Removable Drive Bays

[John H.]

Anyone have any recommendations for removable IDE drive bays on a home
system?

I see prices ranging from $12 plastic bays,
http://www.mycableshop.com/sku/BAYATA100.htm

$29 for the aluminum/plastic ICY Dock MB123
http://www.megahaus.com/Itemdesc.asp?CartId={354260D0-A013-4D9E-A077-0E029260238D}&ic=MB123%2DAK&Tp=

and a dear $79 for the DataPort V
http://www.megahaus.com/Customkititems.asp?CartId={354260D0-A013-4D9E-A077-0E029260238D}&kc=DP2200A66%2DKIT

That's quite a price range. Are the cheap ones junk?

One thing I can't see from the pictures, is there a gaping hole when the
HD is out or does a cover pop up?

Does the fan run with no drive inserted ?


Thanks --JH

 

[Rod Speed]

Anyone have any recommendations for
removable IDE drive bays on a home system?

I recommend they all be avoided and
firewire external drives be used instead.

With firewire you do have a proper formal
standard and not a monsterous kludge.

I see prices ranging from $12 plastic bays,
http://www.mycableshop.com/sku/BAYATA100.htm

$29 for the aluminum/plastic ICY Dock MB123
http://www.megahaus.com/Itemdesc.asp?CartId={354260D0-A013-4D9E-A077-0E029260238D}&ic=MB123%2DAK&Tp=

and a dear $79 for the DataPort V
http://www.megahaus.com/Customkititems.asp?CartId={354260D0-A013-4D9E-A077-0E029260238D}&kc=DP2200A66%2DKIT

That's quite a price range. Are the cheap ones junk?
Basically.

One thing I can't see from the pictures, is there a
gaping hole when the HD is out or does a cover pop up?

Pops down usually.

Does the fan run with no drive inserted ?

Normally not.

 

[Andy Axnot]

Anyone have any recommendations for removable IDE drive bays on a home
system?

Thanks --JH

I see Rod Speed has technical objections to these, and I have heard similar
objections from other technically savvy people.

Well, sometimes ignorance is bliss. Not knowing any better, I use the
cheap (about $8) mobile racks from Genica, available at, for instance:
http://www.compgeeks.com/details.asp?invtid=GN210

I have used these for several years, mostly with Win98SE, without any great
problems that I'm aware of. There were a couple of instances of particular
computers not finding particular drives on boot up. That is, a particular
(usually older) hard drive might not be detected in one computer while
another computer found it. I don't know why this was and didn't have the
time to hunt down an explanation. Perhaps one will be offered in this
thread. Mostly I have had no problems moving data from one computer to
another, and friends and I will occasionally swap data this way.

I use them basically for backups and Ghost images. I then remove the trays
and store them in a closet. I do not attempt to hot swap them (i.e., insert
or remove with power applied). I insert or remove trays with the computer
powered down. When I insert a tray and power up the computer, the OS finds
the drive and its partitions are available.

These racks I use are quite cheap and I don't think they would stand up to
abuse, and they don't help drive heat dissipation any. Maybe the aluminum
racks would be better in this regard. Still, these do the job for me. I
have transferred data among computers using them, restored back to a known
good state from a Ghost image, etc., so I'm satisfied.

Andy

 

[John]

I see Rod Speed has technical objections to these, and I have heard similar
objections from other technically savvy people.

Well, sometimes ignorance is bliss. Not knowing any better, I use the
cheap (about $8) mobile racks from Genica, available at, for instance:
http://www.compgeeks.com/details.asp?invtid=GN210

I have used these for several years, mostly with Win98SE, without any great
problems that I'm aware of. There were a couple of instances of particular
computers not finding particular drives on boot up. That is, a particular
(usually older) hard drive might not be detected in one computer while
another computer found it. I don't know why this was and didn't have the
time to hunt down an explanation. Perhaps one will be offered in this
thread. Mostly I have had no problems moving data from one computer to
another, and friends and I will occasionally swap data this way.

I use them basically for backups and Ghost images. I then remove the trays
and store them in a closet. I do not attempt to hot swap them (i.e., insert
or remove with power applied). I insert or remove trays with the computer
powered down. When I insert a tray and power up the computer, the OS finds
the drive and its partitions are available.

These racks I use are quite cheap and I don't think they would stand up to
abuse, and they don't help drive heat dissipation any. Maybe the aluminum
racks would be better in this regard. Still, these do the job for me. I
have transferred data among computers using them, restored back to a known
good state from a Ghost image, etc., so I'm satisfied.

Andy

I'm with Andy. They are cheap, and the transfer rate is way higher than a
USB2
or firewire connected drive.

I've hotswapped mine under XP with no problems (by disabling within Device
Manager).

Mine are the ViPower ATA/133 IDE cadies (plastic). They have 3 fans, so heat
doesn't seem to be an issue.

 

[John H.]

I see Rod Speed has technical objections to these, and I have heard similar
objections from other technically savvy people.

Well, sometimes ignorance is bliss. Not knowing any better, I use the
cheap (about $8) mobile racks from Genica, available at, for instance:
http://www.compgeeks.com/details.asp?invtid=GN210

Only $6.90. That IS cheap - and hot swappable too. Too late for me
though, I picked up a Kingwin KF-101 at a store a mile down the street
for $20. At least it does have a lot of aluminum in it (for better
cooling - or so they claim) and I didn't get ripped off on shipping cost
or have to wait for it. But there's no hot swapping for the Kingwin.

http://www.kingwin.com/pdut_detail.asp?LineID=&CateID=35&ID=95

I haven't had a chance to try it yet because I'm still working on trying
to save my W2k install. I had never fully installed my newer drive and
had it on the secondary IDE channel, with W98 on first partition of
drive 0 (C) and W2k on the first partition of drive 1 (D). I moved W2K
to the 2nd partition on the newer drive (D), W98 to the 1st partition
(C), removed the old drive and put the newer one on the primary IDE
channel.

What a mess. I had resized the partitions and forgot to create a new
W98 boot sector. I got that fixed so I can use W98 but W2k is still
screwed up, in a loop loading and saving user settings.

I remember having problems with that when I moved W2k from D on the 2nd
partition of the old drive to the 1st partition of the newer drive. But
then I still had the old copy of W2k as well which must have allowed me
to boot up. I could then fix the registry. I don't know how I can do
that this time. It's beginning to look like I'll have to reinstall W2k
(which isn't all bad, I'm overdue for that anyway).

THEN I can try out my new mobile rack.

 

[Rod Speed]

I'm with Andy. They are cheap, and the transfer rate
is way higher than a USB2 or firewire connected drive.

Thats overstating it with firewire. Higher, yes, but not way higher.

I'd rather have a proper standard used than a massive kludge myself.

I've hotswapped mine under XP with no
problems (by disabling within Device Manager).

And thats another area where the swap is done more elegantly with say firewire.

 

[Timothy Daniels]

Well, sometimes ignorance is bliss. Not knowing any better,
I use the cheap (about $8) mobile racks from Genica,
available at, for instance:
http://www.compgeeks.com/details.asp?invtid=GN210

I have used these for several years, mostly with Win98SE,
without any great problems that I'm aware of....

Andy, what was the data rate of the fastest drive you used?
It's possible that the "kludge" and the spec "flout" that some
refer to only has a deleterious effect with the faster drive interfaces.
Also, do you use the standard 80-wire ribbon cable, the standard
40-wire ribbon cable, or some length of "round" cable?

The student computer labs at the local campus of the University
of California use Kingwin removeable drive caddies, and the lab
technicians say that there have been no problems.

*TimDaniels*

 

[Andy Axnot]

:.

Andy, what was the data rate of the fastest drive you used?
It's possible that the "kludge" and the spec "flout" that some
refer to only has a deleterious effect with the faster drive interfaces.
Also, do you use the standard 80-wire ribbon cable, the standard
40-wire ribbon cable, or some length of "round" cable?

