IBM Thinkpad 700C

[Psyberducky]

I can confirm that my CL57SX has a "WD-280" hard drive in it. Someone on Twitter (link) confirmed this is specifically a "DBA" (Direct Bus Attachment) ESDI drive as I described above. Since they found it in their ThinkPad 700/720, it should be compatible!

[ThinkDan]

I am going down now to do some shaking and wobbling and will let you know how it goes...

Hopefully hold on to the drive whilst you do that

Just for giggles, I went off to find the original service tip where I learnt about stiction.

b) If drive does not spin with power applied, turn system off and remove drive from system. Attempt to rotate the drive in a clockwise direction 90 degrees at a time repeatedly. This may free the heads. More inertia will have a better chance in freeing the drive heads. Do not drop the drive or touch drive electronics, as drive is sensitive to G forces and electrostatic discharge. Attempt to twist the drive 3 times and then install to see if drive is able to spin.

The technique we ended up describing was to take the disk right side up, hold it horizontally in right hand against your chest with your arm and elbow horizontal, sweep arm out from the elbow quickly and then give the drive a quick horizontal flick of the wrist away from you. That should do the clockwise rotation. I had to do this a year or two ago when the 820 hadn't been powered up for a while and its 2.5" SCSI drive wouldn't spin up...

[ThinkDan]

I can confirm that my CL57SX has a "WD-280" hard drive in it. Someone on Twitter (link) confirmed this is specifically a "DBA" (Direct Bus Attachment) ESDI drive as I described above. Since they found it in their ThinkPad 700/720, it should be compatible!

Great - I couldn't quite remember if the controller was on the drive or the motherboard on these - i.e. whether reseating the drive was real help or just a placebo! Direct Bus Attachment rather suggests that the ESDI controller is integrated on the drive itself and presents as an MCA device, and is plugged into an MCA bus connector rather than an ESDI connector. These MCA laptops really were nicely made.

[solidpro]

Just a minor update.

I did lots of swinging and the 700C drive spins but doesn't read.

I found a WD-280 on the shelf but it's 1/3rd thicker than the drive sled that goes into the 700C.

So, not a lot of progress...

[ThinkDan]

Had a 'Doh!' moment today when I recalled what IDE, ATA, and ATAPI stand for: Integrated Drive Electronics, AT Attachment, and AT Attachment Packet Interface. So IDE/ATA/ATAPI drives were plugged into essentially an extension of the AT bus, just like ESDI DBA drives into the MCA bus of the 700/720.

This got me thinking: if you had two drives in front of you, one ESDI/DBA and one IDE, from the same manufacturer and with the same capacity and geometry, could you then swap the PCB of a failed ESDI drive onto a bare formerly-IDE drive? With either type of bus attachment, the drive PCB must contain both the bus interface (IDE or MCA) and the drive controller (logical block requests <> physical drive access). Many early IBM portables offered 80MB, 120MB, and 160MB 2.5" drive capacities... could this even be a way to assemble working 2.5" SCSI hard drives for the PowerPC laptops out of contemporary 810MB or 1.2GB IDE equivalents, assuming you had a dead SCSI drive to hand?

[Psyberducky]

Now that you mention it, I have a later production L40SX (from 1992) whose IDE hard drive sounds identical to the ESDI one in my (1992) CL57SX... the WD-280 drive has a very distinct startup/idle sound. My other L40SXs are earlier production (1991), and therefore had failing Conner hard drives in them instead, which is why this particular one stood out to me. I don't remember the exact model number of the newer L40SX's hard drive, but I remember it being very close or identical to "WD-280".

[ThinkDan]

I'd expect that the IDE and ESDI equivalents of the same drive mechanism would certainly have different FRU and part numbers showing, possibly differences on the name labels like "WD280" too. They would have to be visibly differentiated so the right FRU was ordered/dispatched.

What counts is if the drive chassis, PCB connectors, and so on are identical as well as the stated capacity - in which case I'd expect to see manufacturing numbers on castings etc. that were identical, or in very similar ranges. It needs somebody to swallow a brave pill and dismantle two working drives to swap the PCBs across to prove or disprove what is, for now, just speculation.

There are three levels of XXAYYYY IBM part numbers:

• Part (or option) numbers (for a boxed product or accessory)
• FRU (Field Replacement Unit - parts listed in HMM and orderable through dealers and service agents)
• manufacturing part numbers

Most of the IBM manufacturing part numbers are undocumented, despite following the familiar XXAYYYY format, unless they happen to be listed as a FRU. I have a 360C that has a damaged LCD cable, I have the manufacturing part number on it but google finds nothing; the part was never orderable as a distinct FRU.