Sign in with
Sign up | Sign in
Your question

Lots of money on PSU = Lost of wasted cash!

Last response: in Components
Share
October 2, 2006 4:40:50 AM

I am going to start a very big arguement here... and will probably let it go on without me... But, I have a very different approach to computer purchasing when it comes to PSU's

History & Basics

First, Power Supply Units, are essentially AC->DC power converters, with a requirement of supplying multiple Voltages each at different wattages.

Dispite what the manufacturers want you to beleive... this is old school tech... You could build your own PSU using parts from 40 years ago, and with the knowledge that was around then.

What is true, is that the parts that make up a PSU, have gotten better... in some cases, MUCH better... they heat up less, and they provide less margin of error. They also cost less. In the old days you could make up with this margin of error with more detailed designs, fans, heatsinks, and more parts... Now with better part quality, simplier designs can be used.

The new stuff uses newer components, which provide some benifits... These benifits usually come as longer life spans, more power/less heat ratio, cost, veriable fan speeds, higher wattages, etc.

Of course you can go out and buy the BEST transistors on the market, and the BEST capacitors... and these units will perform really well at low temps, and most likely will live longer... and will definately take more of your cash.

But heres the catch... The better the parts, the lower the heat, so the less fans they put on it, so the heat goes back up inside it, but you dont hear it.. or they keep the fans and they get a higher UL approved wattage rating.

Life in the world of Physics

This is were it gets dicey... a UL approved rating is the maximum safe operation. Going past it is usually acceptable for short bursts. Under long term stress the components will die (and might catch on fire burning your house down). The UL rating is there for safety.

So a cheap PSU rated at a 600w UL approved rating verse an expensive 450w UL means that if you run the 600 at 500... the stress put on it is allot less, and the life span will get ALLOT longer. Run the expensive 450 at a steady 500w the life would be very short and your house could catch on fire and burn.

Now if you buy a cheap 600w UL rated powersupply... and you run it with any processor on the market, with a few hard drives and you leave it on all day, with occassional usage, it most likely wont die on you (notice I didnt include a videocard in the HW list)... A PSU of this magnitude will last for MANY years in these conditions.

Buy a 1,000w Power Supply, and the life span of that powersupply will probably outlive the voltage requirements for PC's in the future.

Its like buying a Ferrari, sure you can run the thing at 200mph for 1 days NON STOP, without a breakdown... but on the 2rd day, you find yourself needing a Ferrari parts dealer. (See Le'mans Racing)

If you take that same Ferrari and run it at 50mph, you would easily be able to cover 15+ days and allot more distance. Even if you put that SAME car in one of the HOTEST deserts in the world, and drive it for 15 days. (See Road and Track about 4 months ago),

REALITY
Every PSU that has ever gone bad on me, started to do the same thing... The machine would run for awhile, then I would do something with a heavy load, and it would shut down... or it would randomly just shut down. I have NEVER lost a computer part due to a bad PSU... Not once... Not on any of the machines at any companies I worked at, or any of the machines my family, and my extended family, and friends run... I know because I am often the one all of these people goto when the computer starts acting weird..

My fathers neighbor was hit by lightning, and the static and power surge burnt out a telephone, the electric dog fence, and a few other minor devices. The only computer parts we lost was a router and a single network card... all of the machines run fine, and are on 24 hours a day... The surge protectors are the cheap ones you buy at any store, and they didnt go off...

The time to Buy an expensive PSU?
In servers we often run 2 Hotswapable PSU's each on their own Generators... These PSU's are generally of high wattage, for longer life... We aircondition these rooms, and the rooms airconditioners are also run by either generator...

If you want 2 video cards, then buy a higher watt PSU (600w).

If you want quiet, then spend extra money.

If you want moduler design, with nice wires, lights, etc... Then spend more money.

If you want to save yourself grief from THG Forum PSU Nazi's who will blame everything from why your car doesnt start to the reason your hard-drive went bad on your PSU, then spend more money...

What this reminds me of
I am a bit of an audiophile. Do you know that you can spend $10,000 on 3 feet of SPEAKER cable... Yes, wire... from the Amp to your speaker...

When scientific studies have been performed with direct comparisons of this REALLY expensive wire verse a normal lamp cord wire... There was NO proof that one wire sounded any different then the other... These tests used $110,000 speakers, with the best equipment available... Yet no one could tell the difference between the $10,000 wire and the $10 wire.

Now go to ANY audio store, and see how many "High end" audio cable you can buy... Ask the sales man if that wire makes any difference? He will most certainly say yes...

Go to a High end audio store... and go listen to there $10,000 plus speakers... and look at the 4" radius wire running from the speaker to the amp... then ask the sales man if it makes any difference. He will say yes...

Then ask him to do a double blind study... and watch how he skerts around triing to explain how difficult it would be to set up.... hmmm... a Drape, a $10 cord, and that 4 inch speaker, both out of sight... yep... impossible to set up...

Why? Because companies like to make you feel like your buying a Ferrari... They market themselves as the "BEST", and they often earn that praise... There powersupply probably has a 2% better chance of making it another year then the cheaper one does...

Take that
Mike
a b ) Power supply
October 2, 2006 5:10:56 AM

Hear, Hear could'nt agree with you more my friend. 8) BTW 10,000K on 3ft of speaker cable , you cheapskate :lol:  :lol: 
a c 83 ) Power supply
October 2, 2006 5:24:25 AM

I think we are on the same page, but I'm not sure. First, have you ever ran one of those REALLY cheap PSUs? I think we agree, but I'm checking to see what you think cheap is. I'm talking about one of these:

http://www.newegg.com/Product/Product.asp?Item=N82E1681...

Check out this bad boy. 450W, 100% feedback rating, and a mere $21 after rebate counting shipping. Would anyone on this forum trust their build to this PSU?

While I don't suggest ANYONE uses this powersupply, there are many out there that will work for people in the $35-60 range. Antec, thermaltake, and others make perfectly good powersupplies that I would use in my own system. Sure, you could go hog wild and get an OCZ or a PCP&C, but most of us don't need something that high end. If this is what you are referring to, I agree. If however your claiming that you've never had a psu blow on you and we should stick to supplies like the one I linked you, you're nuts.

P.S. You forgot to talk about the difference between peak and sustained power. The one I linked to is probably 450W peak, while PCP&C and Seasonic list their supplies as sustained. Theres a big difference between those two.
October 2, 2006 5:53:37 AM

You lost your wasted cash? Did it drink too mush?
October 2, 2006 7:33:10 AM

Quote:
I have NEVER lost a computer part due to a bad PSU... Not once...


This is like saying "I never had a car crash due to a flat tire". I have not had a car crash due to a flat tire either.

I have seen psu blow-outs kill components, and they were very "light weight" power supplies. This usually means they lack components to keep the manufacturing costs down. Which circuits you ask? Probably "protective" circuits that keep unsafe voltage / current spikes from passing through to the outputs, aka a crowbar circuit. It's not a 100% failsafe but it greatly reduces the risk of component damage. A good analogy would be having a default install of an unpatched windows pc directly on the DMZ internet without a firewall, and why MS enabled the firewall by default in SP2.

Compare this:

COOLER MASTER eXtreme Power RS-430 for $34 shipped @ newegg. In the "description" tab you see this:

Over Voltage Protection YES
Overload Protection YES

The product link tells you more:

Protection OVP / OCP / OPP / Short Circuit Protection

Now look at this "light" power supply:

XION Simply 400W for $25.11 shipped @ newegg w/ a $10 MIR

There is no "protection" anywhere in the description. The product link is also devoid of that word. Why? Being UL listed does not require to have protective circuitry.

