• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

PSU tier list 2.0

Page 266 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
Hmm... I helped a guy here on Tom's Hardware who had a "poppy" PSU that made a popping sound when the switch was flipped to ON. I think popping sounds could be a sign of a bad PSU.
 
Not necessarily bad, but almost certainly cheap. I've got a couple of switches in my house that'll 'pop' when I turn on the switch. That's just indicative of a decent sized load creating high potential and a really slow switch movement allowing for spark jumping between poles before the poles are fully connected.
 


Hmmm..... interesting I had similar experience with them with the same model. Sadly since the launch of H100V2 there have been no leaks to get firsthand experience of Corsair support team.
 


Don't know why you would WANT a leak to happen.
 
As a consumer you don't. As a builder you do. It's quite easy to just read up on 10yr old reviews of AIO's leaking, and then assume all AIO's will leak, it's another to take brand new, revised, fixed, tweaked units and honestly say you've had leak problems or not. Just spouting nonsense that AIO's will leak and are noisy is just repeating old, ill-advised garbage, no different than big air towers will break a mobo or your 700w TT TR2 is a good unit because you've used it for years on your i3 and GT 720. Builders want to know what will break, what will fry, what will leak, if only to discourage clients from purchasing/using that stuff.
 


I don't want it to leak. I was saying that Corsair improved on quality of its AIO so much that there is no problem of leakage as feared by many. I have never seen a Corsair H100i v2 or any other AIO from that generation and up leak and cause trouble for customers.
 
Anti surge is effective and saved many people from damaging CPU and Memory on the loss of motherboard which was still covered under warranty and got replaced no questions asked. As I said my friend had fried motherboard but CPU and Memory was still working thanks to Anti-surge protection from board
That doesn't mean anti-surge works, or did anything. It means that the board failed, and other components didn't. Far more information would be required to determine the actual cause of the failure and whether other components were at all at risk, and whether software had anything to do with it. A major design criteria for most step-down devices is that failures and overvoltages do not propagate to the secondary side.

Such damage usually happens far too fast for software to react. Crowbar circuits, silicon avalanche diodes, and fuses/fusible resistors are what is typically used.

He got it replaced the board for free by visiting local ASUS service center in a weeks time.
That's typically called a 'warranty', at least unless they can demonstrate it was caused by an outside effect that the board should not handle.

Sadly the Gigabyte GPU did not have that feature and was missing the clause surge protect from warranty so he had buy new one.
If Gigabyte can show that the GPU was damaged by an out-of-spec supply, they are entirely within their rights to deny the warranty, just as Ford won't warranty your diesel truck just because you put petrol in it, or because you did 100km/h over a speed bump. Outside influences are generally for insurance, not warranty.

Hope brands like ASUS Gigabyte MSI and EVGA start including power surge in warranty for GPU's.
Only insofar as it incentivises them to improve overvoltage protection. Otherwise, it just encourages people to cheap out on the PSU.

Losing SSD and HHD is not as painful as losing GPU which is usually most expensive component of the build.
GPUs can be replaced, data can't.



Good old inrush current. Not really a major issue, especially for a switch that is mainly used for isolation and rarely switches load.


It's mostly about the power factor of the load. Lots of capacitors, lots of inrush, lots of spark. Not really high potential (though breaking an inductive load definitely can be).

Most mains switches are required to be snap action for this reason.

/grumble
 

Anti-surge protection is a hardware feature and I think it has nothing to do with software side. When high voltage of current hits the board the fuse cuts out the power from being supplied to CPU and Memory thus saving them from surge. This is an instant effect that takes place.


I think you missed the part where I said that Motherboard was replaced thanks to the warranty provided by ASUS covers the surge protection. So yes it was definitely covered in warranty.


I have nothing against Gigabyte. I only wish GPUs have similar kind of warranty coverage for power surge the way Motherboard is covered.


I expect atleast high end GPUs to have this because people spending on high end stuff are sure to buy good PSU but if it fails then there should be some kind of coverage for consumer who loses a costly component.


Full 100% data recovery is possible and can be done at far minimal cost($15-$150Max) compared to replacing high end GPU($400 and up).

 
I am not entirely sure it is possible on SSD. HDD yea it is possible. Still getting a proper quality PSU avoids the issue altogether. I learned my lesson, the hard way. Fortunately the system that got hit by the junk PSU's death was an inexpensive rig I used to hook up to my TV. It was mostly old parts, other than the Sempron 145 and cheesy 760g biostar board I had in it. Replaced it with an Phenom II x2 and a free 760g biostar board, from microcenter, and an OCZ 500w PSU. Not the greatest, but better than the craptastic Rosewill it replaced. I was still a bit uninformed back then. I was just starting to understand PSU quality. I trusted Microcenter's people that it was decent enough, for what I was using it for, which it was, in all honesty.
 
$15-$150 max? Really? It costs over $5k to have someone recover the data from a dead HDD. Data on a dead SSD is gone. Many ppl spend hundreds or even thousands of man-hours on projects and if the master copy is destroyed, you are now left with a backup that may or may not have current data.


