had any problem. I think the Seagate drives Mr. Archambeau refers to would have also crashed on MFM. It is a well known fact that Seagates aren't the most relaible drives. I know a couple of friends had ST238 drives who crashed (and these ARE RLL certified!), including one of me. Paul Derks
phil@diablo.amd.com (Phil Ngai) (12/20/89)
In article <44809@bu-cs.BU.EDU> ngeow@cs.bu.edu (Yee Ngeow) writes: | I formatted a Seagate ST-412 10Meg MFM drive (Yes, straight out of an |IBM PC/XT) using a WD 1002-27X RLL controller. It formats, yes, no problem. |Did not give me one single bad sector. I was pretty happy, since now I get |15+ Megs out of the drive. After running it for 10 minutes, suddently it |went nuts, making really loud noise, and smell of burned plastic and soder |hit me. I turned it off immediately, and the logic board is totally fried. I simply don't believe that your drive fried because of running RLL. You probably hooked it up wrong or dropped a screwdriver on it. We already know that you don't know what RPM means. I don't believe anything you have to say on this subject. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil Ann Landers says "Let's talk about legalizing drugs."
clyde@hitech.ht.oz (Clyde Smith-Stubbs) (12/22/89)
ngeow@bu-cs.BU.EDU (Yee Ngeow) writes: > Please note that running an MFM drive as RLL DEFINITELY CAN HURT >it physically. Running an MFM at RLL push the drive electronics at >50% higher clock rate then an MFM. As with any electronic components, this Would someone who knows what they are talking about please stand up and clear this up: to my knowledge the difference between MFM and RLL is NOT that the clock rate is different, but that RLL has less redundant information, i.e. to record one MFM bit takes two transitions while to record one RLL bit takes only 1 point something transitions. If this is the case, then the drive electronics surely are unaware of the difference? The only thing that should matter is the stability of the drive (i.e. motor flutter etc.) because of the reduced redundancy. One thing that would affect an old drive being used with a new controller is the difference in step rate. Older drives (e.g. the old Shugart 10MB drives) could not accomodate buffered seek so maybe putting one of these on a new controller that is toggling the step line at a high rate could cause damage to the drive. Virtually all ST506 (really ST412) drives made in the last few years should not be affected by this. Incidentally I have formatted a Micropolis 1335 (MFM) as RLL and it came up ERROR-FREE! When formatted as MFM it has bad sectors. I still entered the factory bad tracks manually to be safe. -- Clyde Smith-Stubbs HI-TECH Software, P.O. Box 103, ALDERLEY, QLD, 4051, AUSTRALIA. ACSnet: clyde@hitech.ht.oz PHONE: +61 7 300 5011 UUCP: ...!nwnexus!hitech!clyde FAX: +61 7 300 5246
phil@diablo.amd.com (Phil Ngai) (12/23/89)
In article <423@hitech.ht.oz> clyde@hitech.ht.oz (Clyde Smith-Stubbs) writes: |Would someone who knows what they are talking about please stand up |and clear this up: to my knowledge the difference between MFM and RLL is NOT |that the clock rate is different, but that RLL has less redundant information, |i.e. to record one MFM bit takes two transitions while to record one |RLL bit takes only 1 point something transitions. If this is the case, then |the drive electronics surely are unaware of the difference? The only thing |that should matter is the stability of the drive (i.e. motor flutter etc.) |because of the reduced redundancy. I think this is an accurate description of the situation. Note that for ST506, the data separator is on the controller, not the drive, meaning the drive electronics only sees the transitions and it is the controller that extracts the clock and turns the transitions into data bits. Moving the data separator to the drive so the analog information didn't have to travel over a noisy cable was one of the advantages of ESDI. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil Ann Landers says "Let's talk about legalizing drugs."
