Real SCSI Hard Drive recommendations?

[Ed.D]

Hey everyone,

Can anybody please suggest a quiet, half height, 50 pin SCSI hard drive for the Amiga? or is SCSI and quiet a contradiction in terms?

Doesn't need to have large capacity. I have some drives that I've been playing with and these are noisy or clunky, listed below.

HP C2490A #Sounds like a jet engine.
Compaq DGHS18Y #Very loud coil latch.
Quantum DSP3107LW #Noisy spindle and loud coil latch.

I do have a coupe of very nice E-Server, X Series 2.5inch drives, 80 pin, but I can't get these working on the 25 pin to 50 pin cable I made + 50 to 80 pin adapter because (I think) I need to find a way to terminate the extra unused wires from what I've been finding on the net.


Thanks,
Ed.

 

[fitzsteve]

I have one, by Seagate IIRC. I will tell you the model no. later. It was pretty quiet IIRC. I keep it as a spare since it's half height and 4gb 50 pin it's a great backup drive.

 

[Ed.D]

Thanks Steve

 

[BLTCON0]

Hey everyone,

Can anybody please suggest a quiet, half height, 50 pin SCSI hard drive for the Amiga? or is SCSI and quiet a contradiction in terms?

Noise (or heat) isn't an issue with SCSI. Designing a cool & quiet SCSI drive those days would be a pointless waste of resources, given the real target group.

Doesn't need to have large capacity. I have some drives that I've been playing with and these are noisy or clunky, listed below.

If you must, grab a Quantum Fireball ST (or maybe SE) SCSI version. They were made circa 1997, narrow-only (it would be pointless to have a wide version given the buffer anyway). It's actually the same IDE drive Quantum made, just reinterfaced for SCSI. 5400 instead of the usual 7200 RPM, so much quieter, and ~128 KB instead of the usual 512 KB buffer, so much less usable. But on the Amiga they'd be more than OK, you shouldn't even have to touch interleaving to get proper performance.

I do have a coupe of very nice E-Server, X Series 2.5inch drives, 80 pin, but I can't get these working on the 25 pin to 50 pin cable I made + 50 to 80 pin adapter because (I think) I need to find a way to terminate the extra unused wires from what I've been finding on the net.

It's not a matter of termination actually, although it can be resolved that way for some drives. But these drives don't provide such an option (and they should really NOT do so - they aren't meant to be installed with common PC-case cabling but on dedicated SCA backplates which take care of all the mundane SCSI details, so it's the right thing that they lack the options as it greatly simplifies proper installation). I'm afraid with a SCA LVD drive you will rarely if ever get anywhere close to success trying to elegantly run it on an old SCSI 1 SE bus, which by your mention of the 25 pin connector is what I assume you're trying to do.

My best advice: Learn to ignore the noise.
My second best advice: Track down some IDE-to-narrow SCSI adapter (Acard etc).

 

[Ed.D]

Thanks for the advice BLTCON0,

What's strange is I can get the amiga to see these E-Server disks if I hot plug them (no flames please :roll: ) and then do a rescan of the scsi bus in the Blizzard Unit Control program. I understand that some drive parameters can be pre-set and stored via a scsi controller, so I might look into this as I have a couple of interfaces I can try in a PC.

I can't ignore the noise sorry and I can't afford Acard's prices.

Cheers,
Ed.

 

[protek]

I/O Data really should make a new batch of R-IDSC-E/R. Hope I'm not breaking the forum rules much by saying that never have I seen a new piece of computer equipment experience such a price inflation in such a short period of time (4 years) as I have with the R-IDSC-E/R. I'm really regretting I sold mine, when I sold it.

 

[Ed.D]

Or perhaps a new breed of A1200 accelerator with DDR2/DDR3 and a Sata interface.

 

[BLTCON0]

Thanks for the advice BLTCON0,

What's strange is I can get the amiga to see these E-Server disks if I hot plug them (no flames please :roll: ) and then do a rescan of the scsi bus in the Blizzard Unit Control program. I understand that some drive parameters can be pre-set and stored via a scsi controller, so I might look into this as I have a couple of interfaces I can try in a PC.

I can't ignore the noise sorry :smile: and I can't afford Acard's prices.

Cheers,
Ed.

Yeah, Acards indeed cost an arm and a leg. And of course noise tolerance varies widely among people

Hot-plugging the drives on an SE bus is of course illegal and requires capital punishment but let's focus on the interesting part here.

[SCSI babbling]
It's worth noting that it's not really the termination that does the trick in such cases, because termination doesn't concern a specific device but the bus itself. I.e. it's the bus that must be terminated at both ends, not the devices which reside at the ends. But it's obviously very convenient to include termination on a device, so that IF said device resides at the end of the bus it can legally (being at the end) terminate the bus.

What a wide SCSI device is actually looking for during intialisation is the presence of some bias voltage on the SCSI lines. A wide SCSI device may simply "AND" the status of all 16 lines, therefore without the voltage on the high lines it'll return 0 and won't start up. The proper check should be more complex to include the possibility for narrow operation (i.e. IF (AND 0-15 == 1) THEN mode = wide ELSE IF [(AND 0-7 == 1) & (AND 8-15 == 0)] THEN mode = narrow ELSE do_not_report), but not all wide drives do that.
A wide-device-on-a-narrow-bus adapter with high-byte active termination will provide said voltage on the high (8-15) lines thus convincing the wide device to start up.
But such an adapter also needs termination power, which AFAIK the blizzard SCSI kit doesn't provide, so you must also have a device (any one device) on the SCSI bus provide termination power (look for a TERMPWR enable jumper).
[/SCSI babbling]

So perhaps the deficient status checking procedure is bypassed via hotplugging, but even so some testing should be done to verify the drive is fully operative afterwards.