BTW those dual-redundant hot-swap server PSUs definitly have protective circuitry, as well as active PFC. I trust your dual-generators have the UPS power behind them to carry the load until the generator can get to up-and-running status. Seriously the generator is useless if you have utility glitches and every server has to crash / reboot on generator power :wink: get a redundant 200+ KVA ups system to compliment the generators. Ours are triple-redundant, we can take a UPS failure and still be load redundant 8) well, unless our generator is dead or out of gas :oops: 

As far as speaker wire goes - zip cord was my preference
October 2, 2006 8:45:57 AM

Quote:

Life in the world of Physics

So a cheap PSU rated at a 600w UL approved rating verse an expensive 450w UL means that if you run the 600 at 500... the stress put on it is allot less, and the life span will get ALLOT longer. Run the expensive 450 at a steady 500w the life would be very short and your house could catch on fire and burn.

Now if you buy a cheap 600w UL rated powersupply... and you run it with any processor on the market, with a few hard drives and you leave it on all day, with occassional usage, it most likely wont die on you (notice I didnt include a videocard in the HW list)... A PSU of this magnitude will last for MANY years in these conditions.
Take that
Mike


Here s a fundamental ERROR in your statments.

UL DOES'T RATE or approve any equipment.


UL is a testing laboratory which only LISTS equipment as complying with various specification.
UL doesn't accept liability for design or performance of the equipment.

UL by listing equipment confirms that the equipment with UL level is built in accordance to electrical and NFPA codes, and that the equipment is built and designed to safely perform under the rated by the manufacturers loads, and does not pose the injury hazards to users and operators.

http://www.ul.com/info/standard.htm

"UL tested products for public safety for more than a century."

UL label (listing) mean that the equipment as designed and manufactured is meeting the safety STANDARDS under rated by the manufacturer loads, and operational conditions and (i.e. NFPA) classifications.

i.e. it only mean that the i.e. wire gage is in compliance with industry standards for given laods, that wire will not brake, that the fasteners are properly attached, and unit wouldn't pose the fire hazards.
UL would refuse listing for hire loads than equipment is designed and built in accordance to industry Specs.

Man by the time you finished your thesis you could earn enough money for the most expensive unit!

It is called TIME management.

PSU is the most important componnent of the computer, so you wouldn't just buy the cheapest but the most stable unit under the rated loads.
There are differnt types of PSU


Basically your disertation is asking for ... trolls.

We a only discussing operstiomnal safety undre the given by manufacturer load.







.
October 2, 2006 2:10:59 PM

Quote:
Man by the time you finished your thesis you could earn enough money for the most expensive unit!


You imply that BS has value...
October 2, 2006 2:30:06 PM

I'll pay to see the look on your face when your cheapo, crappo Ultra-brand PSU blows, and takes your $1000 CPU with it.
October 2, 2006 2:38:49 PM

Quote:
I'll pay to see the look on your face when your cheapo, crappo Ultra-brand PSU blows, and takes your $1000 CPU with it.


Oh, we'll hear ALL about it, and the sig will change...
October 2, 2006 2:46:22 PM

What about the inefficiency of a cheap PSU = lots of wasted energy = lots of wasted cash! Getting a highly efficient PSU will save you money in the long run.
October 2, 2006 2:49:55 PM

I have built 2 computers with the $25 Rosewill psu from Newegg (I think its 450 w I'm not sure) and neither have had any problems. However, they are only running smaller video cards and 1 hard drive. My PC however is running a 7900GT, and 2 hdds (one of them a raptor) with a new Conroe...So...That being said...I wouldn't use a cheapo psu for mine.

In most cases of life...you get what you pay for. There may be some exceptions, but its usually true. So I would spend more $ on a psu for a comp that I really cared about. EVEN if it is ONLY just for piece of mind that I have a 'name-brand' psu, it still makes me feel safer and that is important. I think in a lot of cases, the placebo effect of having name-brand IS important. For example...when you got the drug store...There's Equate or CVS brand for $4, vs. the name brand for $7. I bought the CVS version of Afrin (a nasal spray) and even though it had identical active ingredients...I swear it did not work as well. Nor did the CVS brand of acne med. So...perhaps its all in my head that the name brand works better...but...if thats what makes the difference, then so be it and I'll pay the extra for the placebo effect that somehow makes me feel better.
October 2, 2006 2:53:59 PM

Quote:

Life in the world of Physics

So a cheap PSU rated at a 600w UL approved rating verse an expensive 450w UL means that if you run the 600 at 500... the stress put on it is allot less, and the life span will get ALLOT longer. Run the expensive 450 at a steady 500w the life would be very short and your house could catch on fire and burn.

Now if you buy a cheap 600w UL rated powersupply... and you run it with any processor on the market, with a few hard drives and you leave it on all day, with occassional usage, it most likely wont die on you (notice I didnt include a videocard in the HW list)... A PSU of this magnitude will last for MANY years in these conditions.
Take that
Mike


Here s a fundamental ERROR in your statments.

UL DOES'T RATE or approve any equipment.


UL is a testing laboratory which only LISTS equipment as complying with various specification.
UL doesn't accept liability for design or performance of the equipment.

UL by listing equipment confirms that the equipment with UL level is built in accordance to electrical and NFPA codes, and that the equipment is built and designed to safely perform under the rated by the manufacturers loads, and does not pose the injury hazards to users and operators.

http://www.ul.com/info/standard.htm

"UL tested products for public safety for more than a century."

UL label (listing) mean that the equipment as designed and manufactured is meeting the safety STANDARDS under rated by the manufacturer loads, and operational conditions and (i.e. NFPA) classifications.

i.e. it only mean that the i.e. wire gage is in compliance with industry standards for given laods, that wire will not brake, that the fasteners are properly attached, and unit wouldn't pose the fire hazards.
UL would refuse listing for hire loads than equipment is designed and built in accordance to industry Specs.


I dont think I ever disagreed with a single thing you stated in my original note... I know what UL label is, and what it isnt, and I figured I made it clear in my statement what I ment. Now if I didn't, which seems to be clear now, then thank you for clarrifing what I was meaning.

I personally didnt even KNOW that you could buy one for $30... I WOULD buy one, if it was for an old computer... I do agree that their is a limit to what I ment by inexpensive, but it has less to do with cost then with design... as stated above by another individual.

The "protection" these expensive PSU's provide are like your surge protector, if your house or the electric pole gets hit by lightning then your computer would be fried way before the circuit breaker goes off...

Again, in my 20 years of experiance... both personal and commercial experiance, I have seen and repaired hundreds of computers, and not one of them was damaged by the PSU. This idea that the hard drive, motherboard, optical drive, video card (and other) manufacturers dont provide protection for surges in power is crazy.

The biggest threat for a surge/drop in power isnt from your PSU, but from the electric coming INTO your house. I recommend you go out and spend $100 or more on a UPS. Not because they protect you against total power failure, but because they are the single best way to ensure a constant power supply, without spikes or drops. I have a 7 year old belkin, and the battery in it was suppose to last me for 2 years... I just bought a new battery... With shipping (and this was REAL heavy) it cost me $20... I have used this UPS for 3 different computers during its lifespan, and I only have to replace it because after 450watts it starts to fail. Its a good investment.

I see these $200+ PSU's on the market and laugh... I think, what sucker is buying these... YaY, you spent $100 for less then 1% protection of your equipment... hahahaha.

There are some $100-$200 dollar range PSU's I would buy... but I would buy them because they offer 600+ watts and are ultra-quite.