mitsolid@acf5.NYU.EDU (Thanasis Mitsolides) (12/24/89)
/* acf5:comp.sys.ibm.pc / phil@diablo.amd.com (Phil Ngai) / 5:08 pm Dec 22, 1989 */ In article <423@hitech.ht.oz> clyde@hitech.ht.oz (Clyde Smith-Stubbs) writes: |Would someone who knows what they are talking about please stand up |and clear this up: to my knowledge the difference between MFM and RLL is NOT |that the clock rate is different, but that RLL has less redundant information, |i.e. to record one MFM bit takes two transitions while to record one |RLL bit takes only 1 point something transitions. If this is the case, then |the drive electronics surely are unaware of the difference? The only thing |that should matter is the stability of the drive (i.e. motor flutter etc.) |because of the reduced redundancy. I think this is not exactly accurate. From what I understand, MFM coding needs 0.75 pulses to encode a bit. 2.7RLL takes .5 pulses for a bit. (on the average, since a bit is recorded in 2 clock tiks always) Now, the less the number of pulses needed, the more the number of "silences". 2.7RLL means that the minimum number of silences between successive pulsses is 2, while the maximum is 7. MFM is equivalent to 1.3RLL and FM to 0.1RLL. The trick is to keep the minimum physical distance between pulses constant no matter what the encoding. That decreases the length of a clock tik. That means that the size of silences becomes smaller. Now the controller must be able to recognise smaller silences. Also, the clock (synchronization) is reset on pulses. When the distance of pulses increases, the controller must be able to count more and more silences accurately. Once the controller is RLL and able to perform the task, the question is: Is the HD able to record accurately the smaller silences or in other words, the now finely varying distance of pulses? I don't see that anything would be damaged if the HD cannot keep up to the task. The above is what I understand about RLL encoding. I hope it will clarify some issues. Thanasis ------------------------------------------------------------------------------- Internet: mitsolid@csd2.nyu.edu (mitsolid%csd2.nyu.edu@relay.cs.net) UUCP : ...!uunet!cmcl2!csd2!mitsolid -------------------------------------------------------------------------------
jca@pnet01.cts.com (John C. Archambeau) (12/24/89)
blitter@ele.tue.nl (Paul Derks) writes: >Mr. Ngai is absolutely right, I have formatted some drives on MFM, >then later on RLL an on MFM again and I have never had any problem. >I think the Seagate drives Mr. Archambeau refers to would have also >crashed on MFM. It is a well known fact that Seagates aren't the >most relaible drives. I know a couple of friends had ST238 drives >who crashed (and these ARE RLL certified!), including one of me. The ST225 and ST238 are the drives that tarnished Seagate's reputation, but since Seagate bought Imprimis, I do expect their quality control to increase. I can only state that I've had no problems with Seagates when they were used properly. I had a two year old ST225 until I sold it and upgraded to an ST151. // JCA /* **--------------------------------------------------------------------------* ** Flames : /dev/null | My opinions are exactly that, ** ARPANET : crash!pnet01!jca@nosc.mil | mine. Bill Gates couldn't buy ** INTERNET: jca@pnet01.cts.com | it, but he could rent it. :) ** UUCP : {nosc ucsd hplabs!hd-sdd}!crash!pnet01!jca **--------------------------------------------------------------------------* */
davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (12/27/89)
In article <423@hitech.ht.oz> clyde@hitech.ht.oz (Clyde Smith-Stubbs) writes: | ngeow@bu-cs.BU.EDU (Yee Ngeow) writes: | | > Please note that running an MFM drive as RLL DEFINITELY CAN HURT | >it physically. Running an MFM at RLL push the drive electronics at | >50% higher clock rate then an MFM. As with any electronic components, this | | Would someone who knows what they are talking about please stand up | and clear this up: to my knowledge the difference between MFM and RLL is NOT | that the clock rate is different, but that RLL has less redundant information, There are some really good articles on this, posted by Pete Holzman at Octopus. I have them on my work machine, but I'm on vacation. The gist is that RLL reduces the number of bits by using timing info to *reduce* the flux changes used to represent data. What we call MFM is technically also RLL, by the way, something like RLL 1,3, while what is usually called RLL is 2,7. I could explain this, but why bother when there are articles which do it