Old external drive

If this is not seen by BIOS then it won't be seen by OS. I recall old BIOS-es not recognizing the whole drive when drives got larger, OS was able to use these large drives after loading. But I do not recall any cases when BIOS did not see the drive at all and OS could still see it.

Originally Posted by bricem13 ok got the old computer back up. drive in slave, other IDE drive is master. Bios output drive 2 error... not good anyway I got a knoppix live-CD and can get the computer to start. but then, what to do next? there is only gparted on the live-cd and it only recognizes the master drive.

There is nothing to do next. Based on what you've said:

• [ 1764.352846] critical target error, dev sda, sector 0 op 0x0:(READ) flags 0x0 phys_seg 1 prio class 0
  [ 1764.352860] Buffer I/O error on dev sda, logical block 0, async page read
  [ 1764.353527] sd 1:0:0:0: [sda] tag#0 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=0s
  [ 1764.353537] sd 1:0:0:0: [sda] tag#0 Sense Key : Illegal Request [current] 
  [ 1764.353545] sd 1:0:0:0: [sda] tag#0 Add. Sense: Invalid command operation code

• Drive spins at power up and then stops after some seconds.
• I tried in an old computer directly on an IDE connector and got at start a "disk2 failure" (with or without the original IDE drive).
• drive in slave, other IDE drive is master. Bios output drive 2 error... not good

....the drive is dead. You say you have disk failure messages, and things won't recognize it. Can't be much more plain. The fact that you get direct BIOS level messages saying that the drive is having errors eliminates ANY operating system issues.

Well. If you could get knoppix live gparted to see the drive and it's partitions.

You could unmount it and do a file system check graphically with just mouse clicks.

I used to have to do that on old 128MB sd card with ext2 file system with my pupngo linux install when the card was disconnected improperly like battery going dead on my laptop. Ext2 file systems require that when unplugged incorrectly.

https://sourceforge.net/projects/mx-...unity_Respins/

I use Mx workbench respin Linux instead of Knoppix linux.

There's no point really. He;s getting I/O errors, which are pretty terminal.

WD1600AAJB - 00WRA0: 160 GB PATA udma5 (aka UltraDMA/100 aka ATA/100)
This HDD requires LBA48 support.
This means that the old computer (mainboard) used for attaching this HDD must support LBA48 (ATA drives larger than 128 GB). Normally LBA48 support is provided by mainboards with socket-478 or socket-775 (for Intel Pentium 4 processors – mainboards with socket-423 may support LBA48 too but they are rare) or socket-A (for AMD Athlon XP processors – excluded some very early socket-A mainboards for AMD Athlon processors only). Elder mainboards e. g. with socket-370, slot-1 or slot-A doesn't support LBA48 in most cases. They only support LBA28 (ATA drives up to 128 GB). The LBA28 limit isn't avoidable with software e. g. the linux kernel. Trying this will result in massive data corruption even if software shows full drive capacity! All mainboards for x86_64 processors support LBA48.

Originally Posted by Arnulf WD1600AAJB - 00WRA0: 160 GB PATA udma5 (aka UltraDMA/100 aka ATA/100) This HDD requires LBA48 support. This means that the old computer (mainboard) used for attaching this HDD must support LBA48 (ATA drives larger than 128 GB). Normally LBA48 support is provided by mainboards with socket-478 or socket-775 (for Intel Pentium 4 processors – mainboards with socket-423 may support LBA48 too but they are rare) or socket-A (for AMD Athlon XP processors – excluded some very early socket-A mainboards for AMD Athlon processors only). Elder mainboards e. g. with socket-370, slot-1 or slot-A doesn't support LBA48 in most cases. They only support LBA28 (ATA drives up to 128 GB). The LBA28 limit isn't avoidable with software e. g. the linux kernel. Trying this will result in massive data corruption even if software shows full drive capacity! All mainboards for x86_64 processors support LBA48.

sounds interesting, the old computer is a P4 and after some googling it might be the reason... so how can I be sure of this limitation?

Is the other HDD as large or larger, or smaller than, 160G ?Do you have the manual? Does it refer to a maximum disk size of 64G or 128G or 137G? If not, odds are it has LBA48 support. If your P4 is socket 478 or 775, it has LBA48 support.

Originally Posted by business_kid If you have an old PC, I'd remove it from it's external box, remove the jumpers and try it as a slave.

Only following drives should be connected for a first try:

• primary IDE master: the problematic HDD – jumpered to "master or single drive"
• primary IDE slave: none
• secondary IDE master: optical drive (CD/DVD) – live Linux CD/DVD inserted
• secondary IDE slave: none

This setup avoids any interaction between the problematic HDD and the other HDD, especially data loss on the other HDD caused by any "bus (t)error".

Originally Posted by Arnulf primary IDE master: the problematic HDD – jumpered to "master or single drive"

And if his cable is a CS, as was typical in Dells, HPs and Compaqs at the turn of the century and later, what then?

With IDE drives there are 2 different methods of connection.
Some use a flat ribbon cable and use jumpers to identify master and slave on the ribbon.

Others use a special ribbon that often has different colored connectors and a segment of the ribbon is twisted. For those ribbons both drives must be jumpered as master and the location on the ribbon determines master or slave.

The 2 connection methods do not mix, so you have to pay attention to the cables used and to the jumper positions when attaching 2 devices on one ribbon.

Since you are able to boot the machine with knoppix then it really does not matter where the problem drive is connected. You can use knoppix to mount it and read it as soon as the machine will identify it, and the second drive is not needed for access. This simplifies things because you no longer need to worry about master&slave but can use the drive in question as master only if you wish.

Originally Posted by computersavvy With IDE drives there are 2 different methods of connection. Some use a flat ribbon cable and use jumpers to identify master and slave on the ribbon. Others use a special ribbon that often has different colored connectors and a segment of the ribbon is twisted. For those ribbons both drives must be jumpered as master and the location on the ribbon determines master or slave.

That description differs some from my experience. There are no twists in any of the Dell, HP or Compaq 80 conductor CS cables I've run across, though the connectors do vary in color according to position in most cases, whether standard or CS type: Blue for host; Gray for middle or drive 1; Black for end opposite host or drive 0. Often CS cables are obvious due to a hole through one conductor location. Performance can suffer, or be entirely absent, if an only drive on a cable is on the center connector.

Originally Posted by mrmazda the connectors do vary in color according to position in most cases, whether standard or CS type: Blue for host; Gray for middle or drive 1; Black for end opposite host or drive 0.

This describes a standard 80 conductor 40 pin PATA cable that's reqired for udma4 (aka UltraDMA/66 aka ATA/66) mode or higher. These cables are CS capable but CS jumpering isn't required.

Both PATA drives can be jumpered to CS. Then drive connected to black port becomes master (drive 0 as described above) and drive connected to gray port becomes slave (drive 1 as described above).

Drives can be jumpered as master and slave. Then drive jumpered as master should be connected with black port and drive jumpered as slave should be connected with gray port.

Any mix between the two alternatives isn't recommended.

Is the IDE connector installed correctly? You can install them two ways. If it is hooked up backwards, it can cause similar symptoms. Also did you do a visual inspection of the connectors and the pins? Look for oxidation or bent pins or a connector isn't spread. Due to the age of the equipment these aren't out of the realm of possibility. You can use a DVOM and check the resistance of the wires. Granted it can ohm good, but not be able to carry current or data if it has has some green in it, or has one wire left due to bending over the years. For that you will have to put it under a load, but if one of the wires has much higher resistance, that might be the problem. Also the insulation may look fine, but when you test it, it might be open. Other than this I might suggest a professional data recovery service if the data is important.