Mike
October 2, 2006 3:01:44 PM

BTW, in the life span of that single UPS, it has seen more then 3 different $100 PSU's die on it... One was a 350 watt. then 400 watt, the 450w...

I recommend buying high watt PSU's especially when you run low watt systems, cause its a pain in the ass to replace powersupplies... and every manufacturer admits that their power supply is ment to run at a certain load (usually 75% of the listed raiting).

for thoose who beleive that supper expensive PSU's are going to save the world, then I recommend to them these speaker cables for there $80 radios...

http://www.goodwinshighend.com/cable.htm
October 2, 2006 3:08:48 PM

Quote:
I recommend buying high watt PSU's especially when you run low watt systems, cause its a pain in the ass to replace powersupplies


Are you recommending cheapo 600W PSUs with 10A on the +12V rail?
October 2, 2006 3:13:17 PM

Quote:
This is like saying "I never had a car crash due to a flat tire". I have not had a car crash due to a flat tire either.


Its funny you should mention tires... Thoose $200 bridgestone tires on Ford SUV's... You know, the Top of the line ones! How many people did those tires kill?

Quote:
What about the inefficiency of a cheap PSU = lots of wasted energy = lots of wasted cash! Getting a highly efficient PSU will save you money in the long run.

I cant even reply to this... It is clear you dont know how the PSU, or any electronics, works... An assumption that a Expensive PSU will save you cash on power is CRAZY, and REALLY wrong...

BTW, the Ultra brand PSU I bought has one of the best support department I have ever worked with... I was checking to see if the fan temp gaudge was broken, and they were willing to ship me a new one, without any cost to me on shippment, and was willing to do so without me returning my current one first... I couldnt be happier with the ultra brand, and would recommend them to anyone... Except they are a bit noisy for people running higher watt systems... BTW, I never needed to replace it... and because the Ultra engineers where smart in their design, I was able to create a PSU with the same top end clearance as bottom end. This allows me to install my PSU upside down, allowing me to pull air from outside of the case!

Mike
October 2, 2006 3:27:30 PM

Quote:
I recommend buying high watt PSU's especially when you run low watt systems, cause its a pain in the ass to replace powersupplies


Are you recommending cheapo 600W PSUs with 10A on the +12V rail?

I dont recommend any defective product... It is easy to find out if a product is defective or has a history of problems... just do a google on the PSU... Good PSU's receive lots of High ratings and reviews... Look at new egg, and see what people think... Only people who have purchased the item can complain/praise a product, so you dont get people who havent used the product themselves...
October 2, 2006 3:54:08 PM

Quote:
Its funny you should mention tires... Thoose $200 bridgestone tires on Ford SUV's... You know, the Top of the line ones! How many people did those tires kill?


They should have bought a Hummer :?

"ford" and "top of the line" in the same paragraph 8O
October 2, 2006 4:33:35 PM

The "PSU Nazis" are that way for a reason, it is because they have dealt with so many guys that are having system problems because they are running $3k worth of gear off of one of those perfectly good $38, 600 watt PSU's. I have seen so many issues with these "perfectly" acceptable PSU's that I have lost count. Sorry, but your premise doesn't match my experience. I also know folks who have lost components to a PSU going bad.

As far as your 20 years of experience, which I respect, that is fine but are you saying that you regularly put low end PSU's on your servers? Are you also saying that more efficient PSU's don't save any money? I don't think so and I agree that the cost savings of having a highly efficient PSU on a home rig is minimal at best, maybe a few doalrs a year at most, but generally the more efficient PSU is also the more expensive. Also, I have seen those cheap-o 600 watt units drop to 11.4v on the 12v rail. That is within standards but the high end systems sure don't like it.

Sorry but I can't agree with most of what you say. No, you probably don't need a $200 PSU but I would never recommend you match $3k worth of gear with anything that costs under about $100. Guys that do are usually back in the forums fairly quickly with all those weird system problems that the cheap-o units cause.

Guys, if you are building a high end rig, please do yourself, and everone else on the forums, a favor and get a quality PSU with plenty of amps.
October 2, 2006 4:46:22 PM

There are a few simple rules you can live by.

1. You only get what you pay for....some of the time.

2. You NEVER get more than you pay for.

3. Going cheap is usually the first step of getting fired.

Just some rules of thumb from my business experience. :) 
October 2, 2006 4:51:44 PM

Quote:
I cant even reply to this...

I'm not surprised... Oh wait, you tried anyway:

Quote:
It is clear you dont know how the PSU, or any electronics, works... An assumption that a Expensive PSU will save you cash on power is CRAZY, and REALLY wrong...

So you don't understand efficiency? Check. Not surprising...

Quote:
BTW, the Ultra brand PSU I bought has one of the best support department I have ever worked with...

That's good for you. I suspect you'll be in touch with them again... and again...

Quote:
This allows me to install my PSU upside down, allowing me to pull air from outside of the case!

An upside-down PS seems oddly appropriate for you. I bet that you spent hours trying to figure out why the A/C plug wouldn't go into the receptacle upside down too. Do the upside down electrons underclock your 6800?
October 2, 2006 4:57:43 PM

Quote:
I would never recommend you match $3k worth of gear with anything that costs under about $100. Guys that do are usually back in the forums fairly quickly with all those weird system problems that the cheap-o units cause.


If you look back at Jelly's posting history, you'll see that he's talked about replacing MANY cheap power supplies - with more cheap power supplies. The data is in his face but all he sees are the LED photons emanating from the fan of his Ultra. I bet he's on his knees right now, worshipping his bling. Curtains pulled tight, room lights out, bummer, sunlight leaks lower the LED contrast... Worship LEDs... Worship LEDs...

Jelly, you're the master of bling. Leave the power supply discussion to people that understand the facts.
October 2, 2006 5:09:58 PM

Quote:
What about the inefficiency of a cheap PSU = lots of wasted energy = lots of wasted cash! Getting a highly efficient PSU will save you money in the long run.


Crunch some numbers on a typical 500W @ 75% vs 500W @ 65%.

You will be surprised on how much it actually saves you.
It is not as big as some people think. And remember the difference is usually there only when the PSU is being pushed to the limit.
October 2, 2006 5:48:52 PM

I don't want to be involved in the argument or "discussion" :p  but I am curious if anyone has ever heard of or seen an Ultra PSU die. I know of at least 4 people that use them and 1 of the guys has an SLI setup (2x 7900 gt) and an overclocked D915 running a water kit and he leaves it on almost 24/7 and it is still going strong.

I like my PSU and haven't had any issues with it. Just curious if anyone has a horror story or something. I have yet to see a PSU damage a component, though I am not doubting the fact that it happens.
October 2, 2006 6:09:05 PM

Can't say I entirely agree with your theory. I recently had the experience of using a bottom-feeding power supply that was part of an Aspire X-QPack case. The thing fluctuated power, going up to 20% above spec on some rails, and oscillating up and down like crazy. This was being measured from the motherboard, but the same board, when combined with an Enermax Liberty or Antec PSU had no issues.

I never lost a part that I can point to the PSU, but I can't imagine running at 20% above rated voltage would be good for equipment life-spans. I only had the case for a week before I sent it back. Oh yeah, it blew air from outside INTO the case, pre-warming it nicely for all the components inside. First one I've ever seen do that. And especially dangerous with a small cube-case... The unit was fully factory sealed before I got it, as well.

Clint
a c 83 ) Power supply
October 2, 2006 6:28:15 PM

Quote:
I cant even reply to this... It is clear you dont know how the PSU, or any electronics, works... An assumption that a Expensive PSU will save you cash on power is CRAZY, and REALLY wrong...


