How do I pinpoint SATA/PSU problem?
I have a Maxtor Diamondmax Plus 9 SATA 120 Gig installed on my ASUS A7N8X Deluxe SATA Raid controller – see sig for full spec.
I have 2 separate problems but I’m concerned that one problem may be causing the other – I would just like to run it by some techy boffins to see if I’m on the right line of thinking…
The first problem, below, I was experiencing even before the SATA drive was installed:
Problem One
Occasionally, when I switch on my computer I hear the fans start to spin & then the computer just powers down but power light LED remains lit. I have to switch off the mains power at the back of the computer for a few seconds to get it working again – and it then usually powers up no problem. Is this a sign of not enough initial juice on the power up? (The PSU is a Enermax 300W, 3 years old.)
Problem Two
Also however, just to confuse matters even more, ever since I installed the SATA drive occasionally my system won’t boot due to DISK BOOT FAILURE. I have found that this can be fixed by unplugging then reattaching the SATA cable to both the drive & the MOBO (I’ve tried 2 different SATA cables.).
It will then boot up no problem & yet the HDD was working perfectly well before it was powered down which makes me think the connections are fine - & yet the unplugging & reattaching method works.
So the questions to my confusing post are:
Plec
I have 2 separate problems but I’m concerned that one problem may be causing the other – I would just like to run it by some techy boffins to see if I’m on the right line of thinking…
The first problem, below, I was experiencing even before the SATA drive was installed:
Problem One
Occasionally, when I switch on my computer I hear the fans start to spin & then the computer just powers down but power light LED remains lit. I have to switch off the mains power at the back of the computer for a few seconds to get it working again – and it then usually powers up no problem. Is this a sign of not enough initial juice on the power up? (The PSU is a Enermax 300W, 3 years old.)
Problem Two
Also however, just to confuse matters even more, ever since I installed the SATA drive occasionally my system won’t boot due to DISK BOOT FAILURE. I have found that this can be fixed by unplugging then reattaching the SATA cable to both the drive & the MOBO (I’ve tried 2 different SATA cables.).
It will then boot up no problem & yet the HDD was working perfectly well before it was powered down which makes me think the connections are fine - & yet the unplugging & reattaching method works.
So the questions to my confusing post are:
- Are the 2 problems connected – is this consistent with just a PSU problem - even though voltages check out fine?
Or, if they are separate problems: - How could I go about determining the SATA problem – as I haven’t got a spare SATA drive, controller or MOBO – without RMA’ing both (Tried 2 separate cables)?
- Does the occasional powering down problem on the initial start up sequence stink of a PSU problem or is it consistent with something else?
Plec
0
Comments
Plug a smaller HD into your SATA with a Serillel adapter and see if you can install same O\S and stably run with fewer issues.
If so, one of two or more general power related things may be happening:
1. If your outlet is not power conditioned before power goes into PSU,then the PSU gets to waste work to regulate and is dumping excess using switching capacity (to ground) or is havign to charge caps in order to try to stabilize a low voltage situation.
2. At a guess, given errors, the second general idea as below is more likely though. Basicly, at the time a HD spins up it needs a huge pump of 12 voplt power relative to rest of system (excluding fans). to see if you are straining the 12 volt rail beyond capacity when it is at real high demand, you can unplug or plug into a separate PSU the following things:
DVD
cooling fans
You can use an AT PSU for this trial as it is not a fix, it is an experiment that should not be left in place any longer than required due to the fact that multiple chassis ground routes result in feedback and you create a second the instant you plug anything cabled to mobo or chassis into one. They will have different potentials and you will get some flow from weaker to better or have some things not use the best ground for voltage control feedback. You will not kill anything except data, but one PSU is much more likely to keep a computer stable and keep voltages stable with only one PSU per physical box on at one time.
But, if the PSU add works, you know you have a weak, damaged, or PSU of wrong subtype inso far as how railtorail balancinmg is done to have your system work right and if any of those things are so the best fix is a new PSU.
BOTTOM LINE:
Given a 3 year old PSU and new other stuff, I would replace PSU as soon as you can prove to yourself and whoever might lend money that it really is needed now, but I would also say that changing PSU is very likely to fix problem. Experiments as above will confirm or disprove this working hypothesis.
John Danielson.
Re-reading my post & reading your detailed response of possible solutions I think I’ll try replacing the PSU first. It’s about time I upgraded my PSU plus I was thinking of getting a silent version later this summer anyway – I’ll just make it sooner rather than later.
I think I may plumb for the Thermaltake Silent 480W ATX PSU with Dual Fan – as this seems to pack a lot of punch for the money – plus it’s a name I trust (or at least, I have in the past.)
