• Please review our updated Terms and Rules here

IBM 1501492 MFM Hard Disk Drive

I know this thread is quite old, but I have a question related to what's discussed above. Please excuse my behavour.

As said earlier, the type 1 Xebec contoroller can only take Type 1 drives due to not having any configuration switches. However, I have observed that there are soldering parts for a four-legged DIP labeled "S1" and 1 through 4. I have guessed that this is where a block of configuration switches should be on the variation of the card with support for more than just HDDs of Type 1.

My question is: Can I get support for the other drivetypes by soldering in a set of DIP switches there, or is there some issues with hardware/firmware/software that prevents this?
 
per said:
As said earlier, the type 1 Xebec contoroller can only take Type 1 drives due to not having any configuration switches. However, I have observed that there are soldering parts for a four-legged DIP labeled "S1" and 1 through 4. I have guessed that this is where a block of configuration switches should be on the variation of the card with support for more than just HDDs of Type 1.

My question is: Can I get support for the other drivetypes by soldering in a set of DIP switches there, or is there some issues with hardware/firmware/software that prevents this?
Click to expand...
Because Xebec put out their own branded controllers for the XT, it's best not to refer to the ones that Xebec made for IBM as Xebec, but as either IBM or something like 'Xebec-made-for-IBM'.

According to page 761 of the fifth edition of Mueller's Upgrading & Repairing PCs book, the IBM 10MB controller has different drive entries in it's ROM, compared to the later switchable IBM 20MB controller.

IBM 10MB CONTROLLER (no switches)

Entry 0 --> 306 cyl, 2 heads, no WPC, no LZ, 3mS step rate [probably intended for the ST-506]
Entry 1 --> 375 cyl, 8 heads, no WPC, no LZ, 70uS step rate
Entry 2 --> 306 cyl, 6 heads, WPC of 256, no LZ, 70uS step rate
Entry 3 --> 306 cyl, 4 heads, no WPC, no LZ, 70uS step rate [the entry selected because of no switches]

IBM 20MB CONTROLLER (switches)

Entry 0 --> 306 cyl, 4 heads, no WPC, 305 LZ, 70uS step rate [type 1]
Entry 1 --> 612 cyl, 4 heads, no WPC, 663 LZ, 70uS step rate [type 16]
Entry 2 --> 615 cyl, 4 heads, WPC of 300, 615 LZ, 70uS step rate [type 2]
Entry 3 --> 306 cyl, 8 heads, WPC of 128, 319 LZ, 70uS step rate [type 13]

So adding switches/jumpers to the 10MB controller might work, but are the other drive entries of any use!
 
modem7 said:
Because Xebec put out their own branded controllers for the XT, it's best not to refer to the ones that Xebec made for IBM as Xebec, but as either IBM or something like 'Xebec-made-for-IBM'.

According to page 761 of the fifth edition of Mueller's Upgrading & Repairing PCs book, the IBM 10MB controller has different drive entries in it's ROM, compared to the later switchable IBM 20MB controller.

IBM 10MB CONTROLLER (no switches)

Entry 0 --> 306 cyl, 2 heads, no WPC, no LZ, 3mS step rate [probably intended for the ST-506]
Entry 1 --> 375 cyl, 8 heads, no WPC, no LZ, 70uS step rate
Entry 2 --> 306 cyl, 6 heads, WPC of 256, no LZ, 70uS step rate
Entry 3 --> 306 cyl, 4 heads, no WPC, no LZ, 70uS step rate [the entry selected because of no switches]

IBM 20MB CONTROLLER (switches)

Entry 0 --> 306 cyl, 4 heads, no WPC, 305 LZ, 70uS step rate [type 1]
Entry 1 --> 612 cyl, 4 heads, no WPC, 663 LZ, 70uS step rate [type 16]
Entry 2 --> 615 cyl, 4 heads, WPC of 300, 615 LZ, 70uS step rate [type 2]
Entry 3 --> 306 cyl, 8 heads, WPC of 128, 319 LZ, 70uS step rate [type 13]

So adding switches/jumpers to the 10MB controller might work, but are the other drive entries of any use!
Click to expand...

Thanks a lot.

I have also been able to identify some of the drive-types for the 10MB settings. As we all know, Entry 3 is designed for 10MB drives like the Seagate ST-412 and Miniscribe 2012. Entry 2 seems to be for the 15MB ST-419. Entry 0 matches the specs for the 5MB Seagate ST-406 and Miniscribe 2006 (Can't be 506 since that one has 4 heads), but I find it unusual that this setting has a quite slower step-rate; perhaps it's to keep it compatible with non "ST-412-optimized" drives like the Miniscribe 1006?

