PDA

View Full Version : Is this IBM XT-286 deceased?



Iain
February 26th, 2006, 03:37 PM
Now I didn't initially intend obtaining any 286 computer, but in my quest for XT spares an IBM XT-286 system unit presented itself, and as it looked pretty much like a standard XT I thought I'd take the opportunity to acquire it. It contained a 5¼" floppy disk, a 30Mb hard disk and a video card, so seemed to justify itself in spares value alone.

However, before I'd even got my hands on the unit a little research reminded me of some significant differences that I had not stopped to consider. The board layout is quite different, it uses a 16-bit bus and it has a battery-backed BIOS. There was also the matter of it being just a system unit on its own, so not only no keyboard or monitor, but this meant that there was no way of knowing its working condition.

Anyway, once I had the unit I opened it up and checked the video card. It was an MDA type, so I connected up a suitable monchrome monitor, then added a keyboard and powered it up. There was a good deal of humming and whirring from the power supply, then some response from the monitor - but no activity from the disk drives.

The monitor just showed a cryptic "000000 0002 202", and that was it. Repeated switching off and back on resulted in the same response.

I then decided to have another check of the system. A closer look suggested that it has suffered from neglect for some time. As the XT-286 is actually an AT it has a battery-backed up CMOS. A quick examination of the battery compartment confirmed that this computer has not had much tender loving care. The contents of the old 6-volt battery was seeping out of its container.

Unlike the older XT's with their BIOS in ROM the demise of the XT-286's battery meant that the BIOS settings had disappeared. Hence there was not likely to be any disk activity. Unfortunately, a replacement battery failed to alter any of the above symptoms.

This left me with some questions to ponder:

1. Should I have seen some BIOS information come up on the screen as soon as I applied power?.

2. Would a fault with the graphic card prevent the BIOS info appearing but allow that single line of numbers to appear?

3. Does that numeric sequence ("000000 0002 202") have any diagnostic relevance?

4. Would a memory fault prevent the BIOS info appearing?

If anyone has come across anything similar I'd be interested to hear your suggestions.

Erik
February 26th, 2006, 04:04 PM
1. Should I have seen some BIOS information come up on the screen as soon as I applied power?

I doubt anything other than a memory count down (up) would show on screen.


2. Would a fault with the graphic card prevent the BIOS info appearing but allow that single line of numbers to appear?

Your graphics card is fine! :)


3. Does that numeric sequence ("000000 0002 202") have any diagnostic relevance?

Absolutely, I just don't have my docs handy to look it up. You should be able to find someone with the tech-ref or diagnostics for the XT/286 who can look up the code.


4. Would a memory fault prevent the BIOS info appearing?

It may be a memory fault, but the machine is older than most BIOS boots.

Your motherboard may have been damaged by the battery acid, but since it gets far enough to throw an error message I'd say you can probably salvage it, once you figure out what's wrong.

Hopefully someone will be along shortly to interpret that code for ya! :)

Terry Yager
February 26th, 2006, 05:17 PM
201, 20x is a "memory data error", and the number preceding it most likely the address of the questionable memory.

--T

mbbrutman
February 26th, 2006, 06:09 PM
Another thing to keep in mind - IBM AT class systems (which is where an XT-286 fits in) do not have routines in the BIOS for setting the CMOS RAM. You'll need a utility that runs after DOS boots or the diagnostics disk for the machine.

There may be a jumper to clear the CMOS out and put the contents into a failsafe mode so that you can boot the diagnostics disk. (It probably defaults to no hard disk and a 5.25 360K floppy when it gets cleared out.)

You might have to take care of the memory problem. For that I would recommend finding the 'Advanced Diagnostics' disk. The IBM advanced diagnostics usually will pinpoint the failing chip for you. (That assumes that you can get it to run .. a failure in low memory is kind of catastrophic for running programs, including diags.)

Examine the memory chips mand make sure they are properly seated ...

Jorg
February 26th, 2006, 11:44 PM
Reseating the mem chips might be a point.

The AT diagnostics disk you can find here:
http://www.geocities.com/fcmem/5170/
or directly at my webspace http://members.home.nl/charon.styx/
(its called atdg207.exe)
Not sure if it works on the XT-286 though.
You could also try the generic 286 setup disk (gsetup.zip)

Or maybe the XT setip disk whick is here: http://www.uncreativelabs.net/downloads/

Iain
February 27th, 2006, 04:17 PM
Thank you Erik, Terry, Mike and Yorg for all the information you provided.

Erik, you mentioned that all you would have expected to see on the screen would be an indication of the start-up memory checks. The machine didn’t seem to even get that far, it just displayed the numeric sequence "000000 0002 202".

Although the original battery leaked it is in a compartment attached to the system unit’s back panel, so fortunately none of the acid contents actually got on to the motherboard.

