PDA

View Full Version : AT&T PC 6300 Revisited



clh333
April 20th, 2017, 09:33 AM
I acquired an AT&T PC 6300 recently; it arrived yesterday. Having owned one back in the 1980s I am "sorta" familiar with it, but back then it was just an office appliance. The farthest I got into the hardware was to add a DTC controller and a Conner HD. I passed it on a few years later for a newer PC with 3.5 inch drives and a color display.

So now I have a survivor of that era, and I have gotten as far as removing the CPU's top cover. Already I am asking questions:

In addition to the video card and 2 half-height 5.25 FDDs there are two 8-bit cards in the backplane: One is a Plus HardCard 20, which I'm guessing is an HD on an ISA card.
Anyone got documentation for that?

The other card is a DAK VG-100 and appears to be related to video; there are two RCA jacks labeled video in and video out. No idea.
Anyone recognize this?

I know the motherboard is upside-down under the backplane that holds the cards and that the video card is a connector for the two. I would like to check for dead battery and possible cap leaks.
Can I remove the bottom-half cover or will I have to remove all the internal components to access the motherboard?

The keyboard is missing one key cap and switch (lower right corner, RETURN as I recall).
What is the switch mechanism, and does anyone know of a source for spares?

Thanks for your replies.

-CH-

Pics:

37903 37904

1ST1
April 20th, 2017, 11:01 AM
Yes, the mainboard is "downstairs" and you have to check the battery. The keyboard is special, you won't get spare parts for it, except of from another same keyboard. The Plus HardCard 20 should be a harddisk (so called filecard), I guess it is having 20 Megabytes, but that is not original AT&T/Olivetti part, but some kind of third party add on.

You can download all service manuals etc. from the following sites:
- https://sites.google.com/site/att6300shrine/Home
- http://www.olivettim24.hadesnet.org/download.html

Please note, that AT%T 6300 is originally an Olivetti M24. And Xerox 6060 and Logabax Persona 1600 are the same as well. So you can use software, drivers, manuals, etc. from all of them to learn about this machine.
-

lambda84
April 20th, 2017, 12:35 PM
Can't say much about those cards, but to access the underside, just remove the two captive flat head screws on the back bottom corners and the bottom part of the case should slide off. To remove the motherboard, remove the screws and standoffs holding it in place, take out the video card that came with it, and I believe unscrew the thick red and black power cables, and the it should come out without a problem. I don't think much is on the underside of the motherboard, so just removing the bottom case should reveal everything you want to check without taking out the whole board.

clh333
April 20th, 2017, 12:43 PM
Yes, the mainboard is "downstairs" and you have to check the battery. The keyboard is special, you won't get spare parts for it, except of from another same keyboard. The Plus HardCard 20 should be a harddisk (so called filecard), I guess it is having 20 Megabytes, but that is not original AT&T/Olivetti part, but some kind of third party add on.

You can download all service manuals etc. from the following sites:
- https://sites.google.com/site/att6300shrine/Home
- http://www.olivettim24.hadesnet.org/download.html

Please note, that AT%T 6300 is originally an Olivetti M24. And Xerox 6060 and Logabax Persona 1600 are the same as well. So you can use software, drivers, manuals, etc. from all of them to learn about this machine.
-

Thanks for your reply.

I know (now) that the 6300 was a re-branded Olivetti. I had never seen the add-in cards contained within and still don't have any idea what the DAK card was for, but apparently somebody listed another one on eBay for $200. I'm pretty sure he doesn't know what to do with it either.

I have a bad habit of saving useless stuff for "someday" so I have the original manuals for the AT&T and also the original distribution disks, containing MS-DOS 3.x and the AT&T utilities and help files. Including the mouse driver there are eight disks in all. So I diskcopied the boot disk and booted the machine; all systems passed POST. So far, so good.

Thanks again for your input.

-CH-

1ST1
April 20th, 2017, 01:05 PM
Anyhow, check the battery and take it out. It only holds the date/time, nothing more. But these batteries tend to leak and destroy the board.