Wait, hang on. I need to pull up a good seat. PLEASE explain this seeing as we don't understand how this is wrong. I need a good laugh today.

(btw, although I'm sitting down, you might need to stand up if you're going to pull something amazing out of your @$$.)
October 2, 2006 6:34:31 PM

Quote:
I don't want to be involved in the argument or "discussion" :p  but I am curious if anyone has ever heard of or seen an Ultra PSU die.


I have not been present at the scene of the crime, but I've seen the aftermath. I have a friend that's a Geek Squad guy and he showed me the Best Buy PS boneyard a few months ago. Our discussion started when a guy brought in a cathode-loaded game box that had died and I watched as the tech determined that the PS was dead. It was an LED-enhanced Ultra and when I looked in the PS graveyeard, there were a couple of other Ultras. I mentioned that and my friend told me that the Geek Squad guys replace Ultras often. He's a straight up guy and I do not doubt his honesty.

At that time, I happened to be on a mission to study power supply reliability because of problems we were having with computers we were using in the field in harsh environements. So I talked to service groups online and at local shops like CompUSA, Circuit City and PC Club. All of them mentioned Ultra as being particularly unreliable. I have seen decent reviews of the most expensive Ultra unit but I'd have to see quite a bit of changing attitude in the service industry before I'd go there. There are just too many builders out there with a good rep to go risk an Ultra. I hope you have a good one because it is such a bummer to have your PS take out other components. Been there.
October 2, 2006 6:39:12 PM

I don't know, maybe I'm just coming from a different angle having had to design several power supplies for personal projects and such, but why has no one mentioned ripple? PSUs take in AC and turn it into several different DC values, all these have some percentage of ripple on them. I can speak from experience that this ripple can cause system instability if it's too large. The CPU tends to have it's own power electronics right around the socket to eliminate ripple as much as possible, but most components do, but even the extra electronics for the CPU can't help if the ripple is too large. Guess what, cheap PSUs tend to have a larger ripple.

And as for never seeing a system destroyed by a PSU, you are truely rare. Most modern power supplies use things called switched converters. They create 1 DC value off of the AC, then throw it through these switched converters to get the desired voltages as well as lessen the ripple off even more. If these are implemented with diodes, you tend to get something called a shoot through current with puts a huge strain on the transistors, these can cause the transistors to fail rather unexpectadly and can leave the system open to suddenly have to try to absorb up to 170 volts into the motherboard if you live in the US and up to 340 volts for you europeans. Sorry, but the best UPS in the world isn't going to save you if this happens. Of course the high end ones probably use magnetic isolation or even optical isolation so even if transistors do fail, due to the isolation, the oscilation through the transformer is gone and it stops producing power on the motherboard end, thus your components are saved. I find it hard to believe a low end PSU would have isolation.
October 2, 2006 6:45:06 PM

Quote:
I don't know, maybe I'm just coming from a different angle having had to design several power supplies for personal projects and such, but why has no one mentioned ripple?


All of your points are valid - good post. Perhaps the relevant technical issues have not been discussed more openly because many of the informed members of the forum are avoiding this thread. Read Green Jelly's posting history (just threads he started to save time) and you'll get my point. Facts matter little in his world.
October 2, 2006 7:15:52 PM

Quote:
Crunch some numbers on a typical 500W @ 75% vs 500W @ 65%.

You will be surprised on how much it actually saves you.
It is not as big as some people think. And remember the difference is usually there only when the PSU is being pushed to the limit.


The PSU suppliers very rarely publish the efficency numbers, if ever...

The numbers I refered to where numbers based on the recommended normal operating wattage, and provide a basis for the PSU's lifespan.

Too say that a $200 PSU is better then a $100 unit is exactly my point... MORE MONEY DOESNT MAKE BETTER PRODUCTS!

A company that creates products using good quality improvement standards and management principles will grow and improve a product overtime. This will create products that continue to get better, without an increase in cost... and sometimes will lower cost... Your talking to a Management Information Systems man, who spends his time looking at how systems (not just computers, but human systems) work. I know Q3, ISO9000 and many other quality control practices and principles. I go into a company, look at how products and services are provided... Look at how we can improve them, and then start the process of improving these systems. In the end, I measure success if everyone in the company is looking to improve quality from the ground up... I talk to the guys who are building the machines all the way up to management.

If you look at what happened in the auto industry about 30 years ago. A man named Deming went to the autosuppliers in the US with a new idea (he called it TQM) on how to produce cars and products through a process of continual improvement... well the US car companies called him a quack and dismissed his ideas. So he went to Japan, and started working with Toyota, and Toyota listened to him... They put his principles into practices, and Last I heard, Toyota will be the largest producer of automobiles by the end of the year... Since the early 80s, American car manufactures have been struggling to keep up with Toyota, Honda, and other Japanesee car manufactures simply because they were earlier at adapting these policies.

Xerox, Kodak, and MANY other companies had major problems in the 70s because of there slow implementation into these practices. Now Xerox is one of the model companies that run consistant quality improvement practices (they use among other things ISO and Q3 practices). Everyone in the company, from the man on the assembly line to the man up top should be given the ability to improve the process of product development.

For example: I generally work with computer systems and their interactions with human systems, I dont work on assembly quality... But I got a new client, and as the case maybe, I felt it was important that I be placed on the assembly line doing basic work for a whole week. I was punching things on presses, and learning things I never new how to do... All sorts of tasks... but the part I was there for was to A) gain the guys trust on the floor, and B) it gave me insite on the companies product; in which I never would of gotten any otherway. I ended up recommending implemented red buttons, in certain places, that employees could press to shut down the assembly line if there is a problem. This single change saved the company a million dollars a year, and the product was ALLOT better as a result.

So dont go around preaching to me that more costly product's are better products, cause I can tell you first hand experiance, tons of examples, and hundreds of thousands of documents and books that will prove how wrong this is... and I can give you hundreds of instances in history of companies going out of business or lost major markets, resulting in downsizing because of the lack of focus on improvement. Some of you should go to school and take some business classes... you will learn allot...

Mike

BTW

Here is a link, its the first thing that came up on google... You can always look up "Dr. Deming Quality" on google. Or try "Q3 Quality Practices" and "ISO 9000"

http://www.dharma-haven.org/five-havens/deming.htm

If you think all companies have figured this stuff out, you couldnt be more wrong... Many (if not most) companies fail at this. Its the ones that do it, are the ones you buy products from.
October 2, 2006 7:19:18 PM

Can u tell the difference between your x6800 and an E6700?
Isn´t that a waste of money?

oh, sure, it takes less 50 second to export a 20Gb movie....

So, why can´t a guy have a top notch psu?
Do they need it? Do you need that cpu?:roll:

Edit: I thought the same as you, but when I saw that u have that cpu... I thought...you fuc*ed it all.
a c 83 ) Power supply
October 2, 2006 7:34:20 PM

Quote:
Too say that a $200 PSU is better then a $100 unit is exactly my point... MORE MONEY DOESNT MAKE BETTER PRODUCTS!


So you no longer get what you pay for? Is active PFC no longer better then no PFC? Buying a $150 PSU that has >85% efficiency is no longer better then buying a $60 one that is only 70%? Better is a subjective term, you failed to qualify it. I think you would be hard pressed to link to a $200 PSU and a $100 PSU and claim that they are equals. The $200 PSU will be better somehow. More wattage, higher efficiency, lower noise, etc, it will most likely be better in some way.

