PDA

View Full Version : How am I supposed to hook up a Seagate ST-157A?



6885P5H
January 28th, 2017, 01:46 PM
Hello great forum. I swear I could not find any info on this online. I'd like to hook up a Seagate ST-157A to a computer but do not know how. It won't work in my IBM PC 750, it won't work in this new IBM XT I have with a Western Digital 8-bit IDE card and it won't work with this XT-IDE handmade thingy. In fact, it seems like the XT-IDE will not work at all, after the computer accesses the B: drive the system just hangs, with the drive's light staying on. I tried to play with the DIP switches a bit but could not fix it. Disconnecting the B: drive, removing the floppy controller, none of those work... What's interesting is that when I tried to use the Western Digital hard drive that's in the new IBM XT I got in my IBM PC 750 with the Western Digital card, the computer also stopped working after accessing the B: drive like the IBM XT does with the XT-IDE connected... Same thing with the XT-IDE.

So I really don't know how these old IDE stepper drives are supposed to work, I didn't even know they existed until recently. Do I need some kind of Seagate card to use the Seagate drive? And how does this handmade XT-IDE card work? And by handmade I mean handsoldered.

Chuck(G)
January 28th, 2017, 02:03 PM
This is an AT (16-bit) IDE drive, so the XT-IDE should work. I assume that you've checked the jumper settings (ftp://ftp.seagate.com/techsuppt/at/st157a.txt)

The 157A was a pretty reliable drive, so chances are that if it doesn't work, it's a bad drive.

6885P5H
April 12th, 2017, 10:08 AM
Okay so I read the ROM with an EPROM programmer and if I remember correctly I saw that the BIOS of the XT-IDE card was an AT. I thought that the XT-IDE was for XT systems only but guess not. So after I created this topic I bought an EPROM eraser. I just received it yesterday, yeah it took a while. I was tempted to take actions on ebay but decided to wait for the end of the month. The build quality is pretty atrocious. It didn't even work right out of the package, a wire going to the lamp was disconnected so I had to re-solder it. But it does work, but unfortunately the XT-IDE still does not.

With an XT BIOS burned, the computer complains about the ROM and refuses to use it. It just shows a really cryptic "C8000 ROM" message or whatever I have the memory address set at on the card. I tried different IRQs or none at all... After burning the ROM I run a compare command in the programmer's software which compares the loaded ROM file with what is on the ROM and it compares perfectly, meaning the contents are identical. This function does work properly because if I load a ROM in memory that is not the same as what is on the chip then it complains and lists every instances where the data differs.

So, the XT BIOS burns correctly, and yet the computer complains about it.

Chuck(G)
April 12th, 2017, 10:47 AM
I submit that there's either something wrong with your card or the way you're burning your PROM. The XT-IDE was intended for 8-bit bus systems like the 5160; that's its raison d'etre.

The point of the XT-IDE is that most IDE drives are 16-bit only; what the XT-IDE does is break the 16-bit transfers into two 8-bit ones.

I suggest that you post the problem with the ROM on the XTIDE Universal BIOS (http://www.vcfed.org/forum/showthread.php?17986-XTIDE-Universal-BIOS) thread.

modem7
April 13th, 2017, 01:20 AM
And how does this handmade XT-IDE card work? And by handmade I mean handsoldered.
The term "XT-IDE" now seems to be used by many people for any card that imitates the functionality of the VCF produced XT-IDE card.
Assuming that you have the VCF produced one, there is some XT-IDE info at [here (http://minuszerodegrees.net/xtide/XT-IDE%20-%20general.htm)].


With an XT BIOS burned, the computer complains about the ROM and refuses to use it. It just shows a really cryptic "C8000 ROM" message or whatever I have the memory address set at on the card.
Refer to [here (http://minuszerodegrees.net/rom/misc/ROM%20-%20BIOS%20expansion%20ROM.jpg)]. The "C8000 ROM" means that the POST in the IBM XT found the ROM on the XT-IDE card at address C8000, but when the POST calculated the ROM's checksum, the checksum was not as expected.

Assuming that you have the VCF produced XT-IDE card:
* The XTIDECFG.COM program is normally what is used to configure the XTIDE Universal BIOS software, then write that to the EEPROM on the XT-IDE.
* XTIDECFG.COM will calculate the appropriate checksum before writing to the EEPROM
* Try using XTIDECFG.COM to write to the EEPROM.

Druid6900
April 13th, 2017, 05:34 AM
I have great success putting MFM drives into an XT using the Seagate ST-11M controllers.

Should you need one, I have sitting right next to me along with a ST-07A/08A, a WD-1002A-WX1 and a WD1002s-WX2A, all of which have been tested.

However, I agree that it may be a bad drive or a settings problem.

6885P5H
April 14th, 2017, 03:15 AM
Yes I looked up what the message meant on your great website. Doesn't change the fact that it's still really cryptic... Why not say something like "C8000 ROM CHECKSUM ERROR" instead? But I digress....

