modem7
10k Member
I've added new IBM 5170 information to my web site:
http://www.minuszerodegrees.net/index.htm
http://www.minuszerodegrees.net/index.htm
- VCF South West - June 14 - 16, Davidson-Gundy Alumni Center at University of Texas at Dallas
- VCF West - Aug 2 - 3, Computer History Museum, Mountain View, CA
- VCF Midwest - Sept 7 - 8 2024, Schaumburg, IL
- VCF SoCal - Mid February 2025, Location TBD, Southern CA
- VCF East - April 2025, Infoage Museum, Wall NJ
-
Please review our updated Terms and Rules here
More IBM 5170 information
- Thread starter modem7
- Start date
billdeg
Technician
Thanks for putting all of this in one place. I have one more thing to add that will help people trying to restore an IBM AT with hard drive.
If you have candidate replacement drives that have neither an IBM drive number nor have their specs listed on the web, you have to guess the IBM drive type, not very efficient.
Messing around with the file attachments program in this forum I was browsing through all stored attachments on the server. Amongst other cool things I found was a program called IDE Identify v1.40. Running this program on a diskette in your 286 system will allow you to determine the drive information. It "..shows head, cylinder, and sector counts for CMOS setup parameters. Drive only needs to be hooked-up, does not need to be prepaired. (Date 12/04/92) Req: 286+CPU. ." Here is where it's stored on this forum:
http://www.vintage-computer.com/vcforum/attachment.php?attachmentid=1103
Awesome!
....so, combined with the info on your site...
http://www.minuszerodegrees.net/5170/hdd_type/5170_hdd_type.htm
... one can use the report from the program to determine the IBM type number! Now I can bring a lot more candidate drives in for testing on my IBM AT..
Bill
If you have candidate replacement drives that have neither an IBM drive number nor have their specs listed on the web, you have to guess the IBM drive type, not very efficient.
Messing around with the file attachments program in this forum I was browsing through all stored attachments on the server. Amongst other cool things I found was a program called IDE Identify v1.40. Running this program on a diskette in your 286 system will allow you to determine the drive information. It "..shows head, cylinder, and sector counts for CMOS setup parameters. Drive only needs to be hooked-up, does not need to be prepaired. (Date 12/04/92) Req: 286+CPU. ." Here is where it's stored on this forum:
http://www.vintage-computer.com/vcforum/attachment.php?attachmentid=1103
Awesome!
....so, combined with the info on your site...
http://www.minuszerodegrees.net/5170/hdd_type/5170_hdd_type.htm
... one can use the report from the program to determine the IBM type number! Now I can bring a lot more candidate drives in for testing on my IBM AT..
Bill
MikeS
Veteran Member
Kinda looks like LapLink to me; how about this link:
http://cd.textfiles.com/megmonster/UTIL_FO/IDEID140.ZIP
modem7
10k Member
Yes, noting that it will only be applicable to 5170s that have had their ST506/412 [MFM] class controller and drive upgraded to IDE-ATA class ones.
billdeg
Technician
Correction:
http://www.vintage-computer.com/vcforum/attachment.php?attachmentid=1103
(but it's nice to have a link to laplink, too...if you have the cable.
http://www.vintage-computer.com/vcforum/attachment.php?attachmentid=1103
(but it's nice to have a link to laplink, too...if you have the cable.
billdeg
Technician
darn, I did not catch that. Is there a utility for the stock AT card?
modem7
10k Member
No. ST506/412 [MFM] class drives do not have an equivalent to IDE-AT's 'Identify Drive' command.
But if needed (and that is going to be rare), there are software techniques that can be used (in most cases) to accurately determine cylinder/head information on ST506/412 [MFM] class drives.
IBMMuseum
Veteran Member
- Joined
- Aug 28, 2006
- Messages
- 1,488
I like the NCL adapter I have for my AT: Plug in an IDE drive, and set the system to drive type 1. The BIOS extension on the adapter identifies the drive parameters, and it runs to its full capacity (I haven't tried it with a drive bigger than 512Mb).
billdeg
Technician
I did not know the identify drive command was what it was using, oh well, but my other trick is to put the drive in an XT class machine and run spinrite, to get the info needed to id the drive.
bd
bd
MikeS
Veteran Member
There aren't many ST412/506 drives whose specs can't easily be found on the web...
modem7
10k Member
In PC and XT class systems, certain hard drive software (e.g. SpeedStor) asks the MFM controller for the drive geometry (probably via INT 13h function 08h). The MFM controller reports the geometry that it has been configured for.
As an example, I have an XT clone that has an MFM controller and ST-225 drive, with the controller jumpered for 615 cyl / 4 heads (i.e. geometry of an ST-225). If I run SpeedStor on that machine, it will report that the drive has 615 cyl and 4 heads. If I replace the ST-225 with a ST-4038 (733 cyl / 5 heads), then power the machine up, SpeedStor still indicates 615 cyl / 4 heads.
Chuck(G)
25k Member
Woah there. The controller knows nothing about the drive geometry on an MFM drive. It's as dumb as the drive is. The BIOS software records in in low RAM or uses a pointer into a BIOS table. But the controller knows nothing about all of this.
This is in contrast to ESDI, IDE and SCSI drives, where the controller is able to interrogate the drive for geometry.