better. When running an RLL controller the rate of flux changes is actually less, although the information data rate from the bus to the controller *is* higher. | i.e. to record one MFM bit takes two transitions while to record one | RLL bit takes only 1 point something transitions. That's the difference between FM and MFM, but you have the right idea. RLL works the disk smarter, not harder. If the drive media, drive speed flutter, or electronics are not up to RLL it won't work, but nothing bad will happen to the drive (just the data). Typically the first 24 hours show problems if you are going to get them, although something just on the edge might tend to fail in the summer or winter (air conditioning or electric heat line noise) if your luck is very bad. | Incidentally I have formatted a Micropolis 1335 (MFM) as RLL and it came | up ERROR-FREE! When formatted as MFM it has bad sectors. I still entered | the factory bad tracks manually to be safe. I have noticed this, too. When I went from Adaptek to WD RLL, my ST251 showed no bad tracks. That was about six months ago. The other day I stated getting "can't read" messages, and I checked the whole disk again for bad tracks. Like to rabbit commercial it's "still going." The message came from a program trying to do something fancy with the disk and reading outside a partition. I'm running an ST4096, too, with no problems, although it had a few bad tracks and is way over the rated MTBF. I keep current backups on ALL my systems, so I don't treat RLL as special. -- bill davidsen - sysop *IX BBS and Public Access UNIX davidsen@sixhub.uucp ...!uunet!crdgw1!sixhub!davidsen "Getting old is bad, but it beats the hell out of the alternative" -anon
bradley.grigor@canremote.uucp (BRADLEY GRIGOR) (12/31/89)
cz>From: clyde@hitech.ht.oz (Clyde Smith-Stubbs)
cz>Orga: HI-TECH Software, Brisbane, QLD, Australia
writes:
cz>One thing that would affect an old drive being used with a new
cz>controller is the difference in step rate. Older drives (e.g. the old
cz>Shugart 10MB drives) could not accomodate buffered seek so maybe
cz>putting one of these on a new controller that is toggling the step
cz>line at a high rate could cause damage to the drive. Virtually all
cz>ST506 (really ST412) drives made in the last few years should not be
cz>affected by this.
My Shugart 10 MB drive has been running as RLL for the past 6 months
with no apparent damage -- indeed, I'm happy I replaced the original
controller -- 80 bucks gave the old machine a "new lease on life"!
-- bag -- Newmarket, Ontario, Canada -- The Land of TelePoker
---
~ DeLuxe 1.11a18 #4613 What ain't forbidden is mandatory!jf.messier@canremote.uucp (JF MESSIER) (01/15/90)
>> Is it possible to use a MFM hard disk drive with an RLL >> controller? Are there particular MFM drives that will do it? >> How would I tell? >The basic problem is that the coating on the disk is not designed to >handle the density of data used by the RLL format, and it WILL >'forget' what you have stored there. I had a really bad experience with such controllers. At Revenue Canada, we have a lot of Epson Equiti I+, each with a 30M hard disk (in fact, a 20M and a RLL controller). BUT, the drives used were Miniscribe, NOT APPROVED for RLL, so MFM-ONLY. We had problems, mainly new bad sectors appearing and destroying information. When Epson came, then ADMITTED installing NON-RLL-APPROVED hard disk with RLL controllers in their computer. One of the technician told me that this was to save money to Epson. When Epson came to change the HD, they brought ST-236, hich are RLL-approved, but they had to try 16 of these to fid 5 that worked in the computers. They had to come 4 times and they wasted a lot of time in testing those drives. Even after all drives were installed, we had other bad sectors appearing on the disks.... STAY AWAY FROM RLL, PLEASE......... !!!! --- * Via ProDoor 3.0 * QNet 1.04a1: SuperByte '386
mlm@cs.brown.edu (Moises Lejter) (01/22/90)
