8MB RAM cards

Hi Jay

I had the advantage (?misfortune ) of being "on the case" a couple more pages than you...

...this is not a "normal" PC! - there's on-board SCSI to cope with - unfortunately we haven't been able to track down the motherboard, is Intel manufacture for some OEM (unknown). So I'm the first to admit I'm fumbling in the dark...

==

"prime mover" symptom:
message during POST "Int 13h device not found "
which would seem to mean that it can't talk to the HDD correctly

possibilities:
- a 528MB-type drive overlay was involved and has got lost along the line;
- the SCSI capability is nobbling the IDE drive somehow;
- something to do with an OEM hidden partition which has got lost
- CHS/LBA thing
- probably others too...

the 1st and 4th of these could be disproved if we could try with a HDD less than 528MB.

==

it might be worth trying with an early WD Data Lifeguard toolkit, I don't think the later ones would work. Trouble is I don't think WD distribute the older stuff anymore. I shall have a look around.

WD DLG toolkits are bootable floppies, it's a different sort of bootable floppy to the DOS or Windows SUD, might well "sail through" adverse conditions - they are designed to cope with a variety of OS and misfiring drives. Might make the difference.

suggesting this really because the HDD is "almost" getting there but not quite. If it's the only device on the ribbon this can't be the slave interfering - so it has to be either down to BIOS setting(s) or something amiss with the HDD (but both HDDs are doing the same thing)

Award BIOS (not the same as yours I know, Harpo - but 'tis the closest I can get at the moment) offers the following options for HDD mode:
LBA
LARGE
AUTO
NORMAL

I suppose NORMAL must be what I was calling CHS

==

whoa gotta get some kip ( = stayed up too long!) will return when recuperated to see how it's getting on.

best wishes, HJ

 

The WD utilities would be the ones to use if they are available, if not try Maxtor's Maxblast, it is quite flexible with other makes of drive. Be careful not to install EZ-BIOS though.

Matt

 

Thanks for all the input. I will not be able to get back to the computer until tomorrow evening, and you've given me much to chew on! Please stay tuned...

 

Hi

tomorrow sounds good - got some things I want to check + a bit busy...

best wishes, HJ

 

Hi there,

If you mean model #, yes, it matches.

Yes. As follows:

CHS is auto-configured (default), user defined, or disabled.
- current # of cyl = 1416
- current # of heads = 16
- current # of sectors = 63
- maximum capacity = 697MB

No specific option for LBA. However...

IDE Translation Mode options are
- Standard CHS
- Logical Block
- Extended CHS
- Auto Detected

Multiple Sector Setting options are
- Disabled
- 4 Sectors per Block
- 8 Sectors per Block
- Auto Detected

Fast Programmed I/O Modes options are
- Disabled
- Auto Detected

ISA IDE DMA Transfers options are
- Disabled
- Auto Detected

 

Check the top of the drive and make sure the cylinder head sector setup in the bios matches it. Many of these early ones you had to manually enter it and do it correctly or it won't work.

 

Hi Harpo

LBA is "Logical Block (Addressing) "

==

your last post makes it look like we aren't worrying about a 528MB motherboard limit - which narrows things down (a little bit)

I think it's pointing us back towards
- the SCSI stuff interfering somehow
- hidden OEM partition gone missing.

==

...hoping that by now you've been able to get hold of the old version Data Lifeguard Tools somehow assuming that you have... these tools can check out the interface to the HDD and also can "prepare" the HDD which is like FDISKing it and Formatting it. But the other reason for using them is (I'm hoping) they will ignore whatever fault condition is preventing the W95 & DOS6.22 boot disks from working.

try the BIOS Check thing first:
put floppy in & boot computer: hit return at the splash screen; you end up at a menu of options, Bios Check is among them.

Select BIOS Check and when it's finished whirring, choose to "Display Data" (you get the chance to save a log file again later)

the first and last screens are the important ones:
- the first screen will tell you whether the BIOS is controlling the HDD OK in whatever mode you have selected (Standard CHS, LB, Extended CHS)

but even if it says it's controlling OK, you still have to go on through lots of complicated lookiung stuff until you get to the last two screens, which are titled Int 13h Function 08h Data & Int 13h Function 48h Data

look at the last column on these screens called "Capacity" and check the size it is showing. If it shows the drive has shrunk to 528MB then it means you need to use a different mode.

You should then get a chance to save a log file - to the floppy! - which is a plain text file called DISKDATA.LOG - pls post the first & last bits (as described above) if anything puzzling has occurred.

==

of course this all rather depends upon the floppy actually being able to boot the computer! Here's hoping.

best wishes, HJ

 

from the WD datasheet:

Recommended Setup Parameters
Cylinders 1416
Heads 16
Sectors/Track 63

==

very important to jumper 4 to 6 (or remove) with these earlier WD HDDs, if it's the only device on the ribbon cable - they are seriously picky about jumper settings.

BW, HJ

 

Thanks for this additional info.

