toms@oakhill.UUCP (04/07/87)
How would one go about upgrading a PC-XT to a faster hard disk?
The only faster drives I have seen advertised (avg. seek <30ms) are for
the PC-AT. Is such performance available for a standard PC or PC-XT?
Can AT drives/controllers be modified for the PC?
Furthermore, if such hardware upgrades are possible what kind of
software problems will arise?
By the way, if one reformats the hard disk for an interleave factor
of less than the recomended 6 how does one know how close to the limits
of the hardware they are? How fast has anyone gone successfully?
Thanks in advance for any answers to these questions or any other
disk performance information.
--
=============================================================================
Motorola Semiconductors Tom Spohrer
Austin, Texas {seismo,gatech,ihnp4}!ut-sally!oakhill!tomsdick@zaphod.UUCP (04/09/87)
Summary: Expires: Sender: Followup-To: Organization: In article <859@oakhill.UUCP> toms@oakhill.UUCP (Thomas Spohrer) writes: > > How would one go about upgrading a PC-XT to a faster hard disk? >[...] > By the way, if one reformats the hard disk for an interleave factor >of less than the recomended 6 how does one know how close to the limits >of the hardware they are? I suggest you get a copy of the HTEST/HFORMAT package from Paul Mace and friends up in Ashland, OR. It comes with a program called HOPTIMUM that will determine the best interleave for YOUR particular system/configuration/controller/drive combination. For example, I have a 4.77 MHz XT (a _real_ one!), but it's running a '286 board at 8 MHz. I am also using an Adaptec 2070A RLL controller with Seagate ST225 disk drives. Whose suggested interleave values should I use? IBM's for the XT? The '286 board manufacturer's? Adaptec's? Seagate's? I had been using Adaptec's suggested value of 3, but a real-time analysis by HOPTIMUM suggested that a value of 5 would be a lot better. It was--to a tune of 140% improvement!!! My only affiliation with Paul Mace or Mark Kolod (developer of HTEST) is one of personal admiration and product satisfaction. -- Dick Flanagan, W6OLD UUCP: ...!ucbvax!sun!plx!dick GEnie: FLANAGAN (The usual disclaimers apply)
mvolo@ecsvax.UUCP (04/13/87)
Part of a response to a query about faster hard disks for an XT discusses [Dinterleave factors, and incidently talks about using an Adaptec RLL controller with a Seagate ST225. Everything I've about RLL controllers recently suggests that non-plated media HD's such as the ST225 don't work reliably over the long haul with Adaptec RLL controllers. Am I missing something; or does this sometimes work out in practice. I know this is not a response to the faster hard disk question, but I am cur- ious. --Mike Volow, Psychiatry, Durham Veterans Administration Medical Center Durham, NC, 27712 919 383 3563 mvolo@ecsvax.UUCP
wtm@neoucom.UUCP (04/14/87)
Right you are. It is best not to use the Seagate ST-225 with an RLL controller card (insert your favorite brand). The ST-225 doesn't promise to have plated media, plated being better for the higher bit density of RLL than standard oxide media. A brochure from Seagate also claims that the electronics of the ST-238 have also been "optimized" for RLL 2/7.-- although that probably isn't anything more than a trim pot set to a slightly different value. Rumor has it that ST-225s currently appearing on the market really are 238s, but just have not undergone the level of testing necessary to certify as a 238. What can I say, you pays your money and takes your chance if you use a 225 with RLL. A 238 is only about 50 bucks more than a 225, so I opted for the 238 on my own computer. --Bill
dick@zaphod.UUCP (04/14/87)
Summary: Expires: Sender: Followup-To: Running a Seagate ST225 with an RLL controller (Adaptec 2070A in my case), is like taking a single-sided diskette and formatting it for double-sided use--98% of the time it will work just fine, but none of the suppliers will guarantee that it will. My combination WAS a tad flaky until I thoroughly burned in the pair of them (I actuall have two ST225's installed) with the cont- roller, and they now appear to be running just as smooth as can be. HOWEVER, be careful with _recently_ purchased ST225's and ST238's-- word on the street has it they are both sufferring from quality- control problems that causes their internal track register to get (and sometimes stay) out of sync with their actual track position. This doesn't apparently cause any lost data, but often requires a reset to be able to read again. So please buy your hard disks from a dealer they can be returned to if they turn out to be defective, and be sure to really burn them in when you first get them. -- Dick Flanagan, W6OLD UUCP: ...!ucbvax!sun!plx!dick GEnie: FLANAGAN
henkp@nikhefk.UUCP (04/15/87)
In article <2902@ecsvax.UUCP> mvolo@ecsvax.UUCP (Michael R. Volow) writes: >Part of a response to a query about faster hard disks for an XT discusses I use a fast HD (Microsience HH1050 40MB 28ms) in a M24 (= ATT6300). It is very fast. Controller Adoptek 2070A with a interleave of 3. Within a 33MB partition the average seektime is 20ms (PC labs bench- mark coretest). I have never tested other interleave factors. In the last weeks the servo loop of the HH1050 is not very stable on a small group of tracs around trac 400. The servo pulse pattern is repetive and looks ok. The processing of the servo-data is very tricky. They are adding many signals to drive the voicecoil amplifier. The drive is one of the first series. There are some mods on the print with wirewrap wire. I have no "real technical" documentation. Who knows more about this drive or about servo trac processing. Please email direct. When there is more interest and I will make a summery for the net. Henk Peek, henkp@nikhefk.UUCP ..!seimo!mcvax!nikhefk!henkp.UUCP