Chuck(G)
April 20th, 2017, 02:01 PM
Good Lord, a DAK board--from Drew Kaplan during his heyday, nonetheless. I used to love to browse his catalogs and actually bought some stuff from him back in the day. Here's a sample catalog (https://ia902707.us.archive.org/28/items/1988-Winter-DAK-Catalog/1988-Winter-DAK-Catalog.pdf)

The VG-100 card doesn't ring any bells, but you can drop a line to Sol (who took over from Drew when he retired) at dak.com (http://www.dak.com). If Sol doesn't remember, maybe Drew will.

clh333
April 20th, 2017, 03:07 PM
Good Lord, a DAK board--from Drew Kaplan during his heyday, nonetheless. I used to love to browse his catalogs and actually bought some stuff from him back in the day. Here's a sample catalog (https://ia902707.us.archive.org/28/items/1988-Winter-DAK-Catalog/1988-Winter-DAK-Catalog.pdf)

The VG-100 card doesn't ring any bells, but you can drop a line to Sol (who took over from Drew when he retired) at dak.com (http://www.dak.com). If Sol doesn't remember, maybe Drew will.

I was wondering of that was the same company... the logo looked familiar. What the hell were they doing manufacturing PC cards?

I used to get their catalogs all the time, too. I probably bought something from them but I can't remember what.

Thanks for the link; I'll ask.

-CH-

clh333
April 21st, 2017, 02:42 AM
Anyhow, check the battery and take it out. It only holds the date/time, nothing more. But these batteries tend to leak and destroy the board.

Motherboard Rev. P4, ROM Rev. 1.43, 8087 on board, battery had leaked but perhaps not enough to damage the traces.

Thanks for your assistance.

-CH-

Chuck(G)
April 21st, 2017, 08:03 AM
Fortunately, the battery is "upside down", which makes it a bit more likely that leakage will affect the motherboard less than in "normal" designs. I replaced mine with a lithium coin cell in series with a schottky diode (lower forward drop than a normal silicon diode). Works just fine.

clh333
April 21st, 2017, 11:03 AM
Fortunately, the battery is "upside down", which makes it a bit more likely that leakage will affect the motherboard less than in "normal" designs. I replaced mine with a lithium coin cell in series with a schottky diode (lower forward drop than a normal silicon diode). Works just fine.

Good to know, although I will do without the battery for now.

Found that Plus Card 20 was the second offering from a wholly-owned Quantum subsidiary who pioneered the drive-on-a-card product in the mid-80s. Found a link to some docs and software utilities here: http://ibm-pc.org/firmware/other/plus%20hardcard/plus%20hardcard.htm. Appears I have BIOS version 5.x; second version of the board.

No reply from Sol or DAK yet.

Thanks,

-CH-

clh333
April 26th, 2017, 02:55 AM
Good Lord, a DAK board--from Drew Kaplan during his heyday, nonetheless. I used to love to browse his catalogs and actually bought some stuff from him back in the day. Here's a sample catalog (https://ia902707.us.archive.org/28/items/1988-Winter-DAK-Catalog/1988-Winter-DAK-Catalog.pdf)

The VG-100 card doesn't ring any bells, but you can drop a line to Sol (who took over from Drew when he retired) at dak.com (http://www.dak.com). If Sol doesn't remember, maybe Drew will.

I heard from Drew Kaplan who was kind enough to respond,

"Hi Charles, So I think it was my PC Soul Snatcher video grabber card. I’ve attached a black and white close out ad. I’m sure I have the color somewhere, but this will explain what it did.

Have a great day

Drew".

Essentially this is a 256x256 greyscale digitizer board. The DAK ad is attached.

I also found threads in the VCF archive discussing the Plus 20 Hardcard. Like others I was getting, after a long pause, a controller error, indicating that the BIOS was functional but there was some problem with the drive mechanics. I disassembled the drive and found the gooey stop that others had found. I cleaned the post and replaced the rubber with a bit of heat-shrink tubing, then reassembled and reinserted the card. This time I quickly got a different error. The drive now sleeps with the fishes.

-CH-

Ad:

37973

blaknift
April 27th, 2017, 04:15 PM
I finally got my 6300 That I have been waiting since early April for. I'm going to build a VGA adapter cable and see if It will output video to an old CRT. Mine is in fantastic shape, And someone had taken the battery out long ago and added wires out of the front of the case to hook one up! Thing is an absolute beauty inside.

Mine also came with a hard card! Western Digital Filecard20.

Is what i'm seeing next to the floppy drives another Hard drive?

