Time for new Power Supply?

[b2cm]

NRD and Ozidave, very informative discussion. I'd also be all ears for Cluttermagnet and Nathan's opinion.

[longgone]

Is Ohm's law included here too ..........

[NRD]

NRD, I can see that you are passionate about this, but I think you are wrong.  Unfortunately, I don't know enough about electricity to contend on this subject.But by following the calculations at the link you provided, and using the 1.8 fudge factor, I should be running a power supply with 450 watts.

Using the 1.8 "fudge factor" as you called it, and using your stats as provided, the calculation comes to 300 watts. If you're going to debate, at least do me the service of proper calculation.1 PIII 866 chip= 38w2x128 PC800 Rambus memory= 16w815 dual processor Intel mobo=25w2 10K U160 SCSI drives=20w1 IDE CD/RW=10w1 IDE CD/R=10w1 floppy=5w1 video card=30w1 sound card=5w1 modem card=5w1 86mm fan=3wTotal = 167x1.8=300

Instead, my 4 year old Dell Precision 220 (a server level system) has a 230 watt power supply.  It also had a 3 year warranty.  I doubt that the people at Dell would put in an underpowered supply that would run close to the edge in a system that, back then, was intended to be sold to a business.  Dell may piss all over regular consumers, but they know that they better not do that to a business.

I should not even dignify that with a reply as it is beyond ridiculous. There is an admin of this forum who ran into the exact problem I speak of with another big name maker. They also decided to use a 200w power supply and it was not up to the task of a larger hard disk.

I don't get any weird errors, blue screens, etc.  The system runs nicely.  My REAL LIFE experience would seem to be in conflict with your theoretical posturings as to what will/might happen with such a low output power supply. I still believe that it is a rare person who needs a 400 or 500 watt power supply. What people do need is a GOOD power supply and IMO, a line conditioner or at least a good UPS (which cleans up the power but not as much as a separate conditioner).

These aren't my "theoritical posturings" It did not pull them from the ether as you seem to be doing. To base your total argument on your "personal computing habits" is flawed. I challenge you to place your 200w or even 300w power supply that runs "fine" in a system such as mine, which is not cutting edge by any standard. If it runs at all, I doubt it will run for very long.

[NRD]

It's not my maths, it's a fact, it's an equasion that is constant irrespective of the W/V.Wattage divided by Voltage gives you the Amp-Hour.

I was not reffering to W/V as I used the equation in my reply. I was reffering to the fact that you incorrectly used 500w/12v The full 500w is not available to 12v. In my 500w example, a Maximum of 16a is available to +12v period. Using your method, 42a should be available. Its not. The wattage is divided among the 12 3.3 & 5. as well as -12 -5 & +5 standby.The 500w power supply I listed can only supply a maximum of 16amps at full capacity BEFORE the CPU is attached. The Motherboard draws on +3.3 & 5v

I wasn't being specific! as I said further down "that leaves 19.38A".

No It doesn't. You are starting out with 16 amps maxium with the power supply I listed.

Your picture also says 16A @ 12v, 35A @ 5v, 20A @ 3.3. While the last two seem high

I didn't make the power supply, whether it seems high to you is not the issue, it is what it is. That is what the manufacturer listed as Max amperage for that unit.

What are you calling 'rails', the little skinny tracks on the circuit board the the RED (positive) wire is connected to?

Rails is an old electronics term that used to describe metal bars or strips that provide a particular voltage. Today, in relation to power supplies, it just refers to various voltage lines that power a PC. +12v +3.3v etc etc ;)Wheww..This quoting replying is tiring stuff.

[Agent007]

I'm wondering if this is really related to the PSU. Why is the problem occuring only in Windows? If the PSU was at fault, wouldnt it be obvious at boot time?

[GolfProRM]

I'm wondering if this is really related to the PSU. Why is the problem occuring only in Windows? If the PSU was at fault, wouldnt it be obvious at boot time?

It only happened once, and hasn't done it since (a couple days), so I'm not as concerned about it as I was previously...I'm still going to buy a new PS here in a week or so, but I'm not in as much of a hurry as everything still looks good (all my voltages are normal)

[ibe98765]

Using the 1.8 "fudge factor" as you called it, and using your stats as provided, the calculation comes to 300 watts. If you're going to debate, at least do me the service of proper calculation.

I used the calculator at this URL (which was provided by Longgone in post #2 in this thread) - http://www.jscustompcs.com/power_supply/Below, in parenthesis, I've put the wattage's from the calculator next to your quoted values. It comes up with 227 vs. your 167. Adding in your 1.8 fudge factor gives 409.1 PIII 866 chip= 38w (35)2x128 PC800 Rambus memory= 16w (20)815 dual processor Intel mobo=25w (25)2 10K U160 SCSI drives=20w (25x2=50)1 IDE CD/RW=10w (20)1 IDE CD/R=10w (20)1 floppy=5w (5)1 video card=30w (35)1 sound card=5w (7)1 modem card=5w (4)1 86mm fan=3w (3)Total = 167x1.8=300Calculator total (227 X 1.8 = 409)I realize these wattage's are approximations, but methinks a little less "attitude" would be appropriate!

