12V and HD problem

[Gianni]

Here what I have found on a DEER ATX 300W PS:

Canicon TC(M) nom. value 2200uF 16V measured values 121.7uF 1.634 ohm
Canicon TC(M) nom. value 1000uF 10V measured values 111uF 1.95 ohm
G-Luxon LZ nom. value 2200uF 10V measured values 19.4uF 6.2 ohm

While the G-Luxon was bulged, the Canicon caps have no visible signs of their failure.
All the other caps that I have replaced, most of them were G-Luxon, were still within specs.

Is it possible that, due to the bad cap on 12V rail, tha HD of this PC has been damaged for the high ripple?
On the MOBO there are several caps and inductance to filter the PSU voltages but the HD is connected straight by the Molex connector and I don't think there are large caps inside it; the HD I have replaced had a lot ruined sectors.
What is your opinion?

Ciao
Gianni

[i4004]

Re: 12V and HD problem

bad sectors=damaged platter
and i don't see how can that voltage damage the platter.
thats the voltage that spins the motor.

[arneson]

Re: 12V and HD problem

I could see this possability, becuase your heads are supposed to self park as the disk slows to a stop.
That can't happen if the disk motor is getting brown power.
The heads are riding on a critical air gap determined by platter rpm.
Usually the controller smokes and the drive fails completely before this happens.

[linuxguru]

Re: 12V and HD problem

Hard drives usually have a linear regulator for certain voltages, and none at all for others. There isn't much of a heat sink for the linear regulators - the metal tab of the regulator is soldered to the PCB, that's all. When a bad PSU with high voltage ripple on +12v and +5v is used to power a hard drive, the linear regulators may eventually fail due to higher than designed power dissipation. After that happens, all bets are off - the head-positioning servo may position the head on the wrong track, or even on the servo track, overwriting information that's required for accurate positioning. The drive will go quickly downhill after that.

Even if the linear regulators haven't failed, out-of-spec voltages can cause the same problems outlined above. A hard drive is a precision electro-mechanical device, with very tight tolerances on all electrical/mechanical parameters. Bad rail voltages are a sure-fire way to kill it prematurely.

[i4004]

Re: 12V and HD problem

>After that happens, all bets are off - the head-positioning servo may position the head on the wrong track, or even on the servo track, overwriting information that's required for accurate positioning. The drive will go quickly downhill after that.

key thing there is "quickly".
and i wouldn't call bad sectors quick deterioration.
i can only wish all my drives would get bad sectors first.

head wrongly positioned indeed means quick death, but it won't cause bad sectors(bad sectors=no data, not wrong data) so i don't think that explains bad sectors.

it is surely not good for any component to be supplied with crappy dc power, but i think here we have a case of normal hdd death(worn hdd). data corruption and sudden deaths seem like symptoms bad power should have on hdds.
offcourse, that too is hard to establish, given the crappy quality of some hdd manufacturers. maxtors i had had nice power and were still failing after year and a half...

gradual deterioration? again, i wish i was so lucky...

how old was that hdd anyway? s.m.a.r.t utilities can give you number of working hours, if its working at all....

arneson, you don't really think ripple alone can power that 12v motor?
also, i think it's multi-phase motor (phases produced by driver ic) and i don't think ripple will have much effect there...

[i4004]

Re: 12V and HD problem

to clear up what i said to arneson
i don't think dc with ripple will prevent motor from stopping in any way.

[arneson]

Re: 12V and HD problem

Sorry about the brown or shit power remark.
What I meant was, that condition we've all seen where the whole building dims repeatedly as the feed transformer to the neighborhood lets loose.
Happens to me on a regular basis. Power company just band aid fixes it till next time.
The only way I'm avoiding failures is with UPS on all systems, and even that's risky.

[Gianni]

Re: 12V and HD problem

Hi i4004, linuxguru and arneson, thanks for your reply.