Thanks again for taking the time to write such a detailed response – I’ll probably buy the PSU next week & I’ll let you know how I get on.
Plec
John Danielson.
Could/are the low averages of the 5v & 3.3v rails consistent with the above problems i've been having? Do these results more or less confirm your suspicions of it being a PSU problem?
Plec
You need a digital or analog Volt-Ohmmeter for this, and I prefer a digital one that I have checked the voltage readings on. If you want 100% assurance that only one thing is wrong then you are going to get to swap things outr adn see what happens or does not or measure with good equipment what changes.
As things stand now the PSU is normal. More than + or - 5% is your indicator that a PSU is so bad it needs replacing now.
Wild hairy guess-- does the mobo have a socket that takes a 2X2 power connector???? Is nothing plugged into it??? If so, does the PSU have nothing in its power harness that connects to it???? If all are yes, replace PSU, it is the wrong subtype. If any are untrue of the above three questions I need the answer to very specific question in this post to even have a high probability of guessing right.
John Danielson.
Plec; I'd guess it's the PS. Even if it isn't, you should really replace it. That's a lot of hardware to ask a 300w anything to power up. Skip the Thermaltake btw. The only decent product they have is the SmartFan2 case fan. Get an Antec PS... if they're too expensive, consider Enlight and Allied.
Tex
Anything with a motor in it draws more power to bring it out of rest state than if spun up to speed. At boot almost every motor is working its hardest to spin up and if not spun fully down inertia allows it to change speed with less draw than starting from rest.
So, questions:
How many rotational drive motors spin up at once right at boot???? Versus how many are spinning up to speed AT ONCE while normal RUN is in progress???
My answers, based on hearing and MEASURING quite a few systrems over the years, is that the 12 volt rail's draw is about 15% HIGHER during boot than during normal run.
More questions:
What happens when a PSU is near total (all-rail ) capacity and one rail exceeds capacity??? Do the other rails slump while the one overpeaks??? Can this be a voltage slump out of spec???
To the second and third of this set, yes to both based on physical measurements.
John Danielson.
*decides to go hang elsewhere than where admin\owners contradict common sense*
The problem with Maxtor is very common among NF2 boards with Silicon Image 3112A controller.
It is more noticeable if using SATA-IDE adapters but it happens with native SATA drives as well.
The speed of the drive seems to be the culprit on both cases because when the drive is downgraded from UDMA6 (ATA133) to UDMA5 (ATA100) the problem goes away.
Give it a try if you want by using the Windows RAID utility from Si:
Good luck!
/edit: never mind, I just noticed that you have only one drive. This utility works on RAID arrays only. Sorry, there is no other utility to perform the downgrade on your drive, not even from Maxtor.
I think 80W on its own is substantially more than a few fans would draw powering up.
Equito, thanks for your input, it actually may prove relevant/useful in the future as I may buy a 2nd SATA drive once I get things stable again – so thanks, you’ll have saved me a search & a probable post.
Geeky1, after reading Ageek’s detailed responses I’ve come to the same conclusion. I’ve just looked at the Antec cases & they come with a pretty impressive price tag. I think i'll search out some of your other suggestions - but then stability is what i want so perhaps i should just get the Antec - healthy price tag an' all
Ageek, I really appreciate your efforts in guiding/advising me as to possible solutions with regards to this problem. I’ve read a lot of your other posts on different threads & have gleaned lots of useful information. It would be a shame if you disappeared as your detailed replies on varying topics probably help out a lot more ppl than just the person you're replying to.
It’s the same with a lot of the experienced guys who post here: Equito, Tex, Leonardo, Shivian, Cool Canuck etc – lots of their posts on separate issues have helped me understand my rig & possible solutions to future problems a whole lot better. However, I’ve noticed that there’s sometimes a differing of opinion with regard to possible solutions for particular problems. This is no biggy - & to be expected, to a large degree, as 'original thread starters' invariably can be very short on information making the original question ambiguous from the start leading to lots of varied conclusions.
Basically, I hope you don’t decide to 'hang elsewhere' as I find that threads with differing opinions always produce a lot more possible solutions & loads of detailed explanation as to why a problem may exist.
Personally, I think a degree of contention is a healthy thing on a forum as it usually produces results & an abundance of techy info for us 'lesser mortals' to mull over
Thanks to all that replied, appreciated.
Plec
I can't find any Allied or Enlight PSU's at any of the online sites i normally buy (trust) from but I can get the following PSU's for the amount of money I'm prepared to pay:
I know you recommended the Antec but the it's only 380W - would i be better of buying the Enermax or the Thermaltake? (Like the Antec, both can be auto thermally controlled or manually - & both are AMD approved.)