Taking into account Erik’s suggestion that it could be a memory problem, and Terry’s confirmation that the displayed numeric code referred to a memory error, this appeared to be the area I needed to investigate further.

The XT-286 memory is in the form of two 30-pin SIMMS, each fitted with nine 4256 chips. I had recently been given a selection of SIMMs and so I tried a couple of these (which only had three chips) in place of the original components. This changed the value of the numeric sequence, but everything else remained the same. A second replacement set of SIMMs did much the same, just producing a slightly different numeric sequence.

At this point I began thinking that the fault might be something more than the RAM, when I happened to come across a small plastic box in the recesses of one of my PC spares drawers which contained four ancient 30-pin SIMMs which I didn’t realise I had. These were of the same 9-chip configuration as the original SIMMs. I decided to give these a try. This time the XT-286 began to come to life!

It displayed the following message:
00640 KB OK
161-System Options Not Set-(Run SETUP)
followed by two beeps, then:
(RESUME = “F1” KEY)

Pressing the F1 key seemed to start the 5¼” floppy going, then the BASIC editor appeared:
The IBM Personal Computer Basic
Version C1.10 Copyright IBM Corp 1981
62940 Bytes free

Initially I thought it was running this from the hard disk, but I had seen no indication of any hard disk activity. Next I tried re-starting with a boot disk in the floppy drive, but this didn’t change anything. I then tried another boot disk – and this time I got the welcome sight of the traditional PC message:
Current date is Thu 1-03-1980
Enter new date (mm-dd-yy) :-

Although entering a year greater than 1999 was beyond it, the machine booted into MS-DOS 3.3 and displayed the A:> prompt!

It never ceases to amaze me how narrow is the divide between success and failure in working on PCs. If I hadn’t found those old SIMMs I would probably have resigned myself to believing that the motherboard was a write-off.

So fellas, give yourselves a well deserved pat on the back – you were right on the mark with your diagnosis. This successfully deals with Stage 1 of the XT-286’s resurrection.

For Stage 2 the computer’s configuration settings have to be loaded into CMOS. Now that the machine can boot up I need to do as Mike mentioned and use a utility to select and store the settings in CMOS. I have downloaded the AT diagnostic disk file from the links kindly provided by Jorg to achieve this.

I now have to fit one of my new machines with an old 5¼” drive to create the diagnostic disk. When I have managed to get this done, and attempted to use it on the XT-286, I’ll report back.

Thanks again for the support.

Erik
February 27th, 2006, 04:43 PM
Congratulations!

That's great news! It's always nice to see one of these machines brought back to life.

Enjoy!

Terry Yager
February 27th, 2006, 05:23 PM
Kewl!

--T

Micom 2000
February 27th, 2006, 08:54 PM
The IBMs used Parity chips to my knowledge. Most regular mem chips won't work.
-------------------------------------------------------------------------------
The XT-286 memory is in the form of two 30-pin SIMMS, each fitted with nine 4256 chips. I had recently been given a selection of SIMMs and so I tried a couple of these (which only had three chips) in place of the original components. This changed the value of the numeric sequence, but everything else remained the same. A second replacement set of SIMMs did much the same, just producing a slightly different numeric sequence.

At this point I began thinking that the fault might be something more than the RAM, when I happened to come across a small plastic box in the recesses of one of my PC spares drawers which contained four ancient 30-pin SIMMs which I didn’t realise I had. These were of the same 9-chip configuration as the original SIMMs. I decided to give these a try. This time the XT-286 began to come to life!

It displayed the following message:
00640 KB OK
161-System Options Not Set-(Run SETUP)
followed by two beeps, then:
(RESUME = “F1” KEY)

Pressing the F1 key seemed to start the 5¼” floppy going, then the BASIC editor appeared:
The IBM Personal Computer Basic
Version C1.10 Copyright IBM Corp 1981
62940 Bytes free
---------------------------------------------------------------------------------------------------------------
That would've indicated that the while the machine booted there were no drives attached. A boot to basic is the default boot up till even most PS/2s
------------------------------------------------------------------------------------------------
Initially I thought it was running this from the hard disk, but I had seen no indication of any hard disk activity. Next I tried re-starting with a boot disk in the floppy drive, but this didn’t change anything. I then tried another boot disk – and this time I got the welcome sight of the traditional PC message:
Current date is Thu 1-03-1980
Enter new date (mm-dd-yy) :-

Although entering a year greater than 1999 was beyond it, the machine booted into MS-DOS 3.3 and displayed the A:> prompt!