You might want to stand on your tip toes.
October 2, 2006 7:38:43 PM

Cyprod you are right in what you say... I know because I am an audiophile, and cheap amps use the same type of tech, both in their powersupply and in the way they produce the signal to the speakers... Im no EE, but I know enough about circuit design to understand the problems associated with these issues.

Cyprod, you failed to mention that power spikes from the cord in the wall is not very stable, and that failures tend to happen when the power spikes rapidly.

Most Car audio amps use really slow switching methods that work great on low end sound (Low KHz, aka BASE). The switches are too slow for the high end, and can cause distortion, but they have SUPER high Watt values. So people buy them, and say WOOPIE Im running 1000w... and you get into the car, turn the volume down, and you cant hear the damn high end... I feel like saying, HEY STUPID... YOU BOUGHT A SUB AMP!

The highend Audio Tube amps produce a soft sound because of their properties, and this soft sound is highly prized by many. Solid state amps can be just as good, but the tech costs allot of money... I use sound systems because they do the same things that your PSU's do... they just have to do it at a much more consistant and precise way with very expensive components... Mac amps are GREAT, but $10,000 for a low end 2 channel amp is beyond my ability... of course these techknowledgies are closely monitored, and people are VERY picky about them. Publications also do multiple quality tests, on each type of equipment as well as using the human ear.

My father has allot of engineering magazines, and I rarely see any mention of computer PSU's mentioned, except as mentioned higher and lower control values.

I see all these people on this forum pointing to this brand and that brand as beign bad brands, when they provide no technical references that support their claims... They dont look at the circuits, they dont use (or even know) what a scope is... And they certainly dont provide any references to people who compair power supplies with these approaches.

They simply say... This PSU is $80 and rated at 550w and this PSU is $100 and provides 500w and so the $100 PSU must provide "safer" power.

It is too bad that more computer resources are not testing these things, and coming up with solid reviews and experiments, posting results.
October 2, 2006 7:51:14 PM

Quote:
The PSU suppliers very rarely publish the efficency numbers, if ever...


You want numbers? Look to good ol' Newegg. Here's a PCP&C unit listed on the egg at 83%. What, that doesn't count? How about this Seasonic? Pretty impressive - up to 85% efficient it sez. Still doesn't count? For the OCZ, you'll need to go to their page. Wow, 80%, not bad. so you like Ultra, right? Let's go look for their published efficiency... I don't see it published on this web page of theirs. Hmmm, why do you figure they don't publish an efficiency? Let's look at another Ultra... Whaddya know! Even Ultra publishes efficiency data ! Look here. So basically, you don't even know the specs of the sh!t you buy? Face it Jelly, you're incompetent WRT researching and understanding hardware.

I'll try to leave this thread alone. You're a total waste of bandwidth. Have another brewski, dude... If you're not careful, you'll sober up on us.
October 2, 2006 7:56:44 PM

Quote:
Can u tell the difference between your x6800 and an E6700?
Isn´t that a waste of money?

oh, sure, it takes less 50 second to export a 20Gb movie....

So, why can´t a guy have a top notch psu?
Do they need it? Do you need that cpu?:roll:

Edit: I thought the same as you, but when I saw that u have that cpu... I thought...you fuc*ed it all.


hahah.... You are totally right... and if a $1000 PSU would have made me happy I would have bought it... without question... dispite what people said... So Buy the expensive PSU if it makes you sleep good at night... I bought the CPU cause it keeps me up at night!

In fact all of you are right... well not all of you... The problem I have is the lack of evidence, and of well designed methods of testing PSU's, that are published openingly. Everyone on this site, well almost everyone, are guessing as to which power supply in each price range is better. they also have no idea how much real world protection they get for there ext $50, $100, $150, etc...

hahaha
Mike
October 2, 2006 8:18:43 PM

Thanks alot clueless!
October 2, 2006 8:45:01 PM

Remember it’s not the Watts; it’s the Amps that make a good power supply.
And for PSU’s over $100.00, you are mainly looking at value-added things like LED’s, Plug-in headers, fan-less units and a reputable name (including the warranty that goes with it).
October 2, 2006 9:39:03 PM

WARNING! TECHINAL GIBERISH

I'm sorry, I wasn't going to do any more repies, but you just said something that is so incredibly wrong I just can't let it pass.

You said crashes can be made by "you failed to mention that power spikes from the cord in the wall is not very stable, and that failures tend to happen when the power spikes rapidly" and those it is true I failed to mention this, it's because of a bit of knowlege that I failed to mention. I apologize now, but I'm a graduate of an EE department, though I tend to work more in real time OSs, I've still done a lot with power supplies. First off, for a basic understanding of power electronics, I suggest you read "Fundamentals of Power Electronics Second Edition" by Robert W. Erickson and Dragan Maksimovic, ISBN 0-7923-7270-0. This was the course text for my EE562 class I took back in college and I warn you, it's a difficult read but it will debunk a lot of what you said if you can understand it, namely that the field of power electronics has remained fairly static for the last 40 years (the flourecent light bulb is an example of this, they used to use a ballest, now they use resonant tank circuits).

Also, it's known voltages change, it's AC, it's defined as a sinusoid oscilating between 0 volts and 172 volts, this is in the US, please forgive me. This produces the 120 volts RMS that we all know and love (115 would be more accurate after line resistance, but meh). The power company only gurentees to deliver this within a 20% error margin, and it's not really a gurentee, but an attempt, as such, we're already assuming that the PSU has to take an INCREDIBLY unstable power source and a) not fail and b) make something relatively stable.

Anyway, first off, power spikes tend not to damage electronics because except in the case of reverse biased components, it's usually current that destroys the part, not voltage. Since the first thing one does with any power supply is take the AC voltage and put it through a N:1 transformer to drop that voltage, large current spikes can't happen because transformers are basically two coupled inductors and inductors will not allow current through them to instantaniously change. Now, a large enough spike will cause a decent change in current, but we're all in agreement that a lighening strike isn't exactly a normal spike.

Now, we all know voltage spikes do occur. Well, capacitors do not allow the voltage across them to change instantly. It will absorb moderate sized spikes assuming the capacitance it adequatly large, not only that, but foil capacitors using aluminum foil and aluminum oxide for a dielectric can actually repair themselves in the event of minor failures, neat huh? As such, because of the use of inductors and capacitors one can adequatly absorb entry spikes, in fact, it's long term dips that are more likely to cause the computer to shut off, long term being on electronic terms of around half a second, though high end PSUs such as those made by Advanced Energy can survive up to 2 seconds with no power at 10KW+ without backup, allowing enough time to safely shut down hardware.

Enough of that tangent, continuing on with the spikes, though the inductors and capacitors do a lot for creating a rough DC source, we still need multiple voltages, and usually that first step isn't very clean DC, and though you could run it through a zenor diode to clean it up a lot, that wouldn't be very efficient. Instead you use these switched convertors I mentioned. These are neat because they operate in the mid 90% effiency range often. Basically you open a transistor with some voltage on the back side through an inductor, the current through the inductor begins to ramp up, then you shut the transistor off and the current is then pushed through the load, the load wants to stop said current, but the inductor by nature won't allow it to, which causes the inductor to push harder causing an increase in voltage. Put this in tandem with a capacitor and it's voltage will begin to ramp up as well, while the current in the inductor ramps down. when current through the inductor ramps up again, the capacitor maintains the voltage and feeds the circuit. Change the confgiuration and you can drop the voltage instead of increase it.

