• Please review our updated Terms and Rules here

Using Adaptec SCSI card, MSDOS 6.22 fails to partition (FDISK) and format SCSI drive

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
No, you did not. A 286 is not able to run Windows 95. It has nothing to do with speed, it's simply not possible on a technical level. The code can not run on a 286 CPU.
Fiirst, I worked for the US Navy and we had some of the oldest equipment to test, evaluate and record findings. We had 18-inch 80-MB hard drives, SIMD processors and 512-KB RAM, with OS as CM/2. The recorded findings went on a Zenith 286 with a 44-MB Bernoulli drive and 4-MB RAM. The OS was Win 95. I knew the guy who setup the machine and his job was to make it work so we could generate our reports.

Second, your page is similar to the one shown by Chuck(G) and if you noticed the AHA-1522 page there are drivers for OS's Win 95 and NT along with OS/2. Just food for thought.

Third, I still have a SCSI PCI card that I used with 4-SCSI drives and CDROM on a Pentium Pro that ran Solaris 8. I no longer have the drives or the Solaris 8 for X86 as opposed to SPARC; I think the drives were all Seagate. Anyway the point was that I do happen to have some experince with SCSI even though it was mostly UNIX.
 

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
Everything from a FH HP 9GB 5.25 drive, CDC 330MB FH 5.25", various Seagate 3.5" SCSI drives... down to a Jaz and Bernoulli drives. I've also used SCSI Zip drives as well. I've even imaged a 20SC external hard drive for Apple. Adaptec made some really good controllers.
Since, I started having my difficulties with the SCSI. I plugged in a Zip drive off the parallel port low and behold it works as a hard drive. The Zip drive doesn't boot, yet, mostly because I wanted the SCSI drive to do the booting. Before anyone jumps in and asks if I remove the Zip drive with I try to format the SCSI and that answer is yes, I do. Just curious, I know a Bernoulli would boot; will a Zip drive boot?

Early in the discussion, lazyguy asked if my cable was good or drive might be bad. Even though I can hear the drive, I am wondering if the drive may have a damaged part in the initial cylinders or sectors that we can't test because the drive has low level format but not high level. Thank you for the feedback on the drives and controller. It is much appreciated.
 

Chuck(G)

25k Member
Joined
Jan 11, 2007
Messages
39,833
Location
Pacific Northwest, USA
I don't recall if the Adaptec BIOS allows booting from removable drives. I'd have to check the docs.

None of my SCSI drives can be low-level formatted--you can issue the command, but it returns immediately with nothing being done.
 

Caluser2000

Veteran Member
Joined
Jan 3, 2010
Messages
4,665
Location
New Zealand
As mentioned MS Windows 95 will not run on a 286 at all. a 386SX system is the minimum cpu it will run on. My Adaptic card is also from 1990. It originally came out of my 386DX25 system. The Apple branded 100meg scsi hdd was DOA so pull it the scsi hdd out and put the card in storage as at the time never had a replacement scsi hdd back then.
 

Lutiana

Administrator
Staff member
Joined
Mar 26, 2009
Messages
3,229
Location
Dublin, CA USA
Wait a sec, Isn't the AH-1522 32 bit only? So requires a 386 to work properly... If I recall correctly its a plug and play card and all the utilities are pretty much 32 bit since it came out win Win95 did... I'll have to try to dig up a manual for it. I know for sure the 1510A works fine on a 286, because I've had that working in the past.

For MS-DOS, it will require Dos 5 or higher if I recall correctly. 6.22 is fine.

Here is the manual

Says AT or higher so guess 286 is fine...

Stop leading people to false paths (again). If you are not able to help, just don't post.

There is no such thing as a 16-bit ISA card that uses 32-bit and requires a 386. 16-bit ISA cards are required to work in a 286, otherwise they wouldn't be ISA cards. The system requirements of any software included does not stop the hardware from working. You can even use a SB AWE64 in a 286, despite the software asking for a Pentium. You are not forced to use the included software. And in this case, the card has a BIOS and doesn't require any software at all.

Besides, the AHA-1522 dates back to 1990. Tell me what version of WIndows 95 you were using back then, because I want a copy...

@alejack12001: I would recommend trying the following: enter the SCSI BIOS and format the hard disk there. This will wipe any data that could cause issues (like a broken MBR). Then do a clean boot into DOS by pressing F5. Do FDISK and FORMAT again and see what it does. Do not use SCSIFMT or any other tools not from DOS.