This HD failed suddenly, my friend told me he had a warning about a virus while he was browsing, the PC frozen few seconds after AV warning so he turned off the PC.
When he turned it on he got a message error about system. He thinks the virus has destroyed his HD but AFAIK there are not virus so dangerous.
Before this problem the PC worked fine, no problem at all if I ignore all the time my friend messed up WinXP.
The HD was a Maxtor 60GB, it was built in 2003. I had thrown it in the trash already, so I can't do more test.
I tested it with Maxtor PowerMax 4.23 and the result was an error code that I discovered is top secret, at least what I have found is that the meaning of such error code is known by Maxtor only, you can't find what it really means. Anyway the message also said to contact the vendor for immediate replacement ...

I tried to recover it with a program called HDD Regenerator, my brother successfully repaired 2 HD that had some sectors ruined, but no more than 4/5.
In my case, after 9 hours, the programs had recovered about 11450 sector out of 11500 and the progress bar was only more or less at 4% of the total disk space. I stopped it because I thought all disk was damaged.
Till now, the only HD I have seen to dead so quickly and badly were all Quantum Fireball; I have worked few time on Samsung and Maxtor with problems, but I was always been able to partially recover the data.

This is why I asked about it, I know the +12V feed the HD motor and I was wondering if an high ripple explained such kind of failure.
The HD didn't smell of burning and there were not visible damage or burnt component on the PC, it was only impossible to format/partition it.
By the way I was very surprised when I bought a new Maxtor 160GB: it was thin, the pcb was smaller and no component visible on it compared to the old 60GB.
Ok don't laugh , probably you see more hardware than me and this thing don't surprised you; I usually repair old stuff so I'm not update on new products.
By the way, I have another PC to raise from the dead... probably I will be back with new questions .


Bye
Gianni

[i4004]

Re: 12V and HD problem

given it was sudden then linuxguru's explanation seems plausible. but also...maxtor that did 5 years...i think that's excellent for maxtor..hehe...

here's stuff about maxtor firmware errors
http://www.baixaja.com.br/downloads/...ppy_56466.html

new maxtor? oh...i bet that one won't even do 5 years....

[PCBONEZ]

Re: 12V and HD problem

Part of the air gap is based on the disc spinning.
If disc speed is going fast-slow-fast-slow due to bad +12v then the heads may be bouncing off the disc. -> bad sectors.

[i4004]

Re: 12V and HD problem

you speak as if platter has no inertia of its own...

and why just "part of the air gap"?
if platter is not spinning heads are landed, ergo gap is solely based on platter spinning.

[arneson]

Re: 12V and HD problem

Yes the arms are wings in effect, the heads are flying and the air comes in thru the tiny filter that says 'do not block'.
Long time ago there was a discussion about putting head assemblies and platters in drives.
It's challenging, and then you sneeze.

[PCBONEZ]

Re: 12V and HD problem

you speak as if platter has no inertia of its own...

and why just "part of the air gap"?
if platter is not spinning heads are landed, ergo gap is solely based on platter spinning.

Platters don't have a clutch and don't 'coast' free of the motor.
If the motor slows they do too.

[PCBONEZ]

Re: 12V and HD problem

Heads and arms also have inertia~~~~

[i4004]

Re: 12V and HD problem

>If the motor slows they do too.

but not instantly. take it apart and see for yourself. see what's platter made of and it's weight...


do you have any idea how much platter needs to slow down in order for head to touch it?
and you know for sure voltage ripple can cause this?
you're essentially saying few percentages of motor speed change(btw. why do you think this system is not servo? do you really think motor is plugged directly to 12V?)) can bring head down.
proof?
proof would be monitored 12v that's rippled or just too high or low that causes bad sectors.
i say you can't produce that.