Why did I mention how these work? Because of how this switching works. The voltage result is determined by a sensing circuit, which can be a simple as an op amp, which senses the output voltage and/or current and can adjust the duty cycle of how long the inductor is being ramped up. If the input voltage on this converter increases, and we're dealing with a boost circuit, it will automatically drop the duty cycle so the output voltage remains constant, likewise if the input voltage drops, it increase the duty cycle to the same effect. As such, any remotely well made PSU will not be affected in the slightest by normal minor power fluxuations, even if they are constantly happening like say in the case of a battery which the voltage of which is constantly dropping from the moment you begin using it.

Now, onto why cheap converters are bad. In one word, diodes. Diodes for the unintiated only allow current to travel in one direction (we're ignoring zenors again here). This is in an ideal world, fact of the matter is, diodes take some time to turn off due to parasitic capacitances which must first discharge. As such if the voltage across a diode were to instantly change, like what happens when they inductor goes from ramping up to ramping down, the diode will allow current to freely flow across it as a short for a breif amount of time. This "shoot through current" can be large. I've messured as much as 20 amps in really poorly chosen diodes. That means that the transistor which has just been turned on now has 20 amps going though it. Not many are designed to handle that and this shoot through slowly destorys the transistor over time. Cheap power electronics use diodes though because otherwise one needs to use complementory transistors which requires two driving circuits instead of one. Diodes automatically turn themselves on and off without the driving circuit. Then you consider that these converters usually operating in the KHz range, sometimes even low MHz range, and you see that this shoot through can become rather problematic. The high frequency is used because it allows one to use smaller capacitors and inductors to no ill effect.

Okay, I'm sick of typing, I'm leaving it here, I think I've hit all my major points and hopefully put up a reasonable argument why cheap PSUs are bad without going into isolation, cuz that becomes real complicated and is usually only used in high end PSUs and medical equipment.
October 2, 2006 9:52:59 PM

Quote:
I don't want to be involved in the argument or "discussion" :p  but I am curious if anyone has ever heard of or seen an Ultra PSU die. I know of at least 4 people that use them and 1 of the guys has an SLI setup (2x 7900 gt) and an overclocked D915 running a water kit and he leaves it on almost 24/7 and it is still going strong.

I like my PSU and haven't had any issues with it. Just curious if anyone has a horror story or something. I have yet to see a PSU damage a component, though I am not doubting the fact that it happens.


I have asked that question in another thread, and no one ever responded that they did... Nor would anyone even point to a link of someone who had a problem with them... I asked over and over again, and the only response I got (after about the 3rd time) was that Clueless said he saw a article, that was taken down, that said Ultra had a negative review. But it was for an Ultra product that they made that was a lower costing version me and you have. Not to mention that the review has disapeared.

That was before I started ignoring Clueless.

I have looked through 15 pages of google for a single negative review, and after that, I got tired of looking... If someone could provide proof that the ultra is a bad product, I would be SO quick to replace it... Some reasons people around THG are anti-Ultra and pro-Asus...

Now I have only had good experiances with the Ultra, and only bad experiances with the Asus Mobo... Go figure...

I shouldnt have listened to all the people recommending the Asus Mobo on this site... My product isnt very good... Now maybe the Deluxe version doesnt have all the problems, but I am use to companies that produce a product line that are all simular in construction and that use the same bios... Like I said earlier, quality comes from the process and not the cost. When a MoBo manufacturer creates atleast two seperate boards, and 3 different bios's for the same product line, it leads me to think that company has a problem with their quality control and improvement.

I had someone tell me that Asus and Gigabyte merged and he said that Intel dropped the BTX standard. He provided 3 links that stated Asus and Gigabyte merged... But well, they never actually merged... It was misinformation. So much of what is said and reported on the web is simply misinformation. So I try and confirm what people say.

If someone could provide a link that states the Ultra is a weak powersupply because they have hardware problems after installing one, then this too may be called into question, because the idiot may have shocked the system because hes didnt ware a ground strap. Now If their was 4 or 5 references to failure of devices directly linked to the Ultra, it would catch my attention.

When someone claimed the Ultra was a poor product I did everything possible to prove their claim correct. I simply cannot find anything to back up their statements.

They then tried to claim that the Ultra was made by another company other then Ultra, and that this company made powersupplies for tons of other companies... They reasoned that this made the Ultra a poor product. I think a company that makes hundreds of different powersupplies for many different other companies, is a positive thing. That company would have gathered a huge amount of data on different types of power supplies, and are more likely to build a quality product and to use and implement quality standards and practices. It also makes me feel confortable that the head guys at Ultra shopped around and looked at many different power supply companies, and choose a single company because of their price/quality ratio. Compition creates good products.

I have no doubt that there are better products then the Ultra... Maybe even at lower costs, but their is NO data...

I wouldnt recommend buying a $40 powersupply, but hey... if it works and doesnt cause problems for you, then you just saved yourself a ton of money. I cant even beleive they could build powersupplies for that amount... I bet you Dells have really inexpensive powersupplies in their machines, but no one complains about it... So a $40 powersupply maybe the same type and design you get from a Dell.

Who knows.... Like I said, no one I can find is testing these powersupplies opperating limits...
October 2, 2006 10:09:19 PM

Cheap ass PSU means big pain in my ass.

I recently built a new budget system for a customer, I opted for a mid range case that included a 400W PSU for £30 ($55).

I week later PSU dies. Great, just what I need! :evil: 

Drive to customer and replace the dead PSU but can't RMA it because it was included with case and my customer doesnt want to be without their brand new machine.

Next day customer calls back because DVD-Writer does not work. I presume that I have stupidly not reconnected power to DVD Drive. :cry: 

Take another trip to customer to find that DVD-Writer is blown also!

Order new DVD-Writer and then drive again to customer with replacement. Test everything including his favourite game and system now running OK.

PROBLEM: customer lives 30km round trip away so 4 trips = 120km + 7 hours of my time.

I made approx £110 ($200) on new PC but then lost £25 for new PSU + 7 hours of my time + petrol equalling a net loss for that sale. :cry: 

Cheap ass wannabee PSU's should be made illegal.

Lesson learned....Now all my custom builds come with top notch PSU. :D 
October 2, 2006 10:31:47 PM

cyprod

I knew some of what you were talking about before starting the thread, and learned more... I never ment to state that the world of electronics is static... for if it was, we wouldnt be having this conversation since their would be no computers... What I tried to say is that 40 years ago you could build a reliable powersupply for todays computers.

Would you agree or disagree that a very steady current coming from the wall would decrease the chance of failure from any type of AC->DC converter?

Im asking this because a UPS is infact a battery, aka compacitor, provides steady power at standards much tighter then the powercompany provides?

This is important, because UPS's often cost around $100-$200 but unlike powersupplies, they seem to last FOREVER! So if a UPS provides longer life to powersupplys (Even Diode based powersupplies), wouldnt it be a great investment to buy a UPS?

My father is an EE too, and we have talked about this... but he is unaware of whats going on in each powersupply because you cant take em apart without voiding the warrenty.

My experiance is with $80-130 priced powersupplies, and they have not caused any damage to computers. Of course their is a margin of error to my observations.

My father plugged in a motherboard which must have had a metal shard on it... The thing smoked and sparked... but we still use that computer and Mobo, and it never crashes. Now thats allot of luck, but it also points out that these components are rather tough. I dropped a hard drive down a flight of stars, and that hard-drive lasted 4 years before it died... Again luck...