Entry 1 is quite unusual. I haven't been able to find any info about any MFM drive having 375 cylinders and 8 heads. Perhaps this setting is for a seldom used 8" HDD?
 
per said:
Entry 0 matches the specs for the 5MB Seagate ST-406 and Miniscribe 2006 (Can't be 506 since that one has 4 heads),
Click to expand...
Ah yes. 153 cyinders and 4 heads.
I've got to stop relying on my old-age memory.
A bit like dynamic RAM with intermittent refresh.
 
I did some more research, and figured that the drive-type settings is not in either the hardware or firmware. It's catually stored in the BIOS extension. This means that you can modify a ROM image of the early version of the HDD BIOS ext to contain other values on boot. You can also redefine this table after boot, by pointing interrupt vector 41h to the new table and issue a drive reset.
 
per said:
I did some more research, and figured that the drive-type settings is not in either the hardware or firmware. It's catually stored in the BIOS extension. This means that you can modify a ROM image of the early version of the HDD BIOS ext to contain other values on boot. You can also redefine this table after boot, by pointing interrupt vector 41h to the new table and issue a drive reset.
Click to expand...

Do you know where, off hand, the table is stored in the 5000059 / 6359121 ROM?

/Bill
 
modem7 said:
Use SpeedStor with the 'romlist' switch, i.e.

A:\>sstor /romlist

It will list the table and the address that it found it at.
It doesn't detect the table in all XT HDD controllers, but certainly works for the common ones.
SpeedStor is at http://members.dodo.com.au/~slappanel555/speedstor.htm
Click to expand...

No kidding -- I would never have guessed that. I'll give it a try.

FWIW, I dissembled the ROM image and found the table at offset 03FF.
I'll try SpeedStor and maybe it'll show me what the entries mean.

Thanks.../Bill
 
wrljet said:
No kidding -- I would never have guessed that. I'll give it a try.

FWIW, I dissembled the ROM image and found the table at offset 03FF.
I'll try SpeedStor and maybe it'll show me what the entries mean.

Thanks.../Bill
Click to expand...

The disasembly with comments are located in the IBM technical referance for the XT. You can find a copy of this referance on the internet (Forum user Trixter used to have one on his FTP site).
 
per said:
The disasembly with comments are located in the IBM technical referance for the XT. You can find a copy of this referance on the internet (Forum user Trixter used to have one on his FTP site).
Click to expand...

Duh! I never thought to look there. ;-)
 
IBM 1501492 MFM Hard Disk Drive (modified 5000059 ROM, and Miniscribe repair)

IBM 1501492 MFM Hard Disk Drive (modified 5000059 ROM, and Miniscribe repair)

I added the 4-position DIP switch to the unpopulated pads on a #6135983 10MB controller. Created a source file for the 5000059 ROM contents (disassembler and some hand work), which builds to the same image as the original. The modified one entry in the drive table to match a 20MB Miniscribe with 615 cylinders. Burned an EPROM.

It works!
Running Spinrite seems to give me the same bad spots as for a Western Digital WDXT-GEN2 controller.

Also, the Miniscribe in question came from eBay and was sold Is-As, which of course in this case meant DOA. The stepper motor was seized from water damage. The clearance between the armature (or whatever you call the movable part of a stepper motor) to the field windings is very small. Rust formed *under* the paint on the armature. Rust is larger than the steel from which it formed, increasing the diameter, seizing up the whole thing. This wasn't immediately obvious after I got it apart. The paint looked a little weird and when I looked closer I found the rust under the paint.

Here's pics, shows the arbor press I used to get the seized motor apart. I had to open the disk chamber to get the stepper out. Getting it all back together and aligned for track 0 was a fiddly job.

IBM-XT-Disk-Controller-6135983W.jpg miniscribe-stepper-press.jpg miniscribe-stepper-inside.jpg miniscribe-stepper-spindle.jpg miniscribe-insides.jpg

/Bill
 
The socket on the card uses a different pinout than regular 2764 EPROMs, so if anybody else wants to modify their cards, either make an adapter or use EPROMs with the special pinout. The pinout the card uses is the same as the pinout used for ROM chips in the original IBM 5150.
 
per said:
The socket on the card uses a different pinout than regular 2764 EPROMs, so if anybody else wants to modify their cards, either make an adapter or use EPROMs with the special pinout. The pinout the card uses is the same as the pinout used for ROM chips in the original IBM 5150.
Click to expand...

I used a Motorola MCM68764 EPROM.
 
I have a question.

I wonder the difference between IBM 1816101 and IBM 150149
 
You must log in or register to reply here.
Back
Top