For Stage 2 the computer’s configuration settings have to be loaded into CMOS. Now that the machine can boot up I need to do as Mike mentioned and use a utility to select and store the settings in CMOS. I have downloaded the AT diagnostic disk file from the links kindly provided by Jorg to achieve this.
-----------------------------------------------------------------------------------------
While I have an IBM AT and possibly that program, there was another general program to enter the parameters for your lost BIOS on most ATs.
It sometimes was a problem when your HD parameters were unknown because the company was no longer in existence. One can only hope that the new owners of IBM Personal Computers continues archiving with the same diligence that IBM had. IBM-heads should sequester any of their old IBM set-up files.

-------------------------------------------------------------------------------------------
I now have to fit one of my new machines with an old 5¼” drive to create the diagnostic disk. When I have managed to get this done, and attempted to use it on the XT-286, I’ll report back.

Thanks again for the support.[/QUOTE]

Bonne chance.

Lawrence

Iain
March 1st, 2006, 04:21 PM
Progress on Stage 2.

After locating an old 5¼” drive I fitted it to one of the computers on my home mini-network and set it up as drive A: on that machine. I had a bit of a problem with the atdg207.exe file, and did not manage to get it loaded on to a 360Kb floppy disk. An attempt at putting it on a high density 5¼” disk formatted to 360Kb also failed to work; the problem was most likely the way in which the disk was supposed to be read. (Perhaps someone could confirm the precise procedure for this method). The alternative gsetup.exe was far more straightforward, as this runs as a normal executable file under DOS. Upon running this program the setup procedure was simplicity itself – the old XT-286 didn’t exactly have a lot of options to alter.

The main thing to get right was the hard disk model. The hard disk in the XT-286 was a 20Mb (Not 30Mb as I had originally mentioned) WD25R Type 13. Once I had entered this, and added a few other settings, such as time, date and floppy drive type, I was ready to reboot and see if the changes had taken effect.

The warm reboot was followed by the reassuringly standard set of bleeps and on-screen tellbacks, and finally the long-awaited C: prompt! So all the main components of the system appeared to be up and running. From the files on the hard disk, mainly 6Mb of MS-DOS 6.2 and an installation of PCTools, it seems that the machine hadn’t been used since 1995, which tended to be borne out by the general neglected state of the unit. For those with an interest in bringing a particular piece of vintage hardware back to life this is just about as good as it gets – a system unit purchased with spares in mind has turned into a fully working example of the IBM XT-series.

Perhaps not fully working. A cold boot was rewarded with the old “161-System Options Not Set-(Run SETUP)“ warning message. So even though I fitted a new battery the configuration is not being retained in the CMOS. Hopefully this is something minor, like poor battery contact, or something similar. After a long session at the machine it’s time to take a break, and I’m quite happy at the moment to settle for knowing that the system is operational once the CMOS settings have been made. I’ll look into the battery situation tomorrow.

Chris2005
March 2nd, 2006, 07:55 AM
"Although entering a year greater than 1999 was beyond it, the machine booted into MS-DOS 3.3 and displayed the A:> prompt!"

Funny. I remember upon first hearing of the coming woes regarding Y2K, I set the date on my Texas Instruments Professional Computer to 12/31/99, and waited. Sure enough, 1/1/2000 appeared. And this thing was built in 1983. I'm surprised IBM didn't build Y2K compliance into it's BIOS by 1986!

Iain
March 3rd, 2006, 01:47 AM
Perhaps I was a little over confident. I had assumed that I could sort out the battery problem by cleaning up the terminals and everything would sort itself out.

To begin with all went according to the plan. A check with a multimeter showed no battery voltage was reaching the motherboard. The metal of one of the spring terminals in the battery compartment had some surface oxidisation. After a little judicious work with a nail file the battery was able to make a good electrical connection to the terminal and 6 volts was detected at the motherboard connector.

All I had to do now was refit the battery compartment, put the case back on the XT-286 and power up, and all would be well once I had entered the new BIOS settings. That was the first problem. When I switched the machine on it went through its usual memory checks, but then repeatedly attempted to access the hard disk, which it failed to read, and never switched to boot from the floppy drive. I seem to recall when I had looked at the state of the CMOS before making any changes it showed the hard disk as Type 8, and no floppy drive connected. If this was what the system was now able to read from the newly-available CMOS it would explain the behaviour I was observing.

I then opened the battery compartment on the computer's back panel and extracted the battery and re-booted the machine. This time I got the expected "162-System Options Not Set-(RUN SETUP)" message, and once I had pressed the F1 key it booted to the floppy drive. I re-inserted the battery and then ran GSETUP.EXE . I started to make the BIOS entries for the hard drive and floppy drive. One entry I was not sure of was that of "Expansion Memory". This currently had an odd figure of something like 8346. Its value was only changeable in fixed increments from -94 to 15266. (eg. -94/34/162/290/418/546/.. ../15266).

