SCSI Hard drive suddenly shuts down and restarts

  • Thread starter Thread starter Analabha Roy
  • Start date Start date
A

Analabha Roy

Hi,



I have a Seagate ST336607LW SCSI hard drive where I've installed my Linux
(FC3) & windoze root partitions. It is a 68 pin HD with no internal
termination. My scsi controller is an SIIG AP-20 PCI adapter (50 pin with
internal auto-termination). I connected my scsi card to my scsi hard drive
as follows:




50 pin ribbon
50-pin card---------------->50-pin terminator-->50-68 pin adapter-->SCSI
Drive



The hdd is on a dedicated connection to the power supply (no other devices
chained to it along that power cable).



Occasionally , irrespective of whether I'm on windoze or linux, the hdd
makes a sickening "click" sound, shuts down (I can hear it) and spins up
again. My entire compu freezes & I have to hard-reset.

I checked the cable connections & they seem pretty tight, so I'm rather
confused as to what's up. Is the termination bad? IS the HD overheating
(it's typically at around 50 deg celcuis, & the operating limit is 54 deg,
I'm getting an extra case fan that should cool it down further)? Is there
something wrong with the controller? Is the hard drive itself busted (no
badblocks or anything like that, and I ran Seagate's diagnostics on the HD
and there were no errors).

Where should I start in trying to diagnose the problem? DO I require
special hardware?

Bear in mind that there are no syslog error messages in linux as syslog
can't write to the HD when it shuts down, so...


I would appreciate ANY advice offered on this matter.

Regards,
AR
 
Analabha said:
Hi,



I have a Seagate ST336607LW SCSI hard drive where I've installed my Linux
(FC3) & windoze root partitions. It is a 68 pin HD with no internal
termination. My scsi controller is an SIIG AP-20 PCI adapter (50 pin with
internal auto-termination). I connected my scsi card to my scsi hard drive
as follows:




50 pin ribbon
50-pin card---------------->50-pin terminator-->50-68 pin adapter-->SCSI
Drive



The hdd is on a dedicated connection to the power supply (no other devices
chained to it along that power cable).



Occasionally , irrespective of whether I'm on windoze or linux, the hdd
makes a sickening "click" sound, shuts down (I can hear it) and spins up
again. My entire compu freezes & I have to hard-reset.

I checked the cable connections & they seem pretty tight, so I'm rather
confused as to what's up. Is the termination bad? IS the HD overheating
(it's typically at around 50 deg celcuis, & the operating limit is 54 deg,
I'm getting an extra case fan that should cool it down further)? Is there
something wrong with the controller? Is the hard drive itself busted (no
badblocks or anything like that, and I ran Seagate's diagnostics on the HD
and there were no errors).
Click to expand...

<snip>

though the drive could be bad...
i'd try another cable.

if you have another scsi drive you could try that in the machine
to eliminate any possible controller problem
 
I have experienced a similar case as yours. I have two 50GB Seagate hds in
my system controlled by AHA2940UW2. The solution was to use another Power
supply. My original Power supply is a Generic 450W. I did not know if my
case is applicable to yours. However, it is worthy of a try.
F. Hui
 
Free said:
I have experienced a similar case as yours. I have two 50GB Seagate hds in
my system controlled by AHA2940UW2. The solution was to use another Power
supply. My original Power supply is a Generic 450W. I did not know if my
case is applicable to yours. However, it is worthy of a try.
F. Hui
Click to expand...




My power supply is a generic 250 W (specs here:
http://support.gateway.com/s/PC/310_Series/Shared/6500788/6500788sp6.shtml)

I have the following devices connected to the power supply

1. mobo (obviously heh)
2. 2 weak but quiet Case fans
3. cpu with heatsink and a fan on that
4. IDE hard drive (7200 rpm)
5. The SCSI HDD (10,000 rpm)
6. DVD burner/CD burner/DVD ROM combo drive



What is the best wattage for this type of configuration?

Please throw me a bone here.

Regards,

AR
 
Thanks for the info. I bought the following 400 Watt power supply from
newegg:

http://www.newegg.com/Product/Product.asp?Item=N82E16817159043

The antec ones are like 3 times the price. This one was the cheapest. Don't
know how much of a difference that makes. I hope that the transformer
doesn't fall off after 2 days :).
Click to expand...


The difference it makes is that you didn't get a "real" 400W
PSU, you got a 250W PSU with a (arguably fradulent) 400W
label slapped on it.
 
kony said:
The difference it makes is that you didn't get a "real" 400W
PSU, you got a 250W PSU with a (arguably fradulent) 400W
label slapped on it.
Click to expand...


Whoa! Newegg would sell stuff like that???

That's a bit hard to believe from a cool vendor like Newegg.

If your serious, then how do I test the actual power rating (newegg RMA's,
of course)?
 
First thing to do is make absolutely sure that the power connector to
the drive is good. While it may be hard to pull off, that still allows
a conductor to be loose. Use another power connector (not power supply)
if one is available and/or squeeze the individual sockets with a long
nose pliers. Sometimes the springy steel in the connectors isn't
sufficiently springy and as things heat up and cool down contact is
lost. I doubt very much that the problem is in the adapter, the scsi
cable, or the terminator
 
Whoa! Newegg would sell stuff like that???
Click to expand...

Newegg sells whatever people buy, it's a business.

That's a bit hard to believe from a cool vendor like Newegg.

If your serious, then how do I test the actual power rating (newegg RMA's,
of course)?
Click to expand...

Set up capacitive / resistive loads on each rail, hook up a
scope and a few multimeters, vary the load rapidly like
modern CPUs/GPUs/etc do, and run some tests for a few
months, recording voltages, recovery time, ripple, etc.
Then test shutdown thresholds like overcurrent, overvoltage,
overheat, fan lifespan and response... There is no way you
can do a simple "it's good long-term" type of qualification
for it in a day or two.

Put simply, you won't get the same PSU for $20 as for $60
unless you happened to find some rare OEM closeout surplus
something-or-other, which is not a normal stock retail
purchase. It's certainly not necessary to have cable
sheaths, LEDs, glowing neon or any of the eyecandy to cut
costs a few dollars but PSU are commodity items and one
can't just spend 1/2 as much and expect there weren't
significant corners cut on a generic.
 
kony said:
Newegg sells whatever people buy, it's a business.



Set up capacitive / resistive loads on each rail,
Click to expand...


How? do I get a breadboard, put some resistors and capacitors in them, and
use that? How will I be able to connect that with the power port of the
PSU? Do i have to unscrew and remove the chassis or something?



hook up a
scope and a few multimeters, vary the load rapidly like
modern CPUs/GPUs/etc do, and run some tests for a few
months, recording voltages, recovery time, ripple, etc.
Then test shutdown thresholds like overcurrent, overvoltage,
overheat, fan lifespan and response... There is no way you
can do a simple "it's good long-term" type of qualification
for it in a day or two.
Click to expand...

But isn't it possible to just verify the maximum power threshhold of the PSU
without too much testing?
 
kony said:
That they're peak ratings, not sustainable. Essentially
they use questinable methods to determine those numbers and
if other name-brand manufacturers used similar methods,
their "400W" psu would be rated higher than 400W.
Click to expand...


That's a good point I guess...

But what is a psu exactly? Isn't it just a step-down transformer with some
pots in it to regulate voltage? I mean, it doesn't actually supply any
power, right? That comes from the mains. The 'sustainability' would have to
be more dependent on material properties (if the impedance changes too much
as it gets hotter and if subsequent cooling reverses the trend)like
resilience and stuff. If that is true, then the only manufacturing
difference between a high power and low power psu is in the materials that
make the magnetic core, maybe the CPM of the fan, the # of coils in the
transformer and other similar such. Those should be individually measurable
and benchmarkable, and there should be documentation on the web about ideal
parameters and one should be able to compare them.

Let me know what you think. Also, if you do know of some urls where they
have additional info, I'd be interested in reading it.


Thx,
AR
 
Bennett said:
First thing to do is make absolutely sure that the power connector to
the drive is good. While it may be hard to pull off, that still allows
a conductor to be loose. Use another power connector (not power supply)
if one is available and/or squeeze the individual sockets with a long
nose pliers. Sometimes the springy steel in the connectors isn't
sufficiently springy and as things heat up and cool down contact is
lost. I doubt very much that the problem is in the adapter, the scsi
cable, or the terminator
Click to expand...


Thanks for the info. It seems a bit odd that the connector would be loose
only for the scsi drive and not for all the other devices. Are SCSI hdds
specially sensitive to power fluctuations that may result from marginally
loose connectors? I did tighten the connectors with pliers, though I was
afraid to apply too much strength as I could damage the drive physically
(I'm too poor to risk that, the bloody hdd cost $150 even as an OEM). How
shock-and-shear resistant are these things anyway? I looked at the
published values, but since I don't know how much force I am applying
(exactly in kgwt or pounds or whatever), they don't help much.
 
How? do I get a breadboard, put some resistors and capacitors in them, and
use that? How will I be able to connect that with the power port of the
PSU? Do i have to unscrew and remove the chassis or something?
Click to expand...

It'd have to be quite a hefty "breadboard", considering that
the load would require 400W heat dissipation... put in
perspective, you would need several very large heatsinked
resistors, a heat transfer plate, and fairly high velocity
fans, or an oil bath... whatever your chosen method of
dispersing the heat. As for the capacitance you'll need,
rought guess would be a 10,000 mfd per rail (varies per
system, that might be a rather conservative figure for a
system actually using anywhere near 400W which in itself is
hard to do with a PC, but more importantly some systems may
actually use the majority of the 12V rail.