>Heads and arms also have inertia~~~~

and?
huh, what's more i would say that they have a soft spring that's pushing the heads to the disk, so that disk works even when placed upside down and sideways...but that just reaffirms notion that platter is only thing that's kepping the heads in the 'air'.

i would say only way for bad sectors to appear would be debris coming off from heads(while they were being parked) and that debris coming in between head and platter. or perhaps when head becomes loose.

i would put platter speed as least possible bad sector cause.
possible, but highly unlikely.
driver ic will probably shut down when it detects too low/too high or too dirty voltage...

heck i would state worn bearing as more probably cause for speed issues, but again servo should deal with that untill it becomes real nasty, and by that time it should be audible...

i think problems of recent hdds are not really bad sectors, but general poor build quality.
i think you have a lot of dirty power going to hdds but reasons for their premature filures lie elsewhere. they just fail sooner than power becomes that dirty...

[arneson]

Re: 12V and HD problem

Dirty huh?
That's such an odd expression.
I picture mud and debri coming out of the wall outlet.
Then I open an old power supply ans sure as shit, filthy dirty power inside.
Good point about bearing fail.
The biggest stack of crap drives I have are those head bangers that just make that horrible clunking, themselves to death.
These are all 2.5's, mostly travelstar that have had a hard life.

[linuxguru]

Re: 12V and HD problem

The Travelstars made in Hungary are notorious for their (lack of) reliability. I have several of them, mainly to get donor controller boards to swap with other drives with intact platters. BTW, these board swaps don't work reliably these days, after the drive manufacturers started writing a portion of the HD firmware on the platter itself, to save on costs - damned bean-counters.

OTOH, Fujitsu 2.5" drives seem to be very reliable - even ugly, scratched, dusty drives below about 12GB or so will almost always spin-up, get detected and function, even if they have some bad sectors. Toshiba was reliable, now much less so. That leaves Seagate and Western Digital, which seem to be OK at all sizes from 40 GB to 120 GB. Above that, I have no idea about 2.5" drive reliability.

[i4004]

Re: 12V and HD problem

that travelstar comment should speed up intended backup hdd purchase...or to clean some space on current backup hdd..huh...doing backups is pita..
even when you lost few drives, it's still pita.
and you know that if you stop doing it drive will probably fail tomorrow, because of murphy's law, but still...pita.


dunno if this one is from hungary, but i'll check it out as soon as i finish stuff i'm dloading...

yes, they're writing portions of firmaware on platters....if they weren't disks would probably be much, MUCH more reliable devices...
because SOME damage is surely better than drive that's TOTALLY unusable...
with rom on platters they get totally useless all too often.

and that's really old thing by now: maxtor was doing in in early 00's.

[PCBONEZ]

Re: 12V and HD problem

i4004
By your theory heads would never come into contact with the disk as long as the disk is spinning.
- If you believe that then I think they still have swamp land available in Florida.

If you read up I specified a condition where the drives speed is 'bouncing'.
That will cause the heads to bounce/vibrate in the vertical plane.
Air gap is down to like 0.000003 mm.
You couldn't even SEE that small a vibration.

.

[i4004]

Re: 12V and HD problem

>I specified a condition where the drives speed is 'bouncing'

you said something like
"If disc speed is going fast-slow-fast-slow due to bad +12v then the heads may be bouncing off the disc"
and i said they use servo motors, so what does that do for your argument?
these motors don't work directly with 12v dc. they're not dc motors. they don't have brushes, they have few phases(probably 3) made from the +12v.

now explain to me how will this disc go fast-slow-fast-slow, if we presume driver ic is ok.
if we presume driver ic is faulty then more probable scenario is disk that's instantly dead because it won't reach speed and heads will destroy platter in first few seconds(if driver ic protection doesn't kick in), OR it won't start spinning at all(if its protection kicks in).
if somehow it starts spinning faster than normal, heads will go too far up and there won't be any reading or writing.
all of these are sudden death conditions and not things producing some bad sectors.

so i do see scenario where heads touch platter, but i don't see one where they bounce on/off because speed of motor is varied.
i think that even if you had driver ic that's producing just 2 phases, and not 3, you won't have varied speed, but slower speed.

i just don't see this system as something with varying speed at any considerable amount.

what sort of malfunction would driver ic need to have in order for that to happen, and how probable do you think that is?