Hard Drives
Install RAID Drive
0661-371   320 MB, 3.5x1.6, 4317.8 RPM 
0661-467   400 MB, 3.5x1.6, 4316 RPM 

0662-Sxx    1.05 GB, 3.5x1, ???? RPM

0663-E1x   "Corsair" 1.0/1.2 GB, 3.5x1.6, 4316 RPM  What's the difference between the -E15 and the -E12?

0663-H1x/L1x    1.0 GB, 3.5x1, 4316 RPM What's the difference between -H and -L?

0664-x1H   2.0 GB, 3.5x1.6, 5400 RPM 

DSAS-3720720 MB, 3.5x1, 4500 RPM

SCSI   Position    SCSI   Position
 ID    1   2   3    ID    1   2   3
 6    Off ON  ON     2   Off ON  Off
 5    ON  Off ON     1   ON  Off Off
 4    Off Off ON    0   Off Off Off
 3    ON  ON  Off

WDS-3160   160MB, 3.5x1, 3600 RPM ???

Hi Charles !

>Could you elaborate or point to a reference on this aspect of HDD
>operation with IBM?  IBM make a point that certain points of the
>HDD need to be isolated or that others need to be grounded, and
>I've never been able to get a great handle on the "text" of the
>message & combine that with a graphic that points to different
>spots on the drive. 

I sure can. The 0661 / 0662 / 0664 drives are 1st generation Magneto-Resistive Head drives and the platter cage (the black case inside the drive assembly) has to be ground-free not to interfere the signals under the heads. The heads are driven via a differential driver and grounding the platter-cage would disable the heads from reading any information. At least one of the fixing holes is a bit oddly positioned and a fixing screw longer than 5mm would be one or two threads too long and touch the platter cage, cut through the black paint onto the metallic cage ... and ground it.

Result: the drive will not come up ready, make "marble bouncing sounds" since the lower "servo" head cannot read the tracking informations - and after some time the drive drops the platter motor and a SCSI error code appears.

When you look closer at the drive you will find large black plastic spacers
holding the platter cage in the drive assembly. No screws. Taking apart these drives is a very inferiour job ... most of the screws are 6 or 7 TORX,
printboards are stacked in 3 levels, the motor-controller sits in the front
corners of the outer drive cage and is soldered on flexible circuits. The heads amplifier and the heads cable are attached with a conductive rubber piece with inlayed wires ... 
The whole thing is a darn complex construction. And to make things worse: drive mechanism and drive electronics are paired during manufacturing. So if you have two slightly defective drives you have a 1:50 chance to make one working drive from out of the two bad ones.

>Anyways.  Maybe it's not that complicated.  Why should IBM's
>drives be different than anybody elses? 

They aren't. But IBM is a professional company and has a good public image on reliability. Sometimes caused by the additional care they put in efforts of proper grounding and avoiding electronic interferences where possible - try to get rid of problems right from the start. The additional grounding *can* be ignored - drives will run fine without that ... but in very, very worse cases it might give additional security.

>Do us users just screw this up accidentally all the time? 

Users are the computers only natural enemies.  They are able to louse up even "fool proof" constructions.

>I've no idea of how worried I should be about this.  A real problem, or a

The above-mentioned drives need some caution and care - the later ones are mainly uncritical. With the introduction of the DSAS and DPES series IBM returned to other conceptions of the R/W-amplifiers and different screening of the drive. These are rock-solid general purpose drives, which need no special treatment.
The DFHS / DFRS family of server drives developed from the 0664 is also uncritial about grounding / power sourcing - but has some problems with heat emmission.

Don't forget that the 0661-design is almost 8 or 9 years old. IBM was the
inventor of the MR-technology and the "Spitfire" (a.k.a. "Havant") drive was the first commercially available MR-drive at all. Other companies like Seagate, Quantum and Micropolis later adopted the technology of MR-heads ... while IBM developed the GMR-heads technology (Giantic Magneto-Resistive) as used in the DFHS / DFRS.
Somebody *has* to bring out the first product - and that was IBM at that point. The *experiences* however made the customers ... and the drives were very soon known as "inreliable" - which is only half the truth. 

My 9595-AMF for example still has the 400MB 0661, P/N 73F8958, FRU 82F0012 it came with. Drive is made by IBM (UK), EC 894769.

Compared to other drives it has:
- a thicker white insulation layer between last rear board and platter cage
/ next printboard
- there are insulating stripes under the lower board at the rear to avoid
contact with the computers case
- a large white plastic insulator is sticked from the underside of the
platter-cage to the outer frame, for what purpose.
- the fixing screws are only 3.5mm short (about 0.15")

Still going strong after some billion rounds :-)
Very friendly greetings from Peter in Germany

Can't find it here? Go to the source- IBM Hard Disk Drives

Sometimes IBM doesn't have it either, so go to The Ref

Please email with your IBM Drive type and I'll dig it up and post it...

9595 Main Page