What would cause SCSI to start interfering?
How could a hidden partition go missing?????

Yes, I have a WD diag disk that has identified a cable problem. And I'm afraid the one thing I didn't replace to the original was the cables. I installed round cables just after acquiring the machine, and the computer DID work after installing the round cables. However, the original cables went into a bag with all my other cables, so now I'll never know which ones came with the machine. But since it WAS up & running before swapping out hardware, it probably isn't too important to restore the original cables - only to get the computer to LIKE the cables that I've put in - but then, what set off the "I don't like these cables" syndrome in the first place??? <sigh>

HJ, I haven't had a chance to follow these instructions yet, sorry. Will get back to you on this, probably tomorrow.

Currently, the config I'm working with is just one (or the other) HD, the floppy, and a video card. No sound card, no CD-ROM.

Thanks for sticking with me! Will post back...

 

Hi again

oh heck - cabling

but from what I read earlier up "daisy-chained off the CDROM" you had it the right way round ie HDD (the master device) at the end of the cable and CDROM in the middle

with (newer) 80 way cables, the master device has to be on the end socket (black) and the slave device on the middle socket (grey)

with (older) 40 way cables, this arrangement was preferential but not mandatory

there may be a setting lurking in the BIOS along the lines of "is an 80-conductor cable installed? "

==

good luck - let us know what happens

best wishes, HJ.

 

It sounds like you would have checked, but is the cable itself the right way round, that is, if you stretch a cable flat, the two connectors for the drives are more towards one end, the connector on the far end goes to the motherboard.
I have read that this is important, but always check, so I don't know the consequences of putting the cable in the wrong way round

I have a box of cables (not a bag ). When I went to get one for a system I was making I noticed that one seemed strange. I found that it was actually reversed to normal cables and I suspect that it came from an OEM machine.
...You may have to find those original cables.
Edit: just to explain, the tongue and groove (the lug) on the connectors meant that pin #1 on the motherboard went to pin #40 on the drives.

Matt

 

OK. My answers are all mixed up with the questions, and if I've missed any pertinent questions, please refresh my memory

And the answer is.....
The HD labels both say 730.8MB - I'll change this.

Ummm, I don't remember seeing in in BIOS, but in the WD diagnostics, yes - with an "IF" qualifier.

Yes

Unsure how to proceed with the cable problem... Should I focus on resolving it in particular, or continue with troubleshooting the problem in general, assuming the "cable problem" is possibly related to another "inconsistency "?

 

Hi

if it were me I'd be concentrating on the cabling now; the manufacturer's tools say there's a cable problem, they probably mean it. Won't get any sense out of the thing if it cannot communicate to the motherboard properly.

your CHS values match what WD datasheet says, so don't worry too much about the megabytes (I hope)

if I put those CHS values in I get 730MB

I think this is the difference in how megabytes get measured: sometimes 1 megabyte is 1,000,000 bytes, sometimes it's 2^20 = 1,048,576 bytes

multiply 697 by 1.048567 gives you 730.857472 which is close enough for me!

==

out of the available modes, choose "Extended CHS" (as first guess) - I don't think we are going to hit a 528MB mobo limit because of the way the BIOS "understands" E-CHS

(you may perhaps have to go to LB mode to get it to play; if you do expect the CHS figures to change, probably to 708/32/63, that's normal)

try E-CHS first though

==

can you locate the original cable it might help; I think Matt may well be onto something there...

**edit**
have just seen the clarification about the pin1 to pin 40 thing - if that were the case you wouldn't be getting as far as you are, surely - the BIOS is autodetecting the CHS numbers correctly, can't see that happening if the whole cable's backwards'd?

but it does go to show there are "funny" OEM cables floating around


best wishes, HJ

 

oops

sorry Harpo - just read it through again and saw

evidently the bit that's not functioning properly at this end is the bit between the monitor and the chair...

could you let us know what the WD error message says, and is there any error code or anything else which would help narrow it down pls?

best wishes, HJ

 

How do you guys KNOW this stuff?????

Here's the error message:

"Extended Test
DLGDIAG detected errors while testing the cable for the currently selected drive. The following steps should be taken.
-Verify that the IDE cable is properly connected to the drive and the system board.
-Verify that an 80 pin cable is installed if data transfers are being performed with UDMA66.
Press C to continue with the test or press any other key to return to the main menu. "

Continuing the extended test did not reveal any further errors.

When it came time to create a log file I got this error message:
"Fatal error: Could not create log file. "

Here's notes I took:

IDE drive 1: Regular BIOS operation failed.
System BIOS is NOT controlling your drive properly.

Int 13h #: 80h

Regular Int 13h # of cyl: 707. # of heads: 30. # of sectors/track: 63. Total sectors: 1425312=C*H*S. Capacity: 729MB

Where to from here?

I'll continue trying different cables, etc. The current cable in there is an old one with no external "grooves" or "tab" or whatever it's called, and no missing pin hole. The red line on one side is almost undetectable...