It's possible to interrogate a low-level formatted MFM drive and figure out the geometry by trial and error. i.e. try reading cyl 1023, side 0, sector 1; if that fails, try reading cylinder 511, side 0, sector 1; if that works, try 767,0,1 and so forth in a binary search method. Repeat for sectors and heads.
This is in contrast to ESDI, IDE and SCSI drives, where the controller is able to interrogate the drive for geometry.
It's possible to interrogate a low-level formatted MFM drive and figure out the geometry by trial and error. i.e. try reading cyl 1023, side 0, sector 1; if that fails, try reading cylinder 511, side 0, sector 1; if that works, try 767,0,1 and so forth in a binary search method. Repeat for sectors and heads.
modem7
10k Member
Which statement/post was that in reference to?
Chuck(G)
25k Member
This one:
The controller has no part in that conversation. In the case of an ST412-type drive, it's strictly a function of the BIOS, which, in the case of an AT, is on the motherboard anyway.
The controller has no part in that conversation. In the case of an ST412-type drive, it's strictly a function of the BIOS, which, in the case of an AT, is on the motherboard anyway.
billdeg
Technician
I am talking about drives one can't find. Otherwise what's the point of this thread
modem7
10k Member
Chuck. I don't know why you're bringing the AT into the picture since the post you are referring to contains, "In PC and XT class systems". ATs are a different kettle of fish, and were not in scope of the post.
So, on my XT clone, with XT class hard disk controller, and ST-225, I fire up SpeedStor. SpeedStor reports a drive of 615 clinders / 4 heads. Surely SpeedStor is getting that information via INT 13h Function 08h (a BIOS call). In an XT, the code that that function calls resides in the expansion BIOS ROM on the controller.
Chuck(G)
25k Member
The thread is "More IBM 5170 Information". Nonetheless...
The code could just as easily reside on any card, or in the case of a DDO, in system RAM--the controller itself (not the controller board, that collection of ICs such as he WD1010-5, WD1015-14 and WD1100-13 on a WD1002-WX2) has no part in the conversation. As a matter of fact, wasn't there an add-in card that had nothing more than a ROM on it for the purpose of expanding the drive list? In fact, the controller itself has no connection to the BIOS and cannot even access it.
modem7
10k Member
Of course, one must allow for posts in a thread deviating from topic. A thread can be started on RAM speed and end up on ROMs. This thread deviated into the scenario of ST506/412 class controllers/drives in XTs, and you commented on a post about that scenario.
You write of a controller as being hardware only. It includes any firmware that's on it. I mean, I don't walk into a computer store and ask for a 'printer plus its firmware' - I ask for a printer. So in the example of a Xebec HDD controller for an XT, if I write code that ends up invoking the controller's firmware, then the 'controller' is part of the 'conversation'.
The fact that I can remove the Xebec's ROM, place it on a different card, and observe that the XT still works, does not change the fact that when the firmware is on the controller, it is part of the controller.
Those statements make me think that you're speaking of ATs. The post you commented on was about PCs and XTs. I specifically scoped the post.
But, in regard to my earlier statement of, "In an XT, the code that that function calls resides in the expansion BIOS ROM on the controller.", you are technically right in that the code that gets invoked won't ALWAYS be in the controller. But, you must admit that in the case of a PC or XT, that's going to be pretty rare.
Sometimes one just have to generalise. There are many different hard drive scenarios. To include all possible scenarios can confuse. So for example, what if I had added to one of my posts, "There are some cases where ST506/412 class drives are read to determine the drive's geometry." Those are cases of auto-configure controllers where the controller stores drive geometry on a reserved track. I'm sure that adding that would have just confused the understanding of some readers (including billdeg).
Chuck(G)
25k Member
Well, tomahto--tomayto, I guess.
When I added my first hard drive to a 5150, I used a small homebrew interface card (mostly address decoder and transceivers), a WD1001 controller and a Shugart SA1004 8" drive. Drivers were loaded from floppy. No BIOS support. You're not insisting that the hardware between the 5150 the SA1004 wasn't a controller?
My interpretation of firmware in this context is different also. To me, the "firmware" in the setup above was the program for the Signetics 8x301 bipolar CPU on the thing.
But then, I'm old school.
When I added my first hard drive to a 5150, I used a small homebrew interface card (mostly address decoder and transceivers), a WD1001 controller and a Shugart SA1004 8" drive. Drivers were loaded from floppy. No BIOS support. You're not insisting that the hardware between the 5150 the SA1004 wasn't a controller?
My interpretation of firmware in this context is different also. To me, the "firmware" in the setup above was the program for the Signetics 8x301 bipolar CPU on the thing.
But then, I'm old school.
MikeS
Veteran Member
As a matter of fact I'm not sure any more ;-)
It started with modem7 kindly pointing out that he'd added some info regarding the ST406(MFM) drives, cables etc. in an IBM AT; then you came along with info re identifying an IDE drive which AFAIK was never in any stock AT (and a link to LapLink), and now Chuck and modem7 are arguing about where an XT's drive parameters are stored...
I just thought in case someone got the idea that one had to download/run various utilities, debug, trial and error, etc. in order to get the specs for a drive that it's usually much quicker and easier to just pop the drive model into Google and look it up in one of the many drive parameter charts out there on the web; for example, here are most of Seagate's ST406 drives:
ftp://ftp.seagate.com/techsuppt/mfm/
A little post-holiday testiness here today?
Last edited: