iStar disk rack switching mystery

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View

I'm having problems with my older generation iStar BPU-350 disk rack  
shown here:

This series has activity lights at the top and the pencil switches are  
at the bottom, all of them completely misaligned as usual from the trays  
which makes for a confusing mess.

Before I could switch any of the disks ON but that is NO LONGER the  
case. I have had several mobos since with different BIOSes. The current  
one is Asus Crosshair-IV.

The BIOS lets me set up for drive use on sata leads 1-6 as all IDE, all  
AHCI or 1-4 as AHCI and 5-6 as IDE (for DVD boots).

I have 5 SATA drives in the rack conected to SATA leads 1-5.

I can turn only #1 and #5 ON, no matter what I do #'s 2-4 will not HOLD  
being turned on. The fact that #5 is the other one in addition to #1  
makes me think of the above AHCI/IDE split but the truth is that I  
really have NO CLUE and haven't yet found a way to do it.

Any ideas?


Re: iStar disk rack switching mystery

bad sector wrote:
Quoted text here. Click to load it

Have you ever tried RAID mode ? I think you can use JBOD for
each drive, so the drives don't have to be in an array.

You can certainly dismiss this idea as too much work.
It's a pretty lame suggestion.

Before doing anything like that, make sure you have a backup
of the five drives. Changing to RAID JBOD (i.e. not entering
RAID setup), might or might not add metadata to the drive.
Modern designs put that up near the end of the disk. I think
the one drive I researched with a disk editor, it was in the
last 5MB of storage. If I was experimenting, I'd either
back up the drives, or make sure there was at least
8MB of slack space, with no partition, up near the end.
I was surprised the metadata on my controller, wasn't shoved
right up to the end, and I couldn't figure out how they
determined the exact location.


Re: iStar disk rack switching mystery

On 09/10/2014 02:29 AM, Paul wrote:

Quoted text here. Click to load it

Nothing is lame when you're looking at otherwise unusable hardware :-)

Quoted text here. Click to load it

I tried that and came immediately face-to-face with a raid setup dialog.  
That's not what I want. I DID have this thing going ok initially (by  
that I mean years ago and all the drives were seldom ON at the same  
time) but then somwhere along the way it stopped working right. It was  
just my way of simplifying down from a bunch of larger mobile trays so  
that I could use a mini tower instead of a '24 wheel diesel death-star'.

This thing here seems to have run into some sort of design/support  
conflict or I've got my fingers up the wrong hole, I don't see it being  
a failure given that slot #5 can be switched on. In fact they can all be  
switched on except that 2-4 power off as soon as I let go the pencil switch.

Re: iStar disk rack switching mystery

bad sector wrote:
Quoted text here. Click to load it

Have you had a look at the power converter in the iStar ?

Is there a separate assembly for the  slots 2-4 versus the others ?

You can use a straight, high current switch, to turn electronics
on and off. Or, being a fancy cat, you can use MOSFETs, the switch
would be very low current, and the MOSFET amplifies your request
and turns on the power. Doing so, allows controlled rise and
fall times (capacitor on MOSFET gate to adjust rate). That
helps avoid inductive ringing on power. The hard drive has two
overshoot protection devices, near where power comes in, to stop
inductive surge. (Another poster here, had those two devices burn
on his drive, and that's how we know what they are :-) ) So the
drive itself, Seagate knows there could be a power problem,
and designs the power inlet on the drive itself to handle it. So
the drive can survive hot-swap into a "hot" SATA backplane etc.

Having a switch on each slot, is a "pro" feature. It's better
for the drive, to power off, before using an lever-arm to eject
an enclosed drive. There are other enclosure products around, they
just "pop" out the drive while it is spinning ! So your iStar has
some attention to detail, if it has switches on it.

It could be just the switching circuit is bad.

I can give you an example here. I have a nice multimeter
I've used for years. One day, I go to use it, and the
battery is dead. So I put a new battery in it, and the
multimeter "turns itself on". Immediately, it's evident
the meter turned itself on and drained the battery. So
I know why the battery is dead.

So I get the schematic for my nice multimeter off the web,
and the damn thing has an "electronic switch". Yes, there
is a mechanical latching pushbutton. But the next stage
is a 4000 series CMOS chip, which "times" the meter usage.
If you leave it on for more than ten minutes, the 4000 chip
turns off the power. A semiconductor just after the 4000
turns off the power. The idea is, this electronic switch
is intended to prevent battery draining by inadvertently
leaving the thing on.

So now, the "smart" switch is broken, and I cannot leave
the meter, with a battery in it. So much for "clever"
design. I can't even figure out the capacitor configuration
well enough (there are several caps), to figure out which
cap is leaky and needs to be replaced. And of course, the
circuit is dense enough, I'd probably have a hard time
finding, removing, and replacing it.

If they'd used a simple mechanical switch, and not the
rest of that "junk", the meter would still be useful...

Maybe your iStar has a similar feature. What you're looking
for, is on the power board, some difference between the
way the working channels work, versus the ones that won't
stay powered on. It's hard to believe three channels would
all develop the same defect at the same time. So maybe they
share a chip or something. When you get weird behaviors
like this, it usually hints that something "really clever"
was done in there.


Re: iStar disk rack switching mystery

On 09/10/2014 03:11 PM, Paul wrote:

Quoted text here. Click to load it

Not clever enough to assure failure into the ON mode. I can't pull the  
box out of action, worst case scenario I buy another rack one and it's  
back in service in 5-minutes or less. My worry is that a new one will  
immediately just duplicate the situation for some unsuspected reason :-(

I might add that I'm not set up to tinker with printed boards.

Site Timeline