I should not even dignify that with a reply as it is beyond ridiculous. There is an admin of this forum who ran into the exact problem I speak of with another big name maker. They also decided to use a 200w power supply and it was not up to the task of a larger hard disk.

And this proves something? It should be clear to most people by now that you can't compare different PSU's at different wattage's with different systems.

These aren't my "theoritical posturings" It did not pull them from the ether as you seem to be doing. To base your total argument on your "personal computing habits" is flawed.

So your statements are an ABSOLUTE then? There are no exceptions to your rules, right? And so therefore, my real experience must be flawed? Hmmm... I guess I am that one in a million person then. Were I only so lucky on my lottery tickets.

I challenge you to place your 200w or even 300w power supply that runs "fine" in a system such as mine, which is not cutting edge by any standard. If it runs at all, I doubt it will run for very long.

This is a pure assumptive statement with no evidence to back it up.

[NRD]

but methinks a little less "attitude" would be appropriate! dry.gif

Perhaps you would do well to take your own advice where "attitude" is concerned. Comments like "fudge factor" & "theoretical posturings" is not going to get an endearing response. I'm starting to realize you are now arguing for arguments sake. By your own admission, you don't have the knowledge to make a clear argument, which I'm seeing evidence of.

This is a pure assumptive statement with no evidence to back it up

Kind of like these?"I doubt that the people at Dell would put in an underpowered supply that would run close to the edge in a system that, back then, was intended to be sold to a business. Dell may piss all over regular consumers, but they know that they better not do that to a business""I still believe that it is a rare person who needs a 400 or 500 watt power supply"" And just because a power supply is 500 watts, doesn't mean it is better than a 350 watt one""I don't know for sure, but I'd guess that the bigger the power supply (in terms of wattage), the more power it consumes and the more heat it generates""My max power requirements are 17 watts MORE than the max my power supply provides. And I don't have any problems"

[Ozidave]

Wheww..This quoting replying is tiring stuff.

Hi NRD, I'm all worn out thinking about it. And we haven't even discussed what happpens to AC voltage and current when it's converted to DC. Is there any loss in current and to what degree. If it's the AC tranny that is rated at 500w, and not the DC output, could it be that the current is reduced by about 50% when converting to DC? That would make it about 20A DC, and fit in with your calculations.

[NRD]

Hi NRD,  I'm all worn out thinking about it.    And we haven't even discussed what happpens to AC voltage and current when it's converted to DC. Is there any loss in current and to what degree.  If it's the AC tranny that is rated at 500w, and not the DC output, could it be that the current is reduced by about 50% when converting to DC?

Hi OZI,That 50% conversion loss you're talking about is applicable to linear power supply's. An example is one of those AC adapters used for answering machines. Around 50% is lost in the form of heat. PC power supplies are switching power supplies. Very little is lost in the conversion, the trade off is the interference they produce, which is why they come in metal cases.

[NRD]

Ozi, I think the problem is that you are relying on that figure on the side of the power supply. It may say 500 watts but what it can deliver in usable power is usually much less. The manufacturers are fudging the numbers a bit.A good analogy is PC monitors or Horsepower. It may say "17 inch monitor" but the viewable is usually about an inch or more less. Same thing with horsepower that many auto manufacturers use. 255 hp is what the engine can produce, but that is not what you are getting at the wheels"Same thing with power supplies. If you count up the total watts it will come to 500 but what is actually available for use is usually much less. The don't use the +3.3 V / +5 V combination limit, they count each separately. Thats why its so important to check the amp ratings for each rail. You can judge what you are actually getting. For a modern system, I personally would not touch a PS that can't deliver at least 20a to +12v no matter what the watt rating.

[b2cm]

1 PIII 866 chip= 38w (35)

The P3 866EB could only be 22-23w. NRD's Barton processor would be consuming 68 to 69w.

[NRD]

I got the Pentium III Processor 38W figure from the chart at extremeoverclocking. I used the the same chart for all other calculations, using the lowest watt rating listed.They list an AMD athlon processor & P4 at 70w also.

[longgone]

My turn..... horsepower.... (I know a little about it). Using a sem truck/tractor as an example... a two axle semi with a motor rated at say 600HP is actually going to deliver to the rear axle 85 pct of that rated horsepower. If it is a three axle semi it will deliver 75 pct. The adverstised horsepower referred to as BHP (brake horsepower) is actually measured at the flywheel before it starts to lose horsepower due to the friction of all the components that come after the flywheel. okay back to the power supply now ..