I just gotta find a keyboard for this damn thing....Lol
38003380043800538006380073800838009

1ST1
April 28th, 2017, 11:39 AM
It looks like it has the Toshiba floppy drives. They are 80 tracks DD. Officiallly Olivetti sold them as 640 k drives (8 sectors, 80 tracks, double sided), but they are also capable for 9 sectors (720 kb). I have the Olivetti Service manual for these drives. You only need to set the dip switches for these drives (80 tracks) and use driveparm to tell it about the 9 sectors.

clh333
April 28th, 2017, 12:02 PM
Is what i'm seeing next to the floppy drives another Hard drive?


Looks like it, although I've never seen that arrangement / case. But this is where I intend to put an HD when I get that far. Having the two 5.25 drives is not something I want to sacrifice.

Anyone familiar with this setup?

-CH-

Trixter
May 1st, 2017, 06:56 PM
I would like to check for dead battery and possible cap leaks.

Please remove the battery; it only holds time and is not necessary for operation of the system. (The clock used only 8 bits for the year anyway; it only keeps time from 1984-1992, then rotates around again)


Can I remove the bottom-half cover or will I have to remove all the internal components to access the motherboard?

Consult the documentation; the bottom half is removable separate from the top half. It slides forward.


What is the switch mechanism, and does anyone know of a source for spares?

The key switch mechanism was unique (it's not buckling spring, ALPS, Cherry, etc.). You must get spares from another keyboard or an Olivetti M24 keyboard.


Looks like it, although I've never seen that arrangement / case. But this is where I intend to put an HD when I get that far. Having the two 5.25 drives is not something I want to sacrifice. Anyone familiar with this setup?

Yes, this was common for people adding hard drives to a 2-drive setup. In the AT&T 6300 Retrospective (https://youtu.be/mUCh46_MzZU) video, you can see a third variant, someone who cut a hole into the case to use the side bracket as a 3rd floppy drive.

Chuck(G)
May 1st, 2017, 07:23 PM
Mine came with an 80MB hardcard (although not the Quantum brand).

clh333
May 2nd, 2017, 02:53 AM
As I mentioned earlier the Plus 20 card has been abandoned as FUBAR. The plan is to try to get a SCSI drive installed in the space next to the FDD cage. I obtained a Seagate ST01 8-bit controller for that purpose; I have a 170 Mb Quantum with a 50-pin SCSI I interface that I would like to use with it.

The hard card would have been an easy solution but apparently the early ones weren't all that robust.

-CH-

Stone
May 2nd, 2017, 03:27 AM
I have a Quantum 170MB hardcard if that's of interest (Sorry, I see that's a 16-bit ISA card).

http://www.vcfed.org/forum/attachment.php?attachmentid=38228&d=1493724320


I also have a Seagate ST-238R drive/WD 1002-27X controller combo that I think I removed from a dead 6300 15 or 20 years ago. Anyway, it's running fine and already configured so it's basically plug'n play. :smile:


38228

1ST1
May 2nd, 2017, 08:15 AM
If you swap to another drive I would sugest to go for XT-CF controller.

clh333
May 2nd, 2017, 09:03 AM
Thank you for your offer but there would be no room for anything else in the drive bay with the full-height ST-238 installed.

-CH-

Stone
May 2nd, 2017, 09:07 AM
Thank you for your offer but there would be no room for anything else in the drive bay with the full-height ST-238 installed.

-CH-A Seagate ST-238R isn't a FH drive. :-)

It's basically an ST-225 that has passed RLL certification.

blaknift
May 2nd, 2017, 02:52 PM
I got her fired up today with the adapter cable and sure enough she works a treat! Now I just need a keyboard. Is adding a 3.5" Floppy drive simply a case of just plugging it in and taping over a 1.44MB Formatted drive?

clh333
May 5th, 2017, 04:53 AM
Wishing to retain dual 360k floppies, I obtained a Seagate ST-01/2 SCSI controller with the idea of mounting a SCSI drive vertically next to the drive cage. I configured the ST-01/2 as well as I could from the documentation I found on line, but when I booted the machine with the card inserted it hung. The POST recognized the ST BIOS at C800:0000 but the controller apparently was looking for a bootable drive and couldn't find one (there being nothing attached to it at the time).

I removed the controller board and drive from the PlusCard 20 Mb and modified the frame to accommodate a Quantum SCSI drive and mounted that into the frame. With a SCSI drive attached to the controller the BIOS recognized a drive was attached and allowed boot to proceed from FD A:.

