|Posted on Sunday, July 9, 2000 - 7:22 pm: |
I work for a large computer company that installs 90 watt power supplies in some of their systems. As many articles that I have read, this doesn't even meet a basic configuration without overheating problems and other scenerios. Does anyone "really" know what a minimum requirement would be for a standard basic system? I think the company I work for is short changing the customer.
|Posted on Monday, July 10, 2000 - 4:03 pm: |
As a general rule, I think 250 watts is about the minimum I'd want to see on an AT system, and 300 on an ATX/Intel. It largely depends on the number of devices that are hooked to the case. I recently burned out a Gateway 145 watt P200 power supply when I hooked up an extra hard drive. It had a floppy, HDD, CD-Rom, modem, network card, video, and sound. Realistically, I can't imagine any useable configuration of a PC where 90 watts would be adequate. On the other hand, it'll create lots of high paying work for someone you when the customers start RMAing them back by the dozen...
|Danny Albers (Danny)
|Posted on Tuesday, July 11, 2000 - 12:56 pm: |
There is no predifined minimum, as long as your system provides the correct voltages.
Having said that, you can add up the wattage usage of each device, using peek power, and figure out how small a powersupply you need to keep up.
If you work for a large company, they have probrably done this. If they blow alot of powersupplies, then they are prolly using crappy ones.
Having said all of that, you want to always buy one slightly larger than your needs. If you needed a 75w, why would you use one? Having it run at full capacity all day is pointless when you can get a 250w for like 5 bucks more and have it run at a third capacity.
I have seen PC99 compliant systems, and ISO 9002 systems with as small as 100w powersupplies, yuck, but they meet the engineering specs.
|Posted on Tuesday, July 11, 2000 - 9:36 pm: |
Actually, chances are that a system integrator that's using 90 watt power supplies is counting on the customer not taking advantage of the warranty.
Lets say I design a power supply that will run at 90 watts continuously but will have a average life span of 12 months and then slap a one year warranty on it. I can sleep soundly knowing that the 50% that fail in warranty will probably do so close enough to the expiration of the warranty that I'll never have to honor it.
Yet I'll be "within specs".
Manufacturers write the specifications to suit their own needs, not ours.
Imagine an automobile engine whose idle speed was at redline. Or a bridge with a failure load that was one pound heavier than it's nominal live load.
Is the manufacturer in question breaking the law? Probably not. But if their power supplies are an example of what we can expect from the rest of their componentry, then I think short-changing might be too mild a description for what they're doing to their customers.