Another old cable in my bag actually has a small "hole" in it - a place where it looks like a 1/4" section is missing from one of the wires. How odd.

 

this at least I can answer for you

what you have there is a "DIY" style cable select cable. It's a fudge to convert a 40 conductor cable so that you can use cable select setting on the drives.

these cables break the "normal" rules: with one of these you put the master in the middle and the slave at the end

assuming the cable is 40 conductor, the wire that's broken is #28 (count from the red stripe as #1), this is the cable select wire

mobo-----master--x--slave

x is the break

the master "sees" the motherboard (where #28 is grounded) and seeing ground, decides it's a master; the slave doesn't "see" anything because the wire's broken, and hence configures itself as slave

==

we now have two things telling us that the motherboard cannot "talk" to the HDD correctly:

and the WD tools; clearly this is what needs sorting out first

Harpo - check me pls - the computer worked with one of these AC2700's; you removed the drives; then refitted them and now it won't play any more:

While the HDDs were out of the machine, did you connect them to anything else? - anything that might changed their contents?

(I know you've attempted to FDISK one of them since, but it didn't get far enough to alter the contents...)

(sorry if the answer's already in the thread somewhere)

==

the way I'm thinking:
when you refitted the HDDs the BIOS tried to autodetect - it's got the CHS values correct, but because you have a drive bigger than the 528MB, the mode will matter also. And whatever mode the BIOS has auto-chosen isn't suitable.

we should try with a couple of different modes

the "other" settings should be:

Fast Programmed I/O Modes
- Disabled

ISA IDE DMA Transfers
- Disabled

Multiple Sector Setting options are
- Disabled
- 4 Sectors per Block
- 8 Sectors per Block
- Auto Detected
yikes! - probably leave this on Disabled this has the feeling of something that could mess everything up properly

==

CHS is auto-configured (default), user defined, or disabled.
- current # of cyl = 1416
- current # of heads = 16
- current # of sectors = 63
- maximum capacity = 697MB

starting from this, change the type to "User Defined ", then change the mode to Logical Block

don't panic if the C/H/S figures "grey out" and alter, hopefully to 708/32/63; make a note of what you end up with

==

OK now try booting with the WD floppy

watch the BIOS power-up messages to see if the Int13h device not found is still there

see how far you can get with the WD floppy

from your last run:
Regular Int 13h # of cyl: 707. # of heads: 30. # of sectors/track: 63.
whoa! - that's very very wrong.

make a note of what you get this time round, hopefully it won't be 707/30/63

good luck

best wishes, HJ

 

Correct. And I might have put them into the "games" machine briefly (which is now working just fine), but I don't remember for sure.

When you said

I was confused about setting the C/H/S figures to 708/32/63??? so I set them to 1416/16/63 as these #s were previously confirmed as being good??? They did not grey out when I changed the IDE Translation Mode to Logical Block.

When I rebooted, the computer did not recognize the keyboard, and the Int 13h error is still there.

 

as long as you didn't try to format them or FDISK them in another machine it should be OK; what I'm trying to cover is that if there was something OEM specific on the drive originally, it should still be there

sorry I should have been clearer, I didn't mean you should enter the figures 708/32/63; rather, that when you change mode to LB the figures might get changed for you to 708/32/63 most likely, and not to worry if this happens - just try going with what it says.

The screens you get depend upon what sort of BIOS it is. I've managed to find a WD2850 to play with, but I haven't got a machine with AMI BIOS here to see what happens.

==

if the keyboard's locked then LB was clearly not a good choice!

try repeating the operation selecting E-CHS translation mode

I think we should try all the available modes, clearly what auto chooses isn't working - so it's only sense to work through the other options as "user select" option based upon the 1416/16/63 starting point which we know is good

some drives don't play in some modes

best wishes, HJ

 

I had earlier today had to switch the mouse/keyboard because on boot neither was detected. So after receiving this message, I switched them back before trying LB, but while keyboard/mouse were both detected, no other changes were evident, except that the WD floppy wouldn't read. After changing the setting to E-CHS, the WD floppy boots, but I'm still getting the cable error message, etc.

I noticed the log file was closed, so I tried to open it and got this message:

The log file is currently closed.
The drive has been reinitialized so the log file must remain closed.

**
Note: all day I've been working with the HD that has a jumper in (I think) 4/6 - if I look at the HD from the top, where I can read the label, the jumper is set to the far left.

I tried hooking up the 2nd HD. This one has no jumper. Here are the messages I got with the Quick Test:

DLGDIAG detects 1 hard drive.
If 1 one is not the expected count, then the following steps should be taken for each installed drive:
-Verify the IDE cable is properly connected to the drive and the system board.
-Verify that the power connector is properly attached to the drive.
-Verify that the jumper settings for the each drive are correct.

Immediately after posting, I'll try the previous HD w/jumper set to the far right instead of the far left.

I'll keep trying. I'll try some different cables, too.

 

Jumper set to the far right (from top view) DEFINITELY doesn't work! I get "invalid system disk" error for WD floppy or no floppy.