When I re-booted the XT-286 it went through its 640K memory check then responded with:
100000 FFFE 201-Memory Error
164-Memory Size Error-(RUN SETUP)
(RESUME = "F1" KEY)

Pressing F1 allowed the system to load DOS from the hard disk, ending with the C: prompt being displayed. Repeating the process for different values of Expansion Memory didn't make much difference. At this time I had to down tools again and end my session on the XT.

Unless I have developed a fault in the SIMM RAM it would seem that the "Expansion Memory" entry is upsetting the system. Does this refer to memory allocated to the ROMs on the motherboard, or in some extra memory card that I do not have fitted? If the latter, why can't I set the value to 0? Perhaps there is some utility I can find to give me an indication of the XT’s memory map.

If anyone has some experience of these memory settings their insight would be much appreciated.

Terry Yager
March 3rd, 2006, 07:07 AM
The XT-286 should behave as any other AT class machine, no p'ticular XT knowledge is needed. 'Expansion RAM' simply refers to memory beyond the amount that IBM shipped the machine with. The terms 'Expansion Memory' and 'Expanded Memory' usually mean two different things. IBM uses the term just to make non-techies feel st00pit. Your ROMs, video mem, etc are in the area between 640K & 1Mb. In that area is also a 64K 'page frame' for expanded memory, but you probably don't need to worry about that right now. The memory you're dealing with is most likely Extended Memory, which begins at the 1Mb boundry. A bad chip could be the cause of the error you're getting, but also possible is a poorly seated SIMM. Try reseating your RAM and see if that helps.

--T

Iain
March 6th, 2006, 02:32 AM
After going to the trouble of getting together a selection of DOS memory utilities I found that I didn’t need them in the end. While GSETUP.EXE is a very useful CMOS read and write utility for changing the BIOS settings I was still suspicious of the ‘Expansion Memory’ option and the fact that it could only be altered in fixed increments.

Removing the CMOS battery leaves the BIOS in an indeterminate state. When I first removed the battery the system did not recognise any drive and went into the BASIC ROM. (Finding out how to use this feature has been added to my growing ‘to do’ list). Repeating the process of switching off, replacing the battery, then switching on, a couple of times left the BIOS in a state where the Expansion memory was 0k and it booted to the floppy drive. It was then a matter of running GSETUP again to enter all the correct BIOS settings, but leaving the Extension memory value at zero. Once this had been done the XT-286 start-up proceeded correctly, no more memory error messages appeared, and it ended with the line “Starting MS-DOS …” and the C: prompt.

So it is with some relief that Stage 2 of this project has finally come to a successful conclusion. Not bad for a machine virtually abandoned for over ten years. Thanks again to all those who offered their advice and suggestions which led to getting this XT-286 back in action again.

[For other vintage PC owners who may have queries about the memory, I found this passage in ‘Upgrading & Repairing PCs’ published by Que: “Many older IBM systems used a slightly modified 30-pin SIMM, starting with the XT-286 introduced in 1986 through the PS/2 Model 25, 30, 50, and 60. These systems require a SIMM with different signals on five of the pins. These are known as IBM-style 30-pin SIMMs. You can modify a generic 30-pin SIMM to work in the IBM systems and vice versa, but purchasing a SIMM with the correct pin-outs is much easier”.]

Micom 2000
March 7th, 2006, 05:40 PM
I have about 5 of Muellers editions. It is invaluable for collectors. I believe "Check-it" or another which escapes my memory right now also includes a diagram within the memory-check module which pinpoints the chip which is at fault. Can't remember whether it also includes xpansion memory.

Lawrence.

MoonShadow
March 13th, 2006, 03:52 AM
Wow. I have an IBM PS/2 L40sx (it's a laptop) that was given to me by my mother's uni friend, who evidently knew what he was doing when it came to computers. IBM used only to sell 2MB, 4MB and 8MB "memory modules" (RAM) for this kind of laptop. My particular machine has 6MB, and the ram sticks are identical. Also, it has a chip on it (iono which one) that has BASIC on it. guess what it shows when loading?

The IBM Personal Computer Basic
Version C1.10 Copyright IBM Corp 1981
62940 Bytes free
Ok

Freaky, eh? It don't do graphics or arrays, but it's fun to fiddle with. I think what the previous owner did was to take the chip from a machine like yours and burn its contents on to an EEPROM to put in a socket on the lappy's mobo that it attempts to boot from if nothing else is found to be bootable (I don't have an HD that works for it, or the cable's frocked).

Anyways, well done with your restoration of the XT-286, that truly is a great achievement. Have fun with it!

Bye.

MoonShadow

Micom 2000
March 13th, 2006, 09:50 AM
All the earlier IBM had Basic in ROM, including the PS/2s. That was the default when there were no drives attatched so one could access the debugger program.

Lawrence