...and on the other side, I was writing a DB application on a local computer... it was a week long project so I didnt create a backup... I was working onsite with the consumers computer. I tried to insert the a MSDN disk, and it wouldnt work... Its a Dell, and Im not responsible for the computers operation so I called the IT guy. He came over, and instead of removing the CD-Rom and shaking it, he decided a good slam of the fist would fix it... Well I had no more problems with that CD-Rom!
....The hard drive was now broken.
October 2, 2006 10:49:31 PM

PSU may have caused the problem, but you cant rule out just plain bad luck... Especially when you said it was the optical drive... You should have had a second PSU available, so that you could swap them out, and then return the bad one for the RMA.

Monkey, Optical drives have horrible reliability... take it from me, I use mine heavily, and they have died on me so many times:(  I have 3 in my machine, and its not uncommon for me to send em back... infact I just had a NEC go on me, after less then a month of usage.

Ive lost 3 Plextors, 2 Lite-on's and now 1 NEC... Almost every one of them has been on a different powersupply and/or different machine.

I have had to RMA, RMA'ed Drives. I wish it was easy for us to point out a single cause and be absolutely sure that was the culpret, but there are SO MANY factors...

Like why didnt the HD fail... or the VC, or the CPU, Mobo Etc? Who knows...

I would however not recomend you sell/build machines for people who A) Live so far away from you... and B) If you need to visit their house to do the repairs... Next time, tell them to go buy a DELL. Have them deal with this s@#t


I personally wont build machines for companies. I tell them that DELL's are the best, because they support them. For me to recommend them to build machines, they would have to keep a stock of replacement parts to swap out in an attempt to find the problem. They would have to spend hours looking for the issue, and then they might have to re-install all the software... Go to windows update, etc...

Too much time...

Dell builds so many machines that they create procedures for repairing them. They make software installation CD's/DVD's that dont require all the custom configs, and they do tons of other things. This reduces their support costs, and creates a world were it is cheaper for businesses to go to these mass producers for machines and support.

Major companies like Xerox dont build their own machines... they outsource it.
October 2, 2006 11:10:04 PM

I would disagree that a steady current coming from the wall would decrease the chance of failure from any type of AC->DC converter mostly because the wall is a voltage source and since V=IR, a steady current would mean a drastically changing load in rythem to the voltage, but then again, that's me just being a smart @$$.

A UPS is not a battery, though it does use a battery. And batteries are VERY different from capacitors because capacitors physically store charge while batteries chemically store energy. Capacitors storing straight charge can deliver this energy very quickly, almost instantly, while batteries being chemical have maximum current ratings always associated with them. But all that is just a foot note UPSs anyway, because though they are basically car batteries with extra electronics, they most certainly don't deliver DC power, because if they did, my computer wouldn't be running right now because of the transformer in my PSU which requires a varying current to produce a changing magnetic field which will form power on the other coil of the transfomer.

Though I would agree with your father that it's very difficult to know exactly how a PSU works due to complexity of tracing out the circuit (voiding warrenties has never really bothered me), but you can still have a fairly good idea of how they are made and where they cut corners to make one cheaper then another. Though this isn't fool proof because people are clever and sometimes do ingenious things that cut costs while improving quality, so the more expensive == better isn't a hard fast rule, but it usually applies. As for sparking electronics, I actually had my video card catch on fire recently and my computer still functions fine with a new card, but I know I got lucky. It's just that failures like that are shorts on low voltage lines, failurs in a PSU are generally severe because it's dealing with peak voltages that can be quite high which if applied to any component will instantly smoke them.

On several final notes, an 80-130 dollar psu is what I'd actually consider spending personally, I think the 1000W beast that toms always uses in their benchmarks of SLI systems is WAY overkill for any desktop currently on the market, but if somebody wants it for peace of mind... Also, the problems a PSU and a UPS tackle aren't really related much, though it is true that you get fairly clean and consistant AC power out of a UPS, and I generally recommend them if you have a cheap PSU which might not be able to handle normal voltage changes in the wall power.
October 2, 2006 11:26:56 PM

Quote:


Its funny you should mention tires... Thoose $200 bridgestone tires on Ford SUV's... You know, the Top of the line ones! How many people did those tires kill?

Mike

How about none. The faulty tires on Ford SUV's were Firestone branded tires. After the problems Firestone had with Ford, they fell on hard times and were acquired by Bridgestone.
October 2, 2006 11:54:10 PM

Quote:


Its funny you should mention tires... Thoose $200 bridgestone tires on Ford SUV's... You know, the Top of the line ones! How many people did those tires kill?

Mike

How about none. The faulty tires on Ford SUV's were Firestone branded tires. After the problems Firestone had with Ford, they fell on hard times and were acquired by Bridgestone.
HAHA OWNT!


ANYWAYS...
Some of you people need to learn not to complain. Expensive psu's are not overrated, they have there reasons just like the people that purchase them. Here is the way I see it; just get one that allows you to sleep comfortably at night knowing that your PC is safe.


Here are some good high quality psu's that no one can argue over, I posted these in another post so I just copied and pasted.

Best Power supplies for the price

OCZ GameXStream 700watt $149 (4 12v rails @ 18amps ea., 3 year guarentee)
http://tinyurl.com/glnun

Nanopoint tagan 900watt $299 (4 12v rails @ 20amps ea. Dual transformer quad sli compatible, 3 year guarentee)
http://tinyurl.com/hhfs2

Nanopoint tagan 1100watt $259 (4 12v rails @ 20amps ea. Dual transformer quad sli compatible, 3 year guarentee)
http://tinyurl.com/zvkhp

Enermax 1000watt $349 (5 12v rails @ 17amps ea. quad sli compatible, quad core ready, 3 year guarentee)
http://tinyurl.com/jfv86
October 2, 2006 11:59:35 PM

Quote:
PSU may have caused the problem, but you cant rule out just plain bad luck... Especially when you said it was the optical drive... You should have had a second PSU available, so that you could swap them out, and then return the bad one for the RMA.

Monkey, Optical drives have horrible reliability... take it from me, I use mine heavily, and they have died on me so many times:(  I have 3 in my machine, and its not uncommon for me to send em back... infact I just had a NEC go on me, after less then a month of usage.

Ive lost 3 Plextors, 2 Lite-on's and now 1 NEC... Almost every one of them has been on a different powersupply and/or different machine.

I have had to RMA, RMA'ed Drives. I wish it was easy for us to point out a single cause and be absolutely sure that was the culpret, but there are SO MANY factors...

Like why didnt the HD fail... or the VC, or the CPU, Mobo Etc? Who knows...

I would however not recomend you sell/build machines for people who A) Live so far away from you... and B) If you need to visit their house to do the repairs... Next time, tell them to go buy a DELL. Have them deal with this s@#t


I personally wont build machines for companies. I tell them that DELL's are the best, because they support them. For me to recommend them to build machines, they would have to keep a stock of replacement parts to swap out in an attempt to find the problem. They would have to spend hours looking for the issue, and then they might have to re-install all the software... Go to windows update, etc...

Too much time...

Dell builds so many machines that they create procedures for repairing them. They make software installation CD's/DVD's that dont require all the custom configs, and they do tons of other things. This reduces their support costs, and creates a world were it is cheaper for businesses to go to these mass producers for machines and support.

Major companies like Xerox dont build their own machines... they outsource it.


Holy crap thats alot of bad drives !!! And I have burned probably thousands of DVD's and CD's..... I only buy them in bulk 100 packs as anything less would be a waste of my time. I have had one drive go bad on me and that was a Sony drive (of course I got like 4 years of use out of the darned thing lol) Hmmmm perhaps there is something to this cheaper PS after all. I myself usualy spend in the middle when it comes to PSU's but my last one I spent a little more and got a Antec neo HE, runs much cooler then my old one and in general seems like a better unit.