At this point DOS did not recognize C: as a valid drive. I ran DEBUG with the parameter "G=C800:5" as noted in the Seagate documentation and selected an interleave of 5 from a range of 1-9. Formatting commenced and concluded successfully after a few minutes. After rebooting the controller again recognized the SCSI disk, but DOS did not. I ran FDISK and it produced an "error 2003: Integer divide by 0" message and terminated.

I'm not sure what to do next. Can anyone suggest a strategy? I think I need to get a DOS partition and format on the drive before I can do anything about loading the OS, but I don't know how to get past this error. The drive came from an eBay reseller, BTW; I know nothing about its former life. Try another drive? Reconfigure the controller?

Thanks for your suggestions. Pictures attached.

-CH-

38252 38253 38254 38255

1ST1
May 5th, 2017, 09:24 AM
What's your setting about harddisk on the M24 mainboard dip switches?

clh333
May 6th, 2017, 03:23 AM
What's your setting about harddisk on the M24 mainboard dip switches?

Switch 3 in bank 1 (HDU ROM code select: on mainboard = ON, on controller = OFF) is OFF. There was a hard-card installed when I purchased the machine and apparently was configured then.

After yesterday's post I downloaded Disk Manager 5.x (small enough to fit on a 360K floppy) and ran their program. At first I selected the "easy" configuration, which created two partitions of 100 Mb and 35Mb. Low level format ran but at the end the program announced there was a problem with the partition table and the partition should be deleted. [A fair amount of unsuccessful tries later] I discovered that DOS 3.x can only deal with partitions 32 Mb or smaller so I manually created 4 30 Mb partitions (the most the program would allow) and re-formatted. Same results; the program reported that the partition was incorrect and the drive unusable.

However, there was now a "C:" drive recognizable by DOS; something not present before. But attempts to access the drive (from the C:\ prompt, for example) resulted in a similar division error. At the same time I began to see "Illegal interrupt" messages on re-boot. They were not in the C800: range but lower; they scrolled by and I did not have a chance to write them down.

I began to think I might have a problem with the 8087. When I subsequently attempted to boot from a DR-DOS 6.0 boot disk I got another math-related error (from SuperStor, which apparently was attempting to decompress a file). Boot with DR-DOS was unsuccessful on the AT&T, but the same disk did boot a '586 (which I had to resort to using because the "tweener" decided it would no longer recognize the USB ports; capacitor replacement indicated).

So I pulled the 8087, but the results were the same once again. My next move will be to test the Quantum drive installed on another machine attached to a known-good controller and see if I can format it there.

Thanks for the suggestions.

-CH-

Stone
May 6th, 2017, 04:15 AM
Disk Manager is a DDO and isn't necessary (or useful) on a drive of that size.

The LLF process does not in any way access/use a coprocessor. No program that isn't specifically designed to use a coprocessor does.

Old hard drives can die just sitting around. Would you like to see my knee-deep pile of them? :-)

clh333
May 6th, 2017, 06:31 AM
Disk Manager is a DDO and isn't necessary (or useful) on a drive of that size.

DDO?


The LLF process does not in any way access/use a coprocessor. No program that isn't specifically designed to use a coprocessor does.

Thank you for clarifying.


Old hard drives can die just sitting around. Would you like to see my knee-deep pile of them? :-)

No need; got a few of my own.

Attached the Quantum SCSI drive to the Adaptec SCSI controller in this machine (Win7 64 bit PCI / SATA) and the controller recognized, formatted and verified the drive. With the drive still attached booted Win7 and using disk manager created a DOS partition. I think the drive itself is OK; but still doesn't work in the machine.

In the process I managed to scare myself: Changing configuration DIP switches and adding and removing components I began getting flaky video on reboot. Then after one power-up the machine did not display anything, and boot did not proceed. After disassembling, checking and reassembling, the next power-up the display was scrambled but I could see the drives being accessed. I hit the reset button and the screen synch stabilized but the screen was partial garbage. I could see characters, however, and the boot concluded with an A> prompt. I typed CLS and the display was normal after that. Powered down and rebooted and everything was normal.

I'm thinking there was / is oxidation on the video synch connection that caused this. It may be that the boot process first checks for a stable display before it proceeds, but I'm just guessing here.

Pictures attached.