I know you suggested skipping the thermaltake but it seems to pack a lot of punch for the money - it even supplies a 5.5" front bay fan controller. Have they got a bad reputation?
Plec
BTW, this is completely off topic, but I just have to know:
Is your screen name referring to the fish?
Cool, thanks - but don't make a worry of it. I've looked at all the home sites but TBH it all reads the same to me
Nope
./me looks up this Plec fish...
Plec(trum)
Here's the issue I see:
Enermax comes right out and admits that their power rating is done @ 25*C, which is all well and good, except for one thing- the real operating temperature under normal conditions is going to be ~40*C. PC Power & Cooling compared their 510w to the Enermax 550w. Turns out that @ 40*C, the Enermax puts out 366w, which is ~66% of nominal. Assume the same thing for that 460w, and it's "real" power is just over 300w.
I don't know for sure how Antec and Thermaltake test, but I'm inclined to think they do the same thing. The Antec still has an advantage though- it doesn't share the 3.3v and 5v on one rail; it can actually put out 35a @ 5v & 28a @ 3.3v at the same time, while the Enermax is limited to 200w, which is 45w less than what my 400w antec can do. The Tt PS looks great on paper, but as was proven both by PC Power & Cooling in the ad I mentioned (the pdf of that is http://www.pcpowercooling.com/pdf/Turbo-Cool_510_vs.pdf btw) and by the FX5800 Ultra, specs mean nothing. Real world performance is everything. And I don't know who the OEM for the Tt PSes is, so I can't recommend them. If they're made by a decent company, they should be fine, but I don't know that they are, and some of their other products are real crap. So I'd still recommend the Antec...
About the fish...
Hypostomus plecostomus;
Common names: "Algae eater" "Sucker fish" "Plec" "Pleco" "Plecostomus"
Size: Potentially up to 2' for the real deal; the various aquarium-bred varieties usually stay smaller
Tank: Big. Big is good. Or be prepared to replace the fish with a small one every few months.
Water: Is it liquid? Then it'll probably live. Well, almost anyhow. 75-90*F, 78-83*F preferred, pH pretty much inconsequential, as is hardness.
Nocturnal by nature, they like to dig up plants, eat plants, eat zucchini, eat anything green, eat anything that looks like it might have at one time been green, etc. When they get older, they get lazy, and spend most of the time between meals hiding (presumably sleeping). They've got stiff, spiny scales, so they aren't very flexible, and they can't be netted without risking damage to the fish. You get to catch them by hand instead.
But I run a Barton Box on a 400 Codegen, next to a P4 on an Antec. Both stable as heck. Room air temp is about 84 F nominal. Case about 34 C-- Room plus 5 C. P4 runs about 47 C (Floats in 46-48 range).
Never had so much as a stumble out of either PSU.
AFAIK, ThermoTake used Leadman innards for a while-- no idea about right now, though, as don't have one to dissect of the size you are talking about-- nor even any that are recent.
John Danielson.
IF PSU card is mounted on bottom, and input fan is on bottom or other end, yes. Look careful at some True-Power 2 fan PSUs. The ones I have looked at, Card is mounted to top of PSU(INVERTED), intake fan on bottom blowing up on card, output fan out back. The Antec's two lower power and volume fans keep same coolness or better with less noise than the CodeGen's two fans.
I do not have an Antec I want to dissect, but here is a word picture of the CodeGen 500 Watter:
The CodeGen I have has two fans also. The output side Rulian Science RDH80 vent fan pushes air to about (roughly) what a PanaFlo H1H does. The Input fan is a Rulian Science RDM80 and inputs a decent flow of air.
The Transformers are decent, they are Wyerls, four of them sized for the Rails they feed (and surrounded on two sides by EFI\EMI isolation and heat dump plates that are oriented so they run parallel to airflow for the biggest ones). There is also a Wyerl pre-transformer that can compensate for small slumps or surges, as it is also wired to ground plane for dump and is inline from the input jack on the PSU (where PSU cord plugs into PSU) The CAPS, the large ones, are HECs (HEC in oval). I'd give you visible pics but have not purchased the Camedia 4000 yet-- probably about a couple weeks before I order given finance flows right now.
So, rail details:
3.3 V, 20 Amp +
5.0 V, 45 Amp +, .5 Amp -
12 V, 18 Amp +, .8 Amp -
(Yes, I just took one that had been on shelf except for looking at in detail since tested 3 months ago apart for you-- IT will go in one of my boxes that will get another removable media drive shortly).
The CodeGen would benefit from the transplanting perhaps of quieter Panaflo fans into\onto it, but the other innards are VERY decent. They are more economical than the Antecs also, in my part of the world.
John Danielson.