• Please review our updated Terms and Rules here

Help with IBM XT286

Okay, try the attached file. Unzip and run it to extract to a 360K DSDD 5.25" diskette, then boot the diskette in your XT286.

It will attempt to read every sector on the diskette and report any errors. The diagnostic program is contained entirely within the first sector on the floppy.

Let's see if your diskette drive works.
 
Last edited:
Hi, Chuck,

I have tried extracting to a 360 K drive in my AT but I keep getting a "Diskette error - try another" message. I have tried 5 different floppies, they are all DSDD.
 
I wonder if your AT is set up for a 360K drive (and not a 1.2MB drive). You do know that you have to run the CMOS setup utility when you change drive types, right?

FWIW, I just extracted to a 360K diskette in a 360K drive and one in a 1.2M drive under DOS 6.22 with no errors.

Okay, let's make the disk the hard way.

I'm going to assume that you're running your AT from a hard disk. Unzip the attached file and put it on your AT's hard disk in the same directory that you'll be working in for the next steps. You'll get a 512 byte file called DIAG2.BIN

Format a 360K disk in (let's call it your A: drive) and leave it in the drive.

Run the following command:

DEBUG DIAG2.BIN

you'll get a hyphen prompt. Enter the following command (the hyphen is displayed by DEBUG, you don't type it):

-w100 0 0 1

You should get another hyphen after a bit of floppy activity. Enter the following command:

-q

And you'll be back to the DOS prompt. If your 360K drive was B:, the command above would be w100 1 0 1.

You now have a 360K diagnostic diskette ready to boot. You can perform the same steps on any floppy type, not just 360K.
 
Last edited:
Hi Chuck,

Sorry to take so long to try this, work has been hectic.

I was able to make the 360 K floppy and was successful in reading it. It read all 40 cylinders with the following displayed:

001 SSSSSSSSSHSSSSSSSSSH

through

039 SSSSSSSSSHSSSSSSSSSH
040
All done.

Does this mean the hardware is working?

Thanks, Todd
 
vmfa235 said:
Hi Chuck,

Sorry to take so long to try this, work has been hectic.

I was able to make the 360 K floppy and was successful in reading it. It read all 40 cylinders with the following displayed:

001 SSSSSSSSSHSSSSSSSSSH

through

039 SSSSSSSSSHSSSSSSSSSH
040
All done.

Does this mean the hardware is working?

Thanks, Todd
Click to expand...

Well, it means that your floppy drive is working fine--you succeeded in reading all sectors without a single error.

It also means that your display is working, since you could see the messages.

So we're getting somewhere.

Have you tried booting a really down-level version of DOS (i.e. with a very small footprint)? (PC-DOS 1.1 or even 2.1 would be suitable)
 
Hi Chuck,

I tried booting from DOS 1.1 and I still get the same symptoms. The floppy drive starts to read the disk and then just shuts down after about 3-5 seconds. I tried it with the 360 D drive and the 1.2 M drive. The floppy disk will boot my AT.

Todd
 
vmfa235 said:
Hi Chuck,

I tried booting from DOS 1.1 and I still get the same symptoms. The floppy drive starts to read the disk and then just shuts down after about 3-5 seconds. I tried it with the 360 D drive and the 1.2 M drive. The floppy disk will boot my AT.
Click to expand...

Check your PM; let's take another stab.
 
Last edited:
Anyone have any docs on this unit?

It's starting to look like some sort of strange hardware problem, maybe something on the Mobo, like the timer. Does this thing have dual timer chips?

IIRC, the OP said that the drives work in his AT, I THINK he said the controller work in something else, but none of it seems to work (correctly) in the XT/286.
 
Druid6900 said:
It's starting to look like some sort of strange hardware problem, maybe something on the Mobo, like the timer. Does this thing have dual timer chips?

IIRC, the OP said that the drives work in his AT, I THINK he said the controller work in something else, but none of it seems to work (correctly) in the XT/286.
Click to expand...

I'm beginning to wonder if it's not the DMA Page register. It gets through POST, can read a boot sector, write to the display and read other disk sectors. But a bad page register could cause floppy reads to go to the wrong place.
 
Chuck(G) said:
I'm beginning to wonder if it's not the DMA Page register. It gets through POST, can read a boot sector, write to the display and read other disk sectors. But a bad page register could cause floppy reads to go to the wrong place.
Click to expand...

That's quite possible as well, come to think of it. If we could get some diags into the sucker, we'd know much better.
 
Well, if I can find some time in the next couple of days, I'll throw together something that will check to see that disk reads go in the right place.
 
Dma

Dma

What does it use as the DMA chip, anyway ? I have some new 8237's, I don't mind sending one to the OP to swap out.
Chuck(G), it looks like your diag is calling INT 13h function 02h. It would be interesting to call INT 13h function 08h and see what the computer thinks the drive is.

Wouldn't a bad page register on the DMA 2 channel affect the floppy reads right of the bat ?
Would it be worth while to have the bootstrapper (since OP can at least read 1 sector) test the page register ?

patscc
 
patscc said:
What does it use as the DMA chip, anyway ? I have some new 8237's, I don't mind sending one to the OP to swap out.
Chuck(G), it looks like your diag is calling INT 13h function 02h. It would be interesting to call INT 13h function 08h and see what the computer thinks the drive is.

Wouldn't a bad page register on the DMA 2 channel affect the floppy reads right of the bat ?
Would it be worth while to have the bootstrapper (since OP can at least read 1 sector) test the page register ?
Click to expand...

The drive should default to 360K if the BIOS detects 40 tracks--at any rate, it's possible to read a whole diskette without error. After all, you're supposed to at least be able to boot the setup floppy, even with a bad CMOS.

I could modify things to reset the controller (function 00) and try the read again, even displaying the drive type. 512 bytes is surprisingly spacious for running various diagnostic tests.

That the boot sector can be read to 7c0:0 tells me that the 8237 is probably okay. But if the page mapping reg is stuck at 0 for some reason, any reads past the first 64K of memory are going to be problematic.

I passed the OP a copy of CP/M-86 1.1, just as an additional test. If that fails, then the floppy read logic definitely needs some work. I could even write the CMOS information 10H-31H while I'm at it.
 
I had a Toshiba 286 laptop that had a similar problem. It would start the boot by displaying "Starting MS-DOS.." then just hang at a blinking cursor, and sometimes it gave the error "Divide Overflow"

It's been junked a long time ago, but I seem to remember that the problem was memory related. The RAM bios tests went trough fine, and did not reveal any errors.
 
Pic

Pic

It couldn't be the 8259 in there, could it ?
There aren't any other cards in there, right ?
Does anyone have a BIOS listing for it ?

Just out of curiosity, maybe the OP would like to try turning the machine on, letting it sit for 10~15 minutes, rebooting, and then running Chuck(G)'s test, just on the off chance that it's something thermally related. I doubt it, but it's simple enough to try.

Another thought, it might be worthwhile to disconnect the CMOS battery, let the CMOS drain, and try again with the battery disconnected.

patscc
 
Hi Gang,

Trying to boot CP/M failed, it caused the same symptoms. I have tried disconnecting the battery over night and still can't boot. I will give the thermal theory try and let you know how that goes.

Thanks for all the interest and input,
Todd
 
You must log in or register to reply here.
Back
Top