-CH-

38285 38286 38287

Stone
May 6th, 2017, 07:49 AM
DDO?Dynamic Drive Overlay (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwju26vOztvTAhUISCYKHXriC3kQFggiMAA&url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FDynami c_drive_overlay&usg=AFQjCNHKRj8EUx3gZXmMeIELm8yM0px4hw&sig2=bZ68PKrvOWjc594R8DQDgw&cad=rja)

clh333
May 6th, 2017, 10:26 AM
After leaving the machine off for an hour or two I came back and powered up. Again the display was irregular, but in a different way than before (picture attached). Where the previous anomaly appeared to be out-of-synch, in this time the display looked as though the timing of the horizontal sweep (and maybe the start of the top line of the display) was off. Characters were readable, but in the wrong place. The machine booted to an A> prompt.

Once again I was able to power down, wait and power up. After a few resets the display stabilized into its normal configuration. I no longer think this is an oxide-resistance issue: it appears to me that either a capacitor is failing (capacitance out of spec but warms up?) or one of the ICs that controls display timing is failing.

As always, suggestions welcome.

Thanks.

-CH-

38294

Trixter
May 6th, 2017, 09:53 PM
I ran FDISK and it produced an "error 2003: Integer divide by 0" message and terminated.

Try running "fdisk /mbr", and if that succeeds, run it again to see if it works. If that fails, boot a DOS 5+ disk (like MS-DOS 6.22) and try the same thing.


I think I need to get a DOS partition and format on the drive before I can do anything about loading the OS, but I don't know how to get past this error. The drive came from an eBay reseller, BTW; I know nothing about its former life. Try another drive? Reconfigure the controller?

You're not doing anything wrong, just running into some bad luck. For a system of this vintage, with its various compatibility issues, I would try to stay as simple as possible: Try to use a 20 or 40MB MFM drive with the internal controller if possible. If not, try a small SCSI drive (under 100MB). Try running AT&T-branded MS-DOS 3.3 and keep each partition to 32MB or less just for testing.

Worse comes to worse, you can order one of the many XTIDE variant cards (http://www.vcfed.org/forum/showthread.php?36816-Keeping-Track-of-All-the-XT-IDE-Variants) and use that with a compact-flash adapter (or a real IDE hard drive). The adapters must run in slow/8-bit mode with the 6300, but they do work.



Once again I was able to power down, wait and power up. After a few resets the display stabilized into its normal configuration. I no longer think this is an oxide-resistance issue: it appears to me that either a capacitor is failing (capacitance out of spec but warms up?) or one of the ICs that controls display timing is failing.

It is unlikely that the display ICs are failing; it is much more likely that a capacitor is failing. Fortunately, that kind of thing can be easily serviced.

If you want to remove the monitor itself from your troubleshooting, you can build a simple VGA adapter for the 6300: ftp://ftp.oldskool.org/pub/drivers/ATT/6300/video_converter_output_to_VGA.pdf
That can help you continue your troubleshooting while eliminating the monitor as a potential issue.

clh333
May 7th, 2017, 04:25 AM
Try running "fdisk /mbr", and if that succeeds, run it again to see if it works. If that fails, boot a DOS 5+ disk (like MS-DOS 6.22) and try the same thing..

I downloaded a set of DOS 4.01c 360k disks. I was going to try installing that if I couldn't get going under 3.3. I read that DOS 4.0 and greater could handle 2 Gb at the same time I found out DOS 3.3 could only handle 32 Mb. But I'll try FDISK / MBR first.


You're not doing anything wrong, just running into some bad luck. For a system of this vintage, with its various compatibility issues, I would try to stay as simple as possible: Try to use a 20 or 40MB MFM drive with the internal controller if possible. If not, try a small SCSI drive (under 100MB). Try running AT&T-branded MS-DOS 3.3 and keep each partition to 32MB or less just for testing..

To quote Albert King: "If it wasn't for bad luck I wouldn't have no luck at all." I wanted to use SCSI partly because the controller and drive are more compact. I had a DEC / Conner 100 Mb SCSI drive that used to work but yesterday threw an error. All the other SCSI drives are larger, and many do not have the requisite 50-pin interface. All of the 8-bit MFM and RLL equipment I have seen take up quite a bit more room, and as I still want to run two drives; that means the drive and controller have to occupy board space and possibly cover slots.


Worse comes to worse, you can order one of the many XTIDE variant cards (http://www.vcfed.org/forum/showthread.php?36816-Keeping-Track-of-All-the-XT-IDE-Variants) and use that with a compact-flash adapter (or a real IDE hard drive). The adapters must run in slow/8-bit mode with the 6300, but they do work.

