The 80 G drive may be a way for them to use those defect platters instead
of throwing them away.

With the low cost of drives now I think i'd just buy the bigger one to begin
with.


"Dr. Uwe Meyer-Gruhl" wrote in message
...
Hi,


now, as new harddisks with more capacity per platter come out, I have
noticed that there are some "strange" models in most manufacturer's lines:

For example, Hitachi has an "80" GByte model in their GXP 180 line,
which maybe should be more properly called "90" GByte (see
http://www.hgst.com/hdd/support/d180gxp/d120gxpmod.htm). What can be
observed is that this model fills the manufacturer's gap at 80 GByte
with 60 GByte per platter. The 60 GByte model has one platter, the 120
GByte model has 2 (and uses them fully), but the "90" GByte model has 2
platters and only 3 data surfaces... so they say (see

http://ssddom01.hgst.com/tech/techli...E3B7F87256C160
06B1CFA/$file/HGSTDesklstar180GXP.PDF).

My suspicion is that this is not the whole truth. I think it is easier
for manufacturers to produce just one type of mechanism and stuff it
with the appropriate PCB or just another firmware to reduce the
advertised model number and capacity. In the Hitachi case, this could be
reached by just using 3 heads or - even better, i.e. more performant -
by reducing the number of cylinders instead.

Has anyone opened such a drive (not neccessarily Hitachi brand) and
knows if the missing "head" is really there? Is it possible to exchange
the firmware with one for the next larger model and re-format the drive
afterwards? The IBM/Hitachi "Drive Feature Tool" seems to exist in a
version with firmware update images included and checks the model number
of the drive found. I have found two firmware upgrade tools by Seagate
for the infamous Barracuda IV series, one for update from 3.x to 3.10
and one for 3.10 to 8.12. The first of those does not check the
preexisting version, the latter does. There seem to be some kind of
certificates involved (.CRT files), which seem to be an indication that
the manufacturers protect their firmwares. Adding to that, few hard disk
firmwares are widely available, other than those for optical drives
(http://www.firmware-flash.com).

So: Is this feasible? Has anyone done it?


Kind regards