Another thing cheapo PSU's run hotter.
October 3, 2006 12:08:34 AM

I Disagree that cheap power supplies and expensive power supplies are about the same. Unlike you who have not had any bad things happen with cheap ones, I have. Took my whole motherboard out and my cpu,No longer will i buy a cheap PSU id ather spend the extra $100. The $100 is like insurace on a car true you can buy liablity issurance and hope u dont get in a wreck or buy full coverage and have it just replaced. What im trying to to say is that cheaper=problems that cost alot expensive=damage just to powersupply most of the time, not expensive compared to new board and cpu.
October 3, 2006 12:53:56 AM

It's worth keeping in mind, that despite the numerous PSU brands and price points, there are far fewer actual manufacturers. Many of the brand names people rave about are actually built by obscure Chinese factories that you've never heard of. Price doesn't necessarily equate to quality.
October 3, 2006 1:01:33 AM

Thanks for the info, some good posts in here.
At last a forum with a kill filter, like usenet with a gui.

The problem for people like me, and probably most computer buffs (who aren't EE's), is knowing what what we really need and identifying quality. All caps, diodes, transistor, etc look the same to me.
(I've had 3 mb's with leaky caps for instance).

When buying a PSU I rely on reviews from quality sites and let others do the hard work. They generally know more than me. The reviews do tend to focus on the more expensive, top of the line models though.

I wouldn't buy a psu without seeing it's specs. You may have to go to the manufacturers site to find them. It they don't have them, then I take that as saying that the specs are bad. Of course, the specs may be overinflated as I have often heard reading PSU reviews.

Perhaps a cheap supply from a respected manufacturer is the way to go, but the real bargains will be no name brands built by good manufacturers (since most no names are rebranded from what I have heard)

I think most people buy PSU's that are overpowered for their needs, but some good points about higher PSU's lasting longer. I may rethink my strategy: my highest PSU is 300W.
My gaming rig (sempron 2400, gforce 6600gt.... stop laughing) is running off a 250W Auriga PSU.

I would have thought a good quality 400W PSU would run anything, and I've yet to see any proof that it wont. Maybe if you run a P4 with SLI , a raid array, and play games while you copy dvd's, you need more, but I doubt it. It's not often that you use all your computers components at the same time.

As for power consumption, if you live in a colder climate like me, and run a heating system on a termastat, I figure that I run my pc's for free in winter cos all that power just goes to heat. It only costs me over the 4 months of summer during the day when the heaters aren't going. So in my case, take your power saving calculations and divide them by six.
My study is heated by my computers alone most of the time.
(I'm right here aren't I ?... or do I have to go back to physics 101 ?)

-------
The rest of this post is waffle, ignore it if you like....avoidance behaviour from real work.

I have 5 pc's, 3 which run 24/7 (wifes music pc, firewall, TV computer).

Noise is a big factor for me, so all the expensive power supplies (Q-sonic 300w, zalman 300w fanless, zalman300b-aps) have been bought with noise dominating my decisions. Except for my gaming PC, even the cpu choices have been made because of noise (35w duron 800's, 17W via M10000, next cpu will probably be a 35w athlon)

I've had a few power supplies fail on me over the years but no problems with them taking out other components. I had a hyena (dont know what factory these come out of) fail the other day. This came with a second hand case I bought. It was a new build of old bits (duron 800) and would reboot on startup. I thought I had bought a dud cpu. But then the PSU died, no fan, no nothing. I bought a second hand PSU for $7 and all is ok. I must have had 6 or 8 reboots with a failing PSU before total failure, and no component damage.

Only 1 expensive PSU failure, a q-sonic ultra quiet 300W PSU that cost me around $120 US, which didn't really fail but started to make a high pitched whine (not the fan).

damn, now I have to do real work.
October 3, 2006 1:18:17 AM

Quote:


Its funny you should mention tires... Thoose $200 bridgestone tires on Ford SUV's... You know, the Top of the line ones! How many people did those tires kill?

Mike

How about none. The faulty tires on Ford SUV's were Firestone branded tires. After the problems Firestone had with Ford, they fell on hard times and were acquired by Bridgestone.

Wrong, Firestone and Bridgestone Merged in 1990

Check the website:) 

http://www.firestone100.com/history/clip9.html

The faulty for tires were in 2000. Sorry, dont mess with a car guy:) 
The firestone tires used, were very expensive handmade tires... they where one of the more pricey large tread design. Just another example where money doesnt make quality... If I was a history buff, I could show you thousands of times where "The Best" became "The Worst" overnight...

3lfk1ng now with all that talk.. and that "Owned" comment was right on tract... though you ment it as if I was wrong... which I wasnt, as I proved... See I dont make statements unless I can prove them with fact... and when I do make statements that are wrong, I quickly appoligize and acknowledge my limitations as a human. As far as your "Best" list.... it now has 0 credibility with me.



Quote:

Holy crap thats alot of bad drives !!! And I have burned probably thousands of DVD's and CD's..... I only buy them in bulk 100 packs as anything less would be a waste of my time. I have had one drive go bad on me and that was a Sony drive (of course I got like 4 years of use out of the darned thing lol) Hmmmm perhaps there is something to this cheaper PS after all. I myself usualy spend in the middle when it comes to PSU's but my last one I spent a little more and got a Antec neo HE, runs much cooler then my old one and in general seems like a better unit.

Another thing cheapo PSU's run hotter.


The failures probably have to do with (A) bad luck... (B) I run 3 on top of each other and will burn 3 on 2 seperate ATA ports 1 SATA port and 2 SATA HD's... meaning they burn around 12-16x. If you do Burn as much as you say you do... you know how warm the DVD's are when they come out... Take and stack three drives on top of each other, and you see how the manufacturers design things... My thought is, if they cant take the heat, then they should send me a NEW one...

Last time I bought DVD's were in 1,000 Trays... I got 4,000 two months ago... I have 250 left...

Heat has to do with my failures, and if you read earlier... I had these drives fail on me with a $125 antec 450w powersupply. It was 1 year old.. My history with Plextor is unexplainable... I have had 3 go bad with 2-4 month usage time each... I have had my current one for over a year...

My Ultra is about 2 months old, and was bought for this new system.

The only drives to fail on me with this new one where 2 SATA Lite-ons and 1 NEC.... Turns out the SATA Lite-ons work VERY WELL... Just not on the old BIOS for the Asus P5B... and the lack of any documentation made me RMA too different Lite-Ons, when they both worked fine... So blame thoose on ASUS... The NEC, was iffy... and I think it has to do with the JMicron SCSI->PATA device... since the new Northbridges from intel dont have PATA.

The only reason I spend time on to respond to this is because it once again proves that there is more to a failure then simply blaming a single part. Its easy and relaxing when you can point your finger at something and say "THATS THE CAUSE OF ALL MY PROBLEMS!"

If the Powersupply did fail on this original case, then isnt the optical drive manufacturer as much to blame as the PSU company? Why wasnt the optical drive able to handle the surge, when every other peace of equipment went on harmed?
October 3, 2006 1:21:28 AM

Quote:


As for power consumption, if you live in a colder climate like me, and run a heating system on a termastat, I figure that I run my pc's for free in winter cos all that power just goes to heat. It only costs me over the 4 months of summer during the day when the heaters aren't going. So in my case, take your power saving calculations and divide them by six.
My study is heated by my computers alone most of the time.
(I'm right here aren't I ?... or do I have to go back to physics 101 ?)



The flip side to this, is that if you live in a warm climate and have to run airconditioners, you have to pay for the heat from the PC and the airconditioning costs.
!