Yeah, so I could not get the AH-1522 to work on my IBM AT at all. It would not boot. I am not the only one who has had this issue (eg: here and here), and I have run into more than a few people, who like Timo, go on about ISA being 16bit and therefore has to be supported on the IBM AT as that defined the spec etc etc etc, which is great and all, but my direct personal experience and experimentation says otherwise. That same AH-1522 card worked absolutely fine in any machine that was 386 or higher.

Also, from the MOD perspective. Let's play nice, disagreements are fine, but be professional about it and avoid devolving into streak talk (ie telling someone to sod off).
 
Last edited:

Caluser2000

Veteran Member
Joined
Jan 3, 2010
Messages
4,665
Location
New Zealand
As I mentioned above my AHA-1522 worked fine on my 286 motherboard.
 

Attachments

  • IMAG0079.JPG
    IMAG0079.JPG
    480 KB · Views: 5
  • IMAG0083.JPG
    IMAG0083.JPG
    494.5 KB · Views: 5

Caluser2000

Veteran Member
Joined
Jan 3, 2010
Messages
4,665
Location
New Zealand
May I ask what motherboard are you using?
Chaintech ELT-286B-1000-SM 286/10 cpu (Award bios with C&T chipset)1meg on the mobo and an extra two meg via isa ram expansion card. I'll start a separate thread in a bit but want to get everytihng functional before doing so.
 
Last edited:

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
Chaintech ELT-286B-1000-SM 286/10 cpu (Award bios with C&T chipset)1meg on the mobo and an extra two meg via isa ram expansion card. I'll start a separate thread in a bit but want to get everytihng functional before doing so.
Thank you for sharing this information with the forum.

Mine is slightly different as its was supposed to the an IBM 5162 (PC XT/286). The motherboard is an unbranded 286 that a member of the forum helped me find some documentation on it and upgrade the AMI bios. So far I haven't had much luck with internal drives on this machine e.g., MFM, IDE and now SCSI. The only thing that works is an external Zip drive off a parallel port. Of course, the zip doesn't boot. I have to use a floppy to boot it, but, I am okay with that process.
 

mR_Slug

Veteran Member
Joined
Nov 28, 2006
Messages
941
Location
UK
This thread is getting confusing.

First your autoexec.bat and config.sys files are these located on the SCSI hdd you are trying to access? If so they are unneeded as chuck(g) pointed out.

If you boot from floppy and access the drive, if it is set to id 0 then there is no need for the drivers.

Third the AHA-1522 came out in 1990, the AHA-1522A around '93 and the AHA-1522B in '95 as far as I can tell. Which one do you have? They are all 16-bit controllers. SOME later software (and drivers) for the cards may use 32-bit code. At any rate if you are trying to boot from the drive, there should be no drivers involved.
 

Getafix

Experienced Member
Joined
Apr 28, 2022
Messages
54
Location
Vancouver, BC
Hi,

As far as I know the AHA-1522 is the same as the AHA-1520, but the 1522 also has a floppy controller. If that is true, then the 1522 should work in a 286 since I have a 1520 working in my AT (5170).

As I said before I did have to change my BIOS and this was easy to see since EZ-SCSI didn’t see the 1520 till I changed the BIOS. So, if EZ-SCSI see the 1522 in the OPs machine with the AMI BIOS, I would suspect the drive or the cable.

Stefan
 

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
This thread is getting confusing.

First your autoexec.bat and config.sys files are these located on the SCSI hdd you are trying to access? If so they are unneeded as chuck(g) pointed out.

If you boot from floppy and access the drive, if it is set to id 0 then there is no need for the drivers.

Third the AHA-1522 came out in 1990, the AHA-1522A around '93 and the AHA-1522B in '95 as far as I can tell. Which one do you have? They are all 16-bit controllers. SOME later software (and drivers) for the cards may use 32-bit code. At any rate if you are trying to boot from the drive, there should be no drivers involved.
In response to your first question, the answer is no. I have those files on my floppy boot disk. Without the floppy, following POST, the machine just displays the following picture:
5301AC48-BD28-474E-9F72-E88E524324A8_1_105_c.jpeg

If the floppy is removed from the floppy drive then as shown in the next picture just the upper half of the screen is displayed and the system is reaching out to the floppy to seek the boot disk. and goes no further to provide a prompt.

4E67A7DD-EE20-43A4-9B63-32A10958840D_1_105_c.jpeg

If the floppy is in the drive then the lower half of the screen is displayed. However, when the A prompt is shown and if I enter C: then the machine will go there to C. If I try to copy a file then the response is invalid drive specification. If I enter FDISK then the response is nothing as the lamp on the floppy is lighted and doesn't go out. So, it is hung. I have to reboot.

