[comp.sys.ibm.pc.hardware] Why are SCSI drives so cheap?

scjones@thor.UUCP (Larry Jones) (08/31/90)

In article <481.26db8352@sunset.sedd.trw.com>, rose@sunset.sedd.trw.com writes:
> 
> Also, why are SCSI drives so much cheaper (at least some are)?  There's an
> 84 MB with kit for less than $400.

Primarily because having the controller right on the drive allows you
to get away with things you could never get away with when the low
level signals from the drive have to travel across a foot of ribbon
cable in an electrically noisy environment.  The Seagate 296N (which
I presume is the drive you are referring to) is the same disk as the
251-1 (which is a 42Mb MFM drive) -- since the controller is so close
to the drive they can reliably use an RLL encoding to double the number
of sectors per track.
----
Larry Jones                         UUCP: uunet!sdrc!thor!scjones
SDRC                                      scjones@thor.UUCP
2000 Eastman Dr.                    BIX:  ltl
Milford, OH  45150-2789             AT&T: (513) 576-2070
I've got PLENTY of common sense!  I just choose to ignore it. -- Calvin

rose@sunset.sedd.trw.com (09/08/90)

In article <149@thor.UUCP>, scjones@thor.UUCP (Larry Jones) writes:
> In article <481.26db8352@sunset.sedd.trw.com>, rose@sunset.sedd.trw.com writes:
>> 
>> Also, why are SCSI drives so much cheaper (at least some are)?  There's an
>> 84 MB with kit for less than $400.
> 
> Primarily because having the controller right on the drive allows you
> to get away with things you could never get away with when the low
> level signals from the drive have to travel across a foot of ribbon
> cable in an electrically noisy environment.  The Seagate 296N (which
> I presume is the drive you are referring to) is the same disk as the
> 251-1 (which is a 42Mb MFM drive) -- since the controller is so close
> to the drive they can reliably use an RLL encoding to double the number
> of sectors per track.
> ----

If the 296 is simply a 251-1 with a SCSI controller can you buy the SCSI (ST-01)
and the 251-1 seperately?  The 251-1 with out controller is approx $250-300 
and the ST-01 is $30.  This would save an additional $50-100 (if allowed).

The actual drive listed is a 296NP which is a half-height drive with
controller.

Thanks for any info

Marc

stevewa@upvax.UUCP (Steve Ward) (09/08/90)

In article <488.26e76a41@sunset.sedd.trw.com> rose@sunset.sedd.trw.com writes:
>If the 296 is simply a 251-1 with a SCSI controller can you buy the SCSI (ST-01)
>and the 251-1 seperately?  The 251-1 with out controller is approx $250-300 
>and the ST-01 is $30.  This would save an additional $50-100 (if allowed).
>
>The actual drive listed is a 296NP which is a half-height drive with
>controller.

No, there is a BIG difference.

The ST-01 is NOT a controller.  It is a SCSI host adapter.  This may seem like
a difference of semantics, but in reality it is a BIG difference in function.

A host adapter is akin to a translator.  It converts a disk request from
your computer's peripheral bus into a SCSI command.  The controller, which
is integated with the drive electronics, then executes the command.

This is different from the "traditional" MFM/RLL setup, where the command
is executed by the controller card in the computer, which actually controlls
the disk drive's stepper motor and other parts to find the required data.

IDE is similar to SCSI in that the controller electronics are located on
the drive itself, though the actual interface is quite a bit different
(IDE is much closer in design to the AT bus than is SCSI, which was an
independent development).

Why would you want the controller on the disk instead of on a seperate card?
In the case of SCSI, where a bus architecture is used (components are daisy-
chained together), significant performance gains can be achieved.  Since it
usually takes longer for the drive to execute a command than it does for the
host adapter to issue it, many additional commands could be sent to other
SCSI devices while the computer is still waiting for the disk to finish.
This is why SCSI is very popular with file servers where multiple users are
accessing data from many different places at the same time. An MFM/RLL
controller would have to wait for the first operation to complete before
sending another one, even if the second operation was on a different drive,
because the controller performs the operation directly.

Hummph.  That wasn't as clear as I'd wanted.  Hope it helps anyway.

Steve
-- 
| Steve Ward Jr. appears courtesy of       |            stevewa@upvax.UUCP    |
| Univ. of Portland, Portland, OR          |         !tektronix!upvax!stevewa |
| (insert disclaimer here)                 |  upvax!stevewa@tektronix.TEK.COM |
| --If all else fails, try:      tektronix.TEK.COM!upvax!stevewa@uunet.uu.net |

scjones@thor.UUCP (Larry Jones) (09/09/90)

In article <149@thor.UUCP>, I wrote:
> The Seagate 296N is the same disk as the
> 251-1 (which is a 42Mb MFM drive) -- since the controller is so close
> to the drive they can reliably use an RLL encoding to double the number
> of sectors per track.

In article <488.26e76a41@sunset.sedd.trw.com>, rose@sunset.sedd.trw.com writes:
> If the 296 is simply a 251-1 with a SCSI controller can you buy the SCSI (ST-01)
> and the 251-1 seperately?  The 251-1 with out controller is approx $250-300 
> and the ST-01 is $30.  This would save an additional $50-100 (if allowed).

No.  The ST01 (like all SCSI controllers) is >not< a disk controller --
it's a host adapter that interfaces your computer to the SCSI bus.  The
disk controller is actually a part of the disk drive.  When I said that
the 251 and 296N are the same disk, I meant the actual disk (the heads,
platters, actuator, etc).  The drive electronics are completely different.
----
Larry Jones                         UUCP: uunet!sdrc!thor!scjones
SDRC                                      scjones@thor.UUCP
2000 Eastman Dr.                    BIX:  ltl
Milford, OH  45150-2789             AT&T: (513) 576-2070
I'll be a hulking, surley teen-ager before you know it!! -- Calvin