In article <90011904272373@masnet.uucp> jf.messier@canremote.uucp (JF MESSIER) writes: > From: jf.messier@canremote.uucp (JF MESSIER) > Newsgroups: comp.sys.ibm.pc > Date: 15 Jan 90 07:45:00 GMT > Organization: Canada Remote Systems Limited, Mississauga, ON, Canada > > >> Is it possible to use a MFM hard disk drive with an RLL > >> controller? Are there particular MFM drives that will do it? > >> How would I tell? > >The basic problem is that the coating on the disk is not designed to > >handle the density of data used by the RLL format, and it WILL > >'forget' what you have stored there. > > I had a really bad experience with such controllers. At Revenue > Canada, we have a lot of Epson Equiti I+, each with a 30M hard disk > (in fact, a 20M and a RLL controller). BUT, the drives used were > Miniscribe, NOT APPROVED for RLL, so MFM-ONLY. We had problems, mainly > new bad sectors appearing and destroying information. ... I recently bought a FlashCard-30 from CompuAdd, built from an RLL controller and a MiniScribe 8438. The disk would format properly, but a couple of days after being formatted it would start failing the PC-TOOLS verify test (new bad sectors, and eventually hard seek errors). I'll be calling CompuAdd this week, but my problem certainly seems to be explained by the article above. Does anyone know if in fact the MiniScribes are *NOT* reliable RLL drives, in general? I've heard of other FlashCard 30s that have worked out fine... Moises ----------------------------------------------------------------------------- Internet/CSnet: mlm@cs.brown.edu BITNET: mlm@browncs.BITNET UUCP: ...!uunet!brunix!mlm Phone: (401)863-7664 USmail: Moises Lejter, Box 1910 Brown University, Providence RI 02912
kentkar@polari.UUCP (kent karrer) (01/22/90)
> > STAY AWAY FROM RLL, PLEASE......... !!!! I have used RLL for years and it works fine if you use the right disks intended for RLL use. RLL disks are certified by their manufacturers for use with RLL controllers. RLL has been around for awhile and is a stable, reliable product if used correctly. I perfer using RLL for two reasons: I usually get a bigger bang for the buck (ie, more megabytes for the dollar) and when I sell a used RLL disk I know it is both MFM and RLL compatible. Unfortunately, you were burned by a supposedly reputable vendor who cut corners to line their pockets and you both paid. I appreciate your frustration and obvious hesitancy regarding RLL disks. But I want to make known to others who may read this that RLL works fine when its used as intended. :->
chapman@fornax.UUCP (John Chapman) (02/04/90)
In article <MLM.90Jan21115353@hector.cs.brown.edu<, mlm@cs.brown.edu (Moises Lejter) writes: < In article <90011904272373@masnet.uucp< jf.messier@canremote.uucp (JF MESSIER) writes: < < < From: jf.messier@canremote.uucp (JF MESSIER) < < Newsgroups: comp.sys.ibm.pc < < Date: 15 Jan 90 07:45:00 GMT < < Organization: Canada Remote Systems Limited, Mississauga, ON, Canada < < < < << Is it possible to use a MFM hard disk drive with an RLL < < << controller? Are there particular MFM drives that will do it? < < << How would I tell? < < <The basic problem is that the coating on the disk is not designed to < < <handle the density of data used by the RLL format, and it WILL < < <'forget' what you have stored there. < < < < I had a really bad experience with such controllers. At Revenue < < Canada, we have a lot of Epson Equiti I+, each with a 30M hard disk < < (in fact, a 20M and a RLL controller). BUT, the drives used were < < Miniscribe, NOT APPROVED for RLL, so MFM-ONLY. We had problems, mainly < < new bad sectors appearing and destroying information. ... < < I recently bought a FlashCard-30 from CompuAdd, built from an RLL < controller and a MiniScribe 8438. The disk would format properly, but < a couple of days after being formatted it would start failing the < PC-TOOLS verify test (new bad sectors, and eventually hard seek < errors). I'll be calling CompuAdd this week, but my problem certainly < seems to be explained by the article above. Does anyone know if < in fact the MiniScribes are *NOT* reliable RLL drives, in general? In contrast to the previous posters' experiences I have been running two miniscribe 6850 70mb mfm drives with an rll controller 24hrs a day for 7 months now with no problems except that if I power the system down for some reason and let it get cold (~60F) I get errors when I reboot - I just let the machie run for ten minutes and then no more problems once it is warmed up)
davidsen@sixhub.UUCP (Wm E. Davidsen Jr) (02/06/90)
I've had no trouble with RLL on 4096 or 251 drives, but there *is* a possibility that it won't work. In most cases if it runs for a day of heavy usage it will be okay. I've seen a few which just plain don't work, chance you take. -- bill davidsen - sysop *IX BBS and Public Access UNIX davidsen@sixhub.uucp ...!uunet!crdgw1!sixhub!davidsen "Getting old is bad, but it beats the hell out of the alternative" -anon
besler@egrunix.UUCP (Brent Besler) (02/09/90)
In bith my and a friends experience, WD RLL cards are easier on MFM drives than Adaptecs. The Adaptecs seem to be more prone to read errors. Brent H. Besler