I can turn off the drivers, however, the same response is given as above invalid drive specification. There is no command.com or boot files because i used the SCSIFMT to low level format the drive thinking it needed that to clear out the invalid drive specification. This didn't work. FDISK just will not configure the drive for the FAT or put in a partition and without that I cannot format this drive.

In response to you second question, the SCSI card is AHA-1510A/1520A/1522A with the ROM chip showing the following ROM: ADAPTEC INC., 585201-00 A, BIOS 4700, @ 1993

I tried AFDISK and i get the following response as shown:

1E46C99F-F8B4-43F1-838B-D9222DDEDDA6_1_105_c.jpeg
 

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
Hi,

As far as I know the AHA-1522 is the same as the AHA-1520, but the 1522 also has a floppy controller. If that is true, then the 1522 should work in a 286 since I have a 1520 working in my AT (5170).

As I said before I did have to change my BIOS and this was easy to see since EZ-SCSI didn’t see the 1520 till I changed the BIOS. So, if EZ-SCSI see the 1522 in the OPs machine with the AMI BIOS, I would suspect the drive or the cable.

Stefan
I did use the minuszerodegrees website to get the AMI BIOS upgrade from the copyright date of 1987 to 1989. I didn't see much difference between the two BIOS. The cable was acquired from Amazon about three weeks ago. So, it is presumably new. The Conner drive is recognized as seen in the pictures I provided. It does spin up, so the motor works. Whether it is spinning up to the right speed is another diagnostic question. I don't hear much noise except for high pitched whirling sound coming from the drive. It does get warm but not hot. I don't know if the first sector is damaged or the cylinders are damaged or if the heads are damaged. Since the drive will not write then I may have to assume that the drive could be defective without being able to test it. I don't know.
 

Chuck(G)

25k Member
Joined
Jan 11, 2007
Messages
39,833
Location
Pacific Northwest, USA
AFDISK is complaining that the drive interface has been instantiated during boot time. Change the SCSI ID to something other than 0 or 1 and it will no longer be included by the 1522 BIOS. The idea behind the message is that two different interfaces writing to the disk is not a happy situation, as the synthetic geometry may not agree (SCSI disks are LBA addressed only, not CHS. If an application demands CHS, the driver has to dummy something up.)

Changing the ID should result in AFDISK working. You can then either use the drive with ASPIDISK.SYS or remove the drivers and change the ID back to 0.

That's the way I read it, anyway.
 

1ST1

Veteran Member
Joined
Oct 21, 2014
Messages
1,290
Location
near frankfurt/m, germany
I think it could be one of the following issues:
a) IRQ, ROM-Adress, IO-port of the 1522 is shared with other hardware. Solution: See which of the ressources they are, try to avoid the conflict.
b) the harddrive is broken. See attached LMSCSI, this benchmark tool can access to raw SCSI harddisk as long as there is an aspi-manager like aspi2dos, no partition required. Try to benchmark the drive, if it operates well, the drive and all communications (cable, termination, aspimanager, ...) is Ok
c) SCSI bus termination, it's not only about the terminator resistors, which also can be installed in the wrong direction in the socket (look at the manual of hostadapter and drive), and check that termpower is enabled and working (jumper, fuse, diode)
d) try older/newer version of aspi2dos.sys. Aspi2dos is a driver not only for 1520, 1522, but also for low cost hostzdapters like 1502, 1505 and so, so during the livetime of these controllers Adaptec released a lot of versions of aspi2dos, and to my experience the later aspi manager drivers may not support older hostadapters of the family. Example: very late aspi4dos only works nice with 1542Cx, but not anymore with 1542A. You will find different versions of EZ-SCSI, having different versions of aspi2dos. What is no problem, ist to combine old version of aspi2dos with newer version of aspidisk, aspicd, and so on. This kind of combination can have some advantage: supportg old hostadapter version, support larger partitions/drives and/or faster
 

Attachments

  • LMSCSI.ZIP
    132.9 KB · Views: 3

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
AFDISK is complaining that the drive interface has been instantiated during boot time. Change the SCSI ID to something other than 0 or 1 and it will no longer be included by the 1522 BIOS. The idea behind the message is that two different interfaces writing to the disk is not a happy situation, as the synthetic geometry may not agree (SCSI disks are LBA addressed only, not CHS. If an application demands CHS, the driver has to dummy something up.)

Changing the ID should result in AFDISK working. You can then either use the drive with ASPIDISK.SYS or remove the drivers and change the ID back to 0.