The student computer labs at the local campus of the University
of California use Kingwin removeable drive caddies, and the lab
technicians say that there have been no problems.

*TimDaniels*

Hmmm... that's interesting. There seems to be a lot of us using these
caddies.

Well, I have used up to ATA133 drives. The IDE cables I use are 80 wire
ribbon cables, as are the short internal cables inside the trays. I have
never tested the data transfer in any way other than normal use.

There is really nothing esoteric about these racks or caddies. They
essentially serve as a short IDE "extension cord". A hard drive in one of
these racks is connected to the regular IDE cable through an additional 2 or
3 inches of 80 wire IDE cable and the Centronics connector that connects the
removable and fixed parts of the rack.

In practice, this seems to present no problems. But I can certainly
understand that it might be stretching the standards just a bit and that it
*could* risk data integrity by lengthening the cable run a bit and adding an
2 additional connections. It might be that there is some slight increase in
the occurrence of data corruption. I don't know enough about how this is
handled by either the hardware or the software to know if this could be
significant.

Interesting thread, this.

Andy

 

[John H.]

Hmmm... that's interesting. There seems to be a lot of us using these
caddies.

Well, I have used up to ATA133 drives. The IDE cables I use are 80 wire
ribbon cables, as are the short internal cables inside the trays. I have
never tested the data transfer in any way other than normal use.

Well I got mine installed and did give it a fairly good test by running
the IBM DFT and DiskSpeed32 and by moving partitions on it with
PartitionMagic. It not only works at ATA100 speeds but the drive stays
cooler too. I thought the thing was broken at first until I realized I
had to turn the key on. I even got W2K working again without
reinstalling. It had assigned DGHC to the partitions on drive 0 when
what I wanted was CDEF.with W2k on D. I hate drive letters.

 

[Rod Speed]

But you did say that they arent exactly high performance drives.

Hmmm... that's interesting. There seems
to be a lot of us using these caddies.

Well, I have used up to ATA133 drives. The IDE cables
I use are 80 wire ribbon cables, as are the short internal
cables inside the trays. I have never tested the data
transfer in any way other than normal use.

And you may get quite a surprise if you
checked the error rate seen on those cables.

There is really nothing esoteric about these racks or caddies.

Yes, but they do flout the specs of the ATA ribbon cable.

They essentially serve as a short IDE "extension cord".

Which isnt even allowed for in the ATA standard.

A hard drive in one of these racks is connected to the
regular IDE cable through an additional 2 or 3 inches
of 80 wire IDE cable and the Centronics connector that
connects the removable and fixed parts of the rack.

Which also flouts the ATA standard.

In practice, this seems to present no problems.

You admit you havent tested whether there are problems.

Plenty have found problems with removable drive bays.

But I can certainly understand that it might
be stretching the standards just a bit

A hell of a lot, actually, particularly that extra connector
which was never designed for that sort of use.

and that it *could* risk data integrity
by lengthening the cable run a bit

Thats not the problem.

and adding an 2 additional connections.

Thats the problem.

It might be that there is some slight increase
in the occurrence of data corruption.

It can be rather more than slight.

I don't know enough about how this is
handled by either the hardware or the
software to know if this could be significant.

It can see the OS decide that the drive is flakey.

 

[Timothy Daniels]

But you did say that they arent exactly high performance drives.

Correct. All I know is their rotational speed - 5400 rpm.


*TimDaniels*

 

[John]

:.

Andy, what was the data rate of the fastest drive you used?
It's possible that the "kludge" and the spec "flout" that some
refer to only has a deleterious effect with the faster drive interfaces.
Also, do you use the standard 80-wire ribbon cable, the standard
40-wire ribbon cable, or some length of "round" cable?

The student computer labs at the local campus of the University
of California use Kingwin removeable drive caddies, and the lab
technicians say that there have been no problems.

*TimDaniels*

I transfer data at 2GB/minute daily (according to Norton Ghost).

The caddies seem to handle that just fine.