[NRD]

  My turn.....  horsepower.... (I know a little about it). Using a sem truck/tractor as an example... a two axle semi with a motor rated at say 600HP is actually going to deliver to the rear axle 85 pct of that  rated horsepower. If it is a three axle semi it will deliver 75 pct. The adverstised horsepower referred to as BHP (brake horsepower) is actually measured at the flywheel before it starts to lose horsepower due to the friction of all the components that come after the flywheel.    okay back to the power supply now ..   

That was my point exactly. People generally don't use the engine without the car attached. Its misleading. My car is rated 255bhp.... to the wheels is more like 220. (Friend had his dyno'd)

[Ozidave]

Does anyone want to know how much a 1400w x 240v Air Conditioner pulls at start-up. It's constant rating is about 5.6A and because the motors are inductive they require more current to start (Fan and Compressor).If you multiply 5.6A x 1.8 you get 10A to start, 5.6A to run. Close enough I think and because the compressor cuts in and out, its overall rating must be 10A.CD drives are much the same. So depending on what you do and what you have installed, is what dictates the amount of power you require.The same can be said for most of the peripherals including the sound-card.. the louder the music... the more current required. :hmm:Then there's operating-temperature to consider also. (and on it goes). So Ibe, if you or anyone increase their PS, they'll avoid half of the problems that people write about here.

[NRD]

Be careful Ozi, or IBE's gonna be pissed at you too Seriously though, that last statement of yours couldn't be more true. I read an article that stated over 30% of PC problems are power related and usually one of the first question asked at a Mobo manufacturers support forum is "What are your rail voltages"I doubt its because they have some weird electrical fetish.. its because they have seen the same problems over and over and a common culprit is undervolted components.

[b2cm]

I mentioned the variance in CPU wattage as an explanation for Ibe's system power requirements. With an integrated microATX board (especially with onboard video) the requirements would probably still be lower (than a full ATX with expansion cards)? In which case, a 300w PSU would suffice?

... one of the first question asked ....

For anyone who fixes PCs, too, at even a faint hint that the problem may be hardware-related.

[ibe98765]

So Ibe, if you or anyone increase their PS, they'll avoid half of the problems that people write about here.

But my real life experience with this and other similarly "underpowered" systems says that this isn't necessarily true. And I'd bet that others here have the same experience. Not everyone is running a 300/400/500 watt PSU and not everyone running the smaller PSU's are having problems.Therefore, there must be another more reasonable explanation. I believe that explanation comes down to the QUALITY of the power delivered and the QUALITY of the PSU, not the SIZE of the PSU (in wattage, amperes or whatever).

[ibe98765]

Perhaps you would do well to take your own advice where "attitude" is concerned. Comments like "fudge factor" & "theoretical posturing's" is not going to get an endearing response.

I don't see what you are having a problem with. Your 1.8 factor is a true "fudge factor", isn't it? There is no science you can quote to back up this figure, is there? And while you have written a lot, you not provided any independt articles or cites to backup your contentions. Which is why I [nicely] call what you write "theoretical posturing's".

Kind of like these?"I doubt that the people at Dell would put in an underpowered supply that would run close to the edge in a system that, back then, was intended to be sold to a business.  Dell may piss all over regular consumers, but they know that they better not do that to a business""I still believe that it is a rare person who needs a 400 or 500 watt power supply"" And just because a power supply is 500 watts, doesn't mean it is better than a 350 watt one""I don't know for sure, but I'd guess that the bigger the power supply (in terms of wattage), the more power it consumes and the more heat it generates""My max power requirements are 17 watts MORE than the max my power supply provides. And I don't have any problems"

Nope. The difference is that all of these make sense in context. What you said and that I specifically referred to, didn't...

[NRD]

..Therefore, there must be another more reasonable explanation.

Yeah IBE, How about Magic? That fits in as well as your other theories.

I believe that explanation comes down to the QUALITY of the power delivered and the QUALITY of the PSU, not the SIZE of the PSU (in wattage, amperes or whatever)

"Or Whatever?" I'm unfamiliar with that technical term. To make an assertion that the size of the PS in terms of watts or amps has little or nothing to do with system stabilty furthers my contention that you don't have even a basic understanding of the topic.

[NRD]

I don't see what you are having a problem with. Your 1.8 factor is a true "fudge factor", isn't it? There is no science you can quote to back up this figure, is there? And while you have written a lot, you not provided any independt articles or cites to backup your contentions. Which is why I [nicely] call what you write "theoretical posturing's".

Did you miss this link? http://www.extremeoverclocking.com/article...ly_Guide_2.htmlI guess they are wrong too. I think I've discovered the problem, You didn't actually read the posts, you just scanned and got miffed that someone disagreed with you.As an aside, I apologize to those reading for what this has turned into, but I now realize I can't argue with someone who already knows everything.

[GolfProRM]

I'm sorry such an innocent question turned into a giant egofest... Please remember to have more respect for others... This type of back-and-forth bickering isn't tolerated here. All of you know better and none of you could put your ego aside for a minute to answer a straightforward question. I do appreciate those who tried to help, and everyone who was respectful.Topic Closed.