Xref: utzoo comp.sys.mac:17550 comp.sys.mac.programmer:1406 Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!killer!ames!pasteur!ucbvax!bloom-beacon!athena.mit.edu!lazarus From: lazarus@athena.mit.edu (Michael Friedman) Newsgroups: comp.sys.mac,comp.sys.mac.programmer Subject: Re: ST-506 to SCSI controller info wanted Keywords: SCSI, ST-506, hard disk controllers Message-ID: <5910@bloom-beacon.MIT.EDU> Date: 26 Jun 88 05:17:30 GMT References: <246@iconsys.UUCP> Sender: daemon@bloom-beacon.MIT.EDU Reply-To: lazarus@athena.mit.edu (Michael Friedman) Organization: Murder Inc. Lines: 71 In article <246@iconsys.UUCP> tom@iconsys.UUCP (Tom Kimpton) writes: >I have an opportunity to buy a Toshiba MK50FB ST-506 80 Meg drive >for $400. Considering getting a case, power supply, controller, >driver, and what-not, is this a good deal? And if it is, where >could I get the things I need to get this up and running >with the least amount of hassle? I am a software person, not >really into hardware, so while I would be willing to do some >programming if I could get some kind of sources for a driver, >I would much rather the hardware came out of a box, ready-made. I don't know too much about the hardware side, but the driver will be a nightmare if you have to write your own. Unless you have access to a good oscilloscope, time to burn, and awesome programming skills I suggest that you spend the money for a package deal. The problem you face is that generic SCSI drivers don't exist. You can try ones that go with other drives, but there is no guarantee. As they are intended for use with only one drive they quite possibly have its size built in. That, however, is a minor problem - find it and change it. The real problem is the many differences between the timing characteristics of different drives. Any driver you find will be chock full of little routines whose only purpose is to make the Mac twiddle its thumbs while the drive does something like switch tracks. As your Toshiba is a different drive these routines will be in the wrong places. Therefore, the Mac will sometimes put out data faster than the Toshiba can accept it. (This assumes blind reads and writes. Most drivers do this.) The Mac II is a little different. There is some handshaking going on in the hardware during blind reads and writes. That will probably make things easier for you if you use a Mac II. I happen to know that some drivers that people have written (I think the one in the SCSI Developement package) work on some drives but not others. The problem, of course, is that you won't know that you have to write your own driver until you try every other driver you can get your hands on. If you do write a driver, I strongly suggest that you obtain the SCSI Developement package from APDA. I also recommend SCSI Tool. SCSI Tool is already a fantastic program. The new release is coming out too. (I'm getting a beta test version and boy do I need it!!!). Mac Tutor had a series of articles by someone who did exactly what you are doing. (A different drive, however.) There is also a Tech Note on the subject. In it they say that they will provide minor assistance, but that writing your own driver is not for the faint-hearted. -------- PS. Just be happy you don't have my job - I have to write a routine to acquire data off a board at a CONSTANT RATE and send it to any hard drive. As the SCSI Manager takes a long time when you use a loop, and as drives have unpredictable pauses during which they don't accept data, I am reduced to righting my own program that will include a circular FIFO buffer and direct access to the SCSI chip. When the disk isn't ready for info I keep grabbing data and put it in the buffer. When it is ready I catch up by sending out two values for every one I acquire. As the chip on the MAC II is undocumented I am currently engaged in dissassembling the SCSI Manager and walking through it while keeping track of bus signals with a logic probe. FUN!!! As long as the American Dream is for Americans only |||Mike Friedman it will remain our dream but not our destiny. |||quoting somebody.