 

[Timothy Daniels]

"Timothy Daniels" replied tentatively:

Correct. All I know is their rotational speed - 5400 rpm.

I just called the lab manager, and the channel speeds are
ATA/100, and they've been upgraded to 7200 rpm hard drives
in the Kingwin removeable caddies. The only problems reported
were with the 12" 80-wire ribbon HD cables that came with the
Dell computers - they were too short to reach the expansion
drive bays, and the lab had to substitute standard 18" cables.

Please don't interpret this to mean that all removeable caddy
installations will work as well. They're not part of the ATA specs,
and so their use is, essentially, one's own engineering. But I plan
to use them for backing up my HD, and I think they're a great
convenience.... along with round cables.

*TimDaniels*

 

[Folkert Rienstra]

Hmmm... that's interesting. There seems to be a lot of us using these caddies.

Well, I have used up to ATA133 drives. The IDE cables I use are 80 wire
ribbon cables, as are the short internal cables inside the trays. I have
never tested the data transfer in any way other than normal use.

There is really nothing esoteric about these racks or caddies.

They essentially serve as a short IDE "extension cord".

Only if used at the end of the cable.
When used at the middle connector they add a substantial stub that can cause
signal reflections on the bus.

A hard drive in one of
these racks is connected to the regular IDE cable through an additional 2 or
3 inches of 80 wire IDE cable and the Centronics connector that connects the
removable and fixed parts of the rack.

In practice, this seems to present no problems. But I can certainly
understand that it might be stretching the standards just a bit and that it
*could* risk data integrity by lengthening the cable run a bit and adding an
2 additional connections.

It's the stub that is presenting the bigger problem.

It might be that there is some slight increase in
the occurrence of data corruption. I don't know enough about how this is
handled by either the hardware or the software to know if this could be
significant.

It sometimes leads to the bad sector syndrome where the driver resets the
drive and retries the sector, causing system freezes.

 

[John H.]

Only if used at the end of the cable.
When used at the middle connector they add a substantial stub that can cause
signal reflections on the bus.

That's a good point Folkert. I installed mine on the middle connector
with the drive set as a slave, and have a CD drive on the end connector
set up as master. This seemed to make sense. Placing the CD in the
middle would give me an even longer stub (6" + 2") hanging off the CD
connector when the removable drive was out. I guess the only solution
would be to have nothing but the removable drive on the IDE channel.
Problem is, I would need 3 IDE channels and only have 2. But so far
everything works as is.

 

[Luxor]

I'd rather have a proper standard used than a massive kludge myself.

Massive kludge? Mobile racks have been used in servers for years. What
an idiot. I'm not using any right now but I have used, and own, ATA
133 Lian-Li plastic racks with no issues at all.

 

[Rod Speed]

Massive kludge?

Yep, massive kludge.

Mobile racks have been used in servers for years.

Pity they have flouted the ATA standard for years too.

What an idiot.

What a pig ignorant fool.

I'm not using any right now but I have used, and own,
ATA 133 Lian-Li plastic racks with no issues at all.

The technical term for the is pathetically inadequate sample.

Plenty have had problem siwth removable drive
bays, because they flout the ATA standard.

 

[Folkert Rienstra]

That's a good point Folkert. I installed mine on the middle connector
with the drive set as a slave, and have a CD drive on the end connector
set up as master. This seemed to make sense.

Actually, that adds the 2" stub.

Placing the CD in the middle would give me an even longer stub (6"
+ 2") hanging off the CD connector when the removable drive was out.

That wasn't what I meant. An interface bus is like a tree without
branches: no stubs allowed. However, from a termination stand-
point the unterminated end with the drive removed would indeed
form a stub as termination would now be at the middle connector.

 

[John H.]

Yep, massive kludge.

Would you consider a SATA mobile rack a kludge? If so, why?

 

[Rod Speed]

Would you consider a SATA mobile rack a kludge?

Nope, that approach is allowed for in the SATA standard.

The only real downside is that the mechanical detail
isnt completely standardised, but that has no effect
on its performance, just on physical detail of rails etc.