Got two, from glitch. That was Plan C.


It is unlikely that the display ICs are failing; it is much more likely that a capacitor is failing. Fortunately, that kind of thing can be easily serviced.

I was hoping for an obvious sign like a leak or tilt but have yet to investigate.


If you want to remove the monitor itself from your troubleshooting, you can build a simple VGA adapter for the 6300: ftp://ftp.oldskool.org/pub/drivers/ATT/6300/video_converter_output_to_VGA.pdf
That can help you continue your troubleshooting while eliminating the monitor as a potential issue.

I was going to investigate whether the 6312 color monitor had the same connector as the 6300 but I'll look at the VGA adapter instead.

My thanks to all for their suggestions.

-CH-

Trixter
May 7th, 2017, 09:12 AM
I downloaded a set of DOS 4.01c 360k disks.

FYI, avoid DOS 4 like the plague. Any other version of DOS that came later provides only benefits without any downsides. I recall that 4.01c does fix all of the known bugs, but if you can run something clearly better that takes up less RAM (like MS-DOS 6.22 or IBM PC DOS 7.x), why not?


Got two, from glitch. That was Plan C.

I would get one of those set up (again, make sure 16-bit transfers are configured off in the BIOS that you flash to them, as the 6300 reverses the byte order of 16-bit reads due to an incompatibility) in a known working system, then transfer the card over to the 6300 and see what you get. Once you know your 6300 is 100% working with the XTIDE card, you can then go back to troubleshooting just the SCSI.

clh333
May 7th, 2017, 10:03 AM
I reinstalled the ST controller and the Quantum disk. Once again I began getting "illegal interrupt" errors (see pic attached) but DOS 3.3 boot concluded successfully. I ran FDISK /MBR, which encountered an error and terminated.

I rebooted, this time with the DOS 4.01c installation disk, and got part way through the installation before it, too, encountered an error and terminated. I rebooted DOS 3.3 and re-ran the Ontrack Disk Manager program, with same results as before.

I tried to install the XTIDE card I got from glitch but due to a clearance issue was unable to install the card. I removed the ST controller and Quantum drive.

It's hard to know what the cause of the problem is, especially with the concomitant video issue, but my guess is that the ST controller is either mis-configured or defective. My documentation for the controller is pretty sketchy so I think I'll have to find better documentation to know for sure. In the meantime the Quantum can be used elsewhere.

On further consideration trying to use the 6312's color monitor attached to the 6300 does not seem like a good idea, as the 6300 is sending power to the monochrome monitor. I'll try building the adapter shown in the link.

Thanks again,

-CH-

38329

1ST1
May 7th, 2017, 10:38 AM
If the M280 video connector is the same as on M24, then you can use the monitor. There is also a color monitor for M24 and this one also has it's own power supply and does not use the M24 graphics card as power supply.

clh333
May 7th, 2017, 03:36 PM
If the M280 video connector is the same as on M24, then you can use the monitor. There is also a color monitor for M24 and this one also has it's own power supply and does not use the M24 graphics card as power supply.

Thank you for the suggestion. However, it's looking more and more like the problem is closer to the motherboard.

I was able to resolve the XTIDE clearance problem and after determining its ability to boot on an IBM XT I moved it over to a slot on the AT&T. The machine recognizes the XTIDE and will boot from either the A: drive or the XTIDE (when it boots correctly, that is), although there isn't much more than DOS 6.22 and a few XTIDE utilities on the flash card.

The garbled display on startup is getting worse, in this sense: The screen bar pattern seen before persists for the first few resets, and the characters are still in the wrong place, but now the character set is becoming garbled as well (CHR$ > 127). The cursor flits all over the screen and sometimes the A:> drive light will come on but not boot. After a few minutes - and several resets plus maybe a power cycle - the machine will settle in and complete a boot.

This is looking like a timing issue and it also seems that it is affecting more than just the display. I'm wondering if it is somehow related to the system clock. More investigation is required.

Thanks again,

-CH-

Trixter
May 7th, 2017, 08:25 PM
Either your RAM or capacitors are going bad. It's not a timing problem (unless related to the caps).

Chuck(G)
May 7th, 2017, 10:02 PM
Well, I'd probably start by cleaning and res-eat all of the boards, including the video cards, making sure that the power supply connections are nice and tight.

clh333
May 8th, 2017, 02:45 AM
Good advice, thanks. I'll try that first.

-CH-