Yesterday I took a random floppy controller and copied its BIOS on the EPROM I'm trying to use... The EPROM worked perfectly in the floppy controller. So I do know how to use my EPROM programmer and it isn't broken.

This XT-IDE comes from a lot of old computer parts I got from great forum user sombunall. I got it with an EPROM in its socket, so any kind of utility won't be able to change its content. I don't have any EEPROMs to use with it. What so the XT-IDE BIOS does not come with a correct checksum right out of the archive? Why not?


Seagate ST-157A is an IDE hard drive Druid6900.

Druid6900
April 14th, 2017, 05:56 AM
Seagate ST-157A is an IDE hard drive Druid6900.

You are, of course, correct. Bad day yesterday......

6885P5H
April 14th, 2017, 02:04 PM
You had a bad day yesterday? My day yesterday wasn't bad, other than the computer refusing to use the XT-IDE BIOS of course... But is that really bad, it's not like I ever got this thing to work before anyway. I mean, would that make a day be "bad". I don't think so. But what I'm saying does not make much sense so I'll stop.

If I don't get any help I'm not sure what I'm going to do. All I understand from this is that the ROM returns a bad checksum to the computer, one that is not 00. Whatever that means. Why would the computer even care about that. 55 AA 10. Why care about some checksum thingy when you can detect and read the ROM.

So the ROM has a 4-number checksum, and the last two need to be zeros. I added ones after the end of the data in the 8KB BIN file and I needed to finish it with a 2 to make the last two numbers zeros. I edited it with HxD because it has a convenient update checksum button which allows you to see what the checksum has become after each edit. Finally it works in the computer. However I noticed something with it. It is incompatible with my 1.44MB floppy drive controller it seems. If I select the drive in the boot menu, it just says something like 02h error. I think that might be because the computer sees my 1.44MB drive as a 5 1/4" 1.2MB drive. The controller appears to have no way to change that. The B: drive works since it is a 360KB drive, and is seen as such. Interestingly, if I disable the onboard BIOS of the floppy controller the A: drive is able to load the beginning of the OS but after that it just turns off never to do something again. Uhmmm I wonder if those incompatibilities with the floppy controller were fixed in later XT-IDE BIOS releases (I'm using v1.1.3).

modem7
April 14th, 2017, 03:44 PM
Why not say something like "C8000 ROM CHECKSUM ERROR" instead? But I digress....
Possibly the combination of:
1. Conserving space in the BIOS ROM for additional functionality.
2. The error message is listed in the Hardware Maintenance Service manuals ("C8000 ROM Error. ............. Replace Fixed Disk Drive Adapter").
3. IBM's 'get the thing going again' philosophy is having technicians (not owners) 'replace the faulty module with a good one', so simply pointing the technician to the faulty module is good enough.


All I understand from this is that the ROM returns a bad checksum to the computer, one that is not 00. Whatever that means. Why would the computer even care about that. 55 AA 10. Why care about some checksum thingy when you can detect and read the ROM.
Bits in PROMs and EPROMS sometimes go bad. Executing corrupted code is undesirable. Better 'C8000 ROM', a pointer to the problem, rather than the computer locking up during boot.


However I noticed something with it. It is incompatible with my 1.44MB floppy drive controller it seems. If I select the drive in the boot menu, it just says something like 02h error. I think that might be because the computer sees my 1.44MB drive as a 5 1/4" 1.2MB drive. The controller appears to have no way to change that. The B: drive works since it is a 360KB drive, and is seen as such. Interestingly, if I disable the onboard BIOS of the floppy controller the A: drive is able to load the beginning of the OS but after that it just turns off never to do something again. Uhmmm I wonder if those incompatibilities with the floppy controller were fixed in later XT-IDE BIOS releases (I'm using v1.1.3).
The default address setting of the XT-IDE card is D0000. In part, that will be so that some other BIOS expansion ROMs (those sitting below D0000) will execute first. How about you try addressing your XT-IDE card at D0000 instead of C8000, and putting your 1.44MB floppy controller at an address lower than D0000 (if not already there).

Note too that (putting the XT-IDE card aside), there is a known problem with some IBM XTs booting from 1.44M diskettes. See [here (http://minuszerodegrees.net/5160/bios/5160_bios_09MAY86_problem_1.htm)]. Can your IBM XT boot from a 1.44M diskette without the XT-IDE fitted ?

6885P5H
April 15th, 2017, 11:54 AM
That makes sense. If the ROM gets corrupted the checksum would change, so the computer won't execute corrupted code.

Yeah I changed the memory address of the XT-IDE because it was getting in the way of other things in the computer and now everything works fine.

Now I wonder if there exists a program that will run on this computer, that could image an hard drive into a file. Time to resurrect another thread...