Storage: Intel - P3700 1.6TB PCI-E Solid State Drive ($1920.99 @ SuperBiiz)
Storage: Intel - 2TB PCI-E Solid State Drive ($2399.99 @ SuperBiiz)
Storage: Intel - DC P3608 4TB PCI-E Solid State Drive ($6789.89 @ Newegg Marketplace)
Storage: Samsung - 850 EVO 4TB 2.5" Solid State Drive ($1523.99 @ SuperBiiz)
Storage: Samsung - 860 Evo 4TB 2.5" Solid State Drive ($1373.55 @ Newegg Marketplace)
Storage: Samsung - 860 Pro 4TB 2.5" Solid State Drive ($1899.99 @ Amazon)
Storage: Hitachi - He6 6TB 3.5" 7200RPM Internal Hard Drive ($635.00 @ Amazon)
Storage: Western Digital - Gold 8TB 3.5" 7200RPM Internal Hard Drive ($688.95 @ Newegg Marketplace)

That's just a sample. When talking about maybe replacing every storage device in even a household NAS if that psu surges, it could very easily extend past $20k just in parts alone, never mind what data was on it. Should have backed up the data? That was the backup.

You do realize that Microsoft Office alone costs over $300, that's not to mention other software like Adobe Premier or Vegas Pro or any of the other production software.

$15-$150? Get real.
 


You are probably talking about high level server or workstation data recovery or some other high end stuff.
The SSDs you listed are all used in workstation build or builds used for productive purpose. Builds like that usually pack in multiple GPUs worth $5K or high each which again is much costlier than SSD.

I got my WD 500GB HDD recovered which had its internal disk badly scratched for INR2,500($40) they also recover data from SSDs and even Memory cards.
 


They are providing recovery service for SSD and Memory cards. I think Seasonic Focus Gold Plus is high quality component which failed and took down Motherboard, GPU and Storage with it. From this experience I learnt that no matter how high quality stuff one invests on there will always be a possibility of failure it may be minimal but it surely exists.
 


You keep believing that. I've had clients AND friends, in situations where I was personally involved with negotiations, that have dealt with the very highest end labs in the United States AND one VERY high end lab outside the US, and in all three cases, where a lot of money was spent (Although none was spent in one case due to the guarantee of no recovery of any kind equals nothing owed), nothing could be recovered.

These were not drives damaged by severe electrical issues either. Just drives that stopped working one day, well within the warranty, so not even that old really. Yes, in some cases data can be recovered. In many cases, NOTHING can be recovered. Not even partial data, in a clean lab, by the best technical staff in the world. And even when it CAN, it costs FAR more than the best gaming card would run you so it's a lose-lose either way. Serious thinking error with your thought process on this one.

Minimum data recovery cost you will see charged by ANY and ALL reputable labs, for even PARTIAL recovery of data, is about 1500.00. This is not an opinion. This is a fact being relayed by somebody who has been intimately part of the process at the highest levels on more than one occasion.
 


i5-7600K
ASUS Z270-A
Corsair LPX DDR4 16GB
Samsung 750EVO 250GB
WD 1TB
Gigabyte GTX1070 Gaming Series
Corsair R200
Seasonic Focus Plus Gold 550

Cyberpower UPS not sure which model. That was bought by my friend. I don't think it is UPS problem because it was perfectly fine after that incident and the router and monitor connected to it are also not damaged.
 
While I'll take your word that that's exactly what you were told happened to that Seasonic, I've personally never in 30 years of funkin around with pc's ever heard of a Seasonic doing that on its own. Even the craptasyic of Corsair psus are known to just die quietly, even with basic ATX protections. So either this was an act of God, a fluke of Nature or your friend tried cramming the pcie into the EPS.
 


I had a friend recently do this! Killed everything! He wasn't paying attention to the cables even though there was a warning!
 
Yeah, and the pinouts for eps and pcie are opposites...

It's not an entirely uncommon happening, especially on larger units that'll actually have 8pin pcie. The EPS can be just as hard to seat as the mains, so unless you pay attention to the connector (there are 8pin pcie and 8pin eps, not just 4+4) mistakes can happen.
 
Stellar data recovery Nehru Place is charging decent amount for recovery bit high than what I quoted above but not too high. For electrical surge/PCB failure(this is what our concern here) INR3,500 for HDD INR5,000 for 2.5"SSD INR8,000 for M.2 for storage limit up-to 2TB and non encrypted data. For above 2TB there is increase but probably minimal than what you guys were expecting. If the consumer is not sure of the problem(most consumer are not) then they charge extra INR650 to find the cause of failure in addition to cost applied to recover the data.

5K USD for recovery is crazy amount. These guys will probably recover entire data on a local server for that price.
 
Stellar data recovery Nehru Place:
Read about INR49,000 (730$) for some hard drives in the feedback comments / reviews of google. For physical damaged hard drives like your friend´s SSD, the charge is much higher there too, if even possible to recover the data.
"Physically damaged (dead hard drive, not detecting hard drive, not responding drives etc.), data recovery charges are different"
 
Status
Not open for further replies.