That's the way I read it, anyway.
First let me say this process worked. So, you posted this at the beginning of the post threads. I didn't have AFDISK to initiate the suggested change. However, I did find the software but didn't change the SCSI ID. My bad. I changed the SCSI ID to 2 and when I started the machine the ROM didn't display the opcode 13h. The APSI2DOS came up but quoted as unsuccessful because the partition hadn't been set by AFDISK as yet. I triggered AFDISK it found the drive and quoted it was SCSI ID 2. It's configuration was the following:
80 cylinders
64 heads
32 sectors/cylinder
80 MB
I accepted the setting and AFDISK put in the partition on the drive then I rebooted. The ASPI2DOS and the ASPIDISK were successful, which made the drive D. I was able to format it with DOS 6.22.

I want to thank all who posted to this problem and it's eventual solution.
 

1ST1

Veteran Member
Joined
Oct 21, 2014
Messages
1,290
Location
near frankfurt/m, germany
I am currently quite confused... aspi2dos is not a driver for logical drives/partitions. aspi2dos (like aspi4dos) is a hostadapter driver (for 152x, 1502, 1505, ..., like aspi4 dos is for 154x and aspi7dos is for 174x and 274x hostadapters, aspiedos.sys for the 294x hostadapters) which provides an aspi software interface. It just is middleware for the hostadapter hardware. Aspidisk.sys docks on that aspi software interface to communicate over the hostadapter to the drive to see/detect logical drives and assign them to dos logical drive letters.

How comes that aspi2dos mounts logical drives at your place? That is the job of aspidisk.sys (as long the drive has not be mounted already by the hostadapter BIOS to int 13h?

There is normally also no matter if you use afdisk or ms-dos fdisk. The result is the same as long as aspidisk.sys is loaded after aspiXdos hostadapter driver.
 

maxtherabbit

Veteran Member
Joined
Apr 23, 2019
Messages
1,906
Location
VA, USA
ASPIxDOS doesn't mount any logical drives, however it does enumerate any drives managed by the HBA BIOS and hook the INT13 calls for them, managing them partially in RAM and partially in ROM.
 

alejack12001

Experienced Member
Joined
Jul 26, 2020
Messages
302
I'm bowing out of this one, folks. I've used Adaptec SCSI adapters for a long time and can usually get one working.
For ISA, the most common one seems to be the AHA-1542CF if that matters.
Everything from a FH HP 9GB 5.25 drive, CDC 330MB FH 5.25", various Seagate 3.5" SCSI drives... down to a Jaz and Bernoulli drives. I've also used SCSI Zip drives as well. I've even imaged a 20SC external hard drive for Apple. Adaptec made some really good controllers.
The original Adaptec AHA-1522 SCSI controller did in fact configure the Conner SCSI hard disk to act is a repository for programs like MSDOS, Windows 3.1, etc. Of course, the drive will not boot because I set it up for the SCSI ID 2. The boot disk is a floppy drive and surprisingly, I got a GoTek 1.44 to be the B drive on this 286 machine.

I removed the 1522 and the Conner drive to experiment with the suggested AHA-1547CF controller that Chuck(G) made mention along with a 4.3 GB hard drive to see if the drive would be recognized by the ROM BIOS on the 1547. When the system starts, the 1547 displays the ROM splash screen giving a number along with version and date. Following is a message stating the pressing CTL-A would get me to a drive configuration screen built into the ROM. If I press the CTL-A as requested, the board doesn't seem to receive the keyboard message and if I wait more than 15 seconds or too long there is an error displayed. In first try, the machine would not jump to the floppy or anything else so I had to reset and start over with the same result, except for the new problem with the floppy boot disk. I was hoping that the machine would make it to the floppy prompt long enough to use the debug command along with the ROM BIOS address to call up the BIOS to configure the hard drive. The boot floppy presented a new error as described below.

The next action seemed strange as the floppy drive on the controller, which was pre-configured from the users manual for the 1547, wasn't recognized either. I got another FDD error along with a F1 resume. Pressing the F1 results in a new error as the controller cannot see the boot floppy drive. Even though the 286 machine ROM sets up the floppies as 1.2 MB for the boot drive and the GoTek as the B or 1.44. It seems like the signature of the boot floppy made under the 1522, which is not recognized by the 1547. Unless this 1547 is a real junker; I don't have a clue to the nature of the problem. I thought about disconnecting the GoTek and tried that with the same result. I didn't think that signatures on boot floppies are unique to controllers; is that possible?

I'd like to see if I could get this to work as a substitute for the old 1522 and Conner drive as the drive is nearing it's 80 MB capacity.
 
Top