How to connect to the power port? Buy the mating molex
connectors. Setting up a testbed isnt' the kind of thing
one can usually do with bits in their spare parts bin,
unless they have a pretty generous stock of parts. You do
not need to open the power supply at all, unless you also
wanted to measure temps.

Frankly all of this is a waste of time, as opening the PSU
probably voids the warranty then you can't return it.
Testing until it fails is also a bit of a loss, really only
useful to quality fhe design for larger scale deployment
over many systems.

But isn't it possible to just verify the maximum power threshhold of the PSU
without too much testing?
Click to expand...

No, you could get a peak sustainable value that's useful to
the extent that you know if the rails would drop with that
degree of system load, but it won't tell you if there's
potential for damage from massive ripple, won't tell you
what the PSU does in a fault scenario, and doesn't tell you
how long it will run at that output level. Even a poor PSU
may put out 300W for a day, maybe even weeks to months.
That's the worst scenario, because then after the warranty
is up it may fail, or even if warranty remains you generally
aren't covered for any OTHER parts that could be damaged.


Look at it another way- if there was no significant drawback
then how would the other manufacturers stay in business
selling them for far more when they all (until recently)
just looked like boring grey metal boxes?
 
That's a good point I guess...

But what is a psu exactly? Isn't it just a step-down transformer with some
pots in it to regulate voltage?
Click to expand...

Not exactly it swithces HV across a transformer at high
frequency with a duration or rate controlled by (a
controller, to regulate the voltage through at least one or
multiple feedbacks) into rectification diodes and (typically
inductive and capacitive) filters...

That's an oversimplification and ignores complimentary
circuits like the shutdown, fan control, 5VSB, and more,
but thinking "it's simple therefore should be cheap" isn't
true, as any kind of electronic components have tolerances,
values, and more or less elaborate designs including the
filtering and lifespan. There are a few PSU schematics
scattered over the web, Google may find some. Unfortunately
those I've seen are rather simple schematics, more typical
of low end PSU that decent ones.

I'd return the PSU you bought, but you might pop open some
old generic and compare it to a decent PSU. WIthout more
experneice in switching PSU you may not realize the
significance of all of the differences but even an untrained
eye can spot a lot of difference between two "supposedly"
400W PSU, one good and one not.
I mean, it doesn't actually supply any
power, right? That comes from the mains.
Click to expand...

You might be trying to overthink the definition of supply,
and should see it within the context of an electronic
device... within that context, it is suitably called a
"power supply".

The 'sustainability' would have to
be more dependent on material properties (if the impedance changes too much
as it gets hotter and if subsequent cooling reverses the trend)like
resilience and stuff. If that is true, then the only manufacturing
difference between a high power and low power psu is in the materials that
make the magnetic core, maybe the CPM of the fan, the # of coils in the
transformer and other similar such. Those should be individually measurable
and benchmarkable, and there should be documentation on the web about ideal
parameters and one should be able to compare them.
Click to expand...

Yes the parameters can be physically compared, though again
you're oversimplifing. You have a good point about the core
though, the size of the transfomer is indeed "one" of the
common differences between a properly rated and a misrated
PSU, and/or the construction of the core, whether it be a
ferric composite material or old school laminate type. If
you dig deep enough you will find that budget grade (junk)
PSU cut corners in many, many ways. You will notice more
similar build in a generic 400W and a name-brand 250W than a
generic 400W and a name-brand 350W.


Let me know what you think. Also, if you do know of some urls where they
have additional info, I'd be interested in reading it.
Click to expand...


I would first suggest some background reading, either on
switching power supplies in general or if necessary, even
more basic like electronics in general. It's not the kind
of think I can condense into a few usenet posts and I don't
know of any website that does so either, as you first need
the requisite background on power, electronics, SMPS,
computer power demands, then to tie it all together in
context plus survey enough power supplies (including failure
modes of dead ones) to get a better picture.
 
If the floppy or tape drive or whatever cuts in and out, your PC won't
lock up. You should be tightening the female connectors connected to
the PS, not the males on the HD (but perhaps that is what you did and
I misunderstood what you said.) I suggest again that you try a
connector on a different cable in the PS; it may be that there is a bad
solder joint in the PS on the wires you are using.

It wouldn't hurt to reseat the SCSI adapter if you haven't already done
that.
 
kony said:
Not exactly it swithces HV across a transformer at high
frequency with a duration or rate controlled by (a
controller, to regulate the voltage through at least one or
multiple feedbacks) into rectification diodes and (typically
inductive and capacitive) filters...
Click to expand...


Oh, so an AC/DC Converter. Right.
Thanks for the info. I"ll look it up and dust off my old undergrad
electricity/electronics textbooks.
 
Back
Top