(e-mail address removed) (Bob) wrote in server.houston.rr.com:
NB: I crossposted this to <comp.sys.ibm.pc.hardware.storage> so that
some real experts can comment. That's why I included the last post so
they can see what the problem is.
--- Snip
I wonder what would have happened if you put the second drive on the
other IDE cable.
The other IDE channel has optical drives on it. I can't say that I want to
d/c those opticals to run a hard drive. The other option there is running a
hard drive and an optical on the same IDE channel. That can bring out other
gremlins and glitches. I've learned the hard way that data is safer and
transfers are faster if I keep hard drives mated with hard drives on the same
IDE channel along with opticals mated with opticals on the same IDE channel.
I've already got a PCI-IDE adapter to run other optical drives as well. Yes,
I've got lots of drives, but, I've got the power to run all of it. PS is 550
watt with 30A on the 12V rail and 18A on the 5V rail. I really don't think
it's a power supply problem because, truthfully, after I connected the drives
straight to the motherboard, there have been no problems whatsoever. Full
surface scans of each drive have been run at least 5 times over the last 2
weeks and no bad blocks. Just pure perfection as long as those removable
drive bays are not incorporated into the configuration.
Oh, I forgot to say, it's gets wierder. I was going crazy trying to figure
out why my computer wouldn't regenerate from a state of hibernation. Lo' and
behold, now that the hard drives are connected straight to the motherboard,
this computer will return from hibernation. With both of those removable
bays connected, the computer reported that it could not establish previous
state when starting up from a state of hibernation.
I triple-boot to W98SE, W2K, and WXP. Both W2K and WXP would fail to return
from "hibernation" with those two removable bays connected. W98SE? I don't
know how it would have responded to "hibernation," because I only use it for
making boot floppies and for hosting PowerQuest DriveImage and Partition
Magic.
Once I ripped out the removable bays and connected the drives straight to the
motherboard, I thought I'd try the hibernate function. Sure enough, removing
the bays fixed my hibernation problems. Computer now hibernates and returns
from hibernation without a glitch. I've used the hibernate state probably 10
to 15 times and it now works perfectly.
Yesterday, however, I tried re-installing the KF-20 drive bays. Immediately,
at first POST, BIOS couldn't find the IDE drives. The POST stalled while
trying to enumerate connected IDE drives. So, I took out the bays, scratched
my head, and Googled the heck out of the internet. Can't say that I've found
any one any where who mentions situations like mine. The only thing I've
found is one KB article at Western Digital which gives very few details.
That KB article says, however, that "it is not uncommon" for removable drive
bays to cause problems.
---
The latest post:
Yes, they all do. It's supposedly the 80-wire ribbon connector that
makes it ATA compatible.
Well, actually, not all of these bays use that "Centronics" connector.
There's a model called KF-101 that uses an internal connector with 64
cavities. (It seems that these model numbers are shared by many of the
Chinese/Taiwanese manufacturers. It's like your KF-23 and my KF-20. The
same designs are sold under a whole bunch of different brand names. I
suspect one factory in Taiwan sells these to all sorts of companies and
brands them accordingly. So, I just happen to be looking at KingWin KF-101
bays at NewEgg, but, I believe the same connector specification would exist
in any removable bay designated as KF-101. And, by the way, I'm not touting
NewEgg. It's just that NewEgg gives some really large detailed views of the
insides and outsides of the drive bays systems they have for sale. I can
examine bays there and see the connectors for most models.)
The KF-101 connector has 64 cavities where each cavity houses a very small
connector pin and each pin is shrouded in it's own cavity of a plastic
enclosure--much like a standard IDE cable arrangement utilizes. That
connector in the KF-101 looks like an IDE or SCSI connector.
There's also a SanMax branded model P96i that utilizes the same 64 pin cavity
connector as the KingWin KF-101.
SanMax is, for some reason I can't figure out, is calling theirs a "96 pin"
connection even though it's a 64 cavity connector. I don't know how they
calculate 96 pin functions into that array unless 80 "pins" are considered to
be the 40 IDE connections with their associated ground wires + 16 utilized
pins for something else involved with the LED's, power, and so forth.
On the other hand, there's your KF-23 and my KF-20 with the "Centronics"
connector. That "Centronics" connector has relatively very large metal
terminals that are in an "open array." Nothing segregates one connector from
the next except an air gap. (The "Centronics" connector is used for parallel
port connectors, SCSI connectors ((slow-SCSI)), and so forth.)
I'm only wondering if some "cross-talk" occurs in those Centronics connectors
on my system? I know that as SCSI got faster, it seems that specs eliminated
the Centronics connector. I'm wondering if I might benefit by eliminating
that Centronics connector in this IDE array?
I'm just wondering if these new WD drives are implementing ATA to the limit
and maybe all my data corruption and drive clicking problems stem from the
incorporation of the "Centronics" connector in these KF-20 bays?
I'm just about to order some removable bays with that "new" 64 cavity
connector to see if that cures my ills. If it does, I'll post in this NG.
I'm hoping, however, someone in the NG community has experienced this before
me and I'm hoping that person will shed some light on this situation before I
waste about $120 U.S. on a bunch of bays/trays that might only going to fail
me again.
He only has one. He opted to build the OS on a smaller drive and put a
250GB in the removeable tray. He has other drives with stuff on them
so he bought some extra trays.
Let's see what the experts have to say about this.
I almost went for a dual bay system like yours until I found the
Enermax 352 unit (which uses only one IDE channel).
Maybe two removeable bays on one IDE channel is too much to ask in
terms of ATA compatibility.
Yes. I think you are right, at least, in my configuration. One removable
bay causes no apparent problems. I hook up two of them and the thing won't
even POST.
I'll shut up now and see what gets posted. I'm really missing the removable
bays, though.
//rus//
