J's new PC
TF2 looks totally different. It's like I'm playing a cartoon. The graphics card rocks. However...
During my build I did run into some annoying problems.
Gigabyte issues:
1 - Gigabyte needs to supply the correct amount of cables for there motherboards. They have 4 SATA ports but just 2 cables? Come on...really? 2 SATA cables and they give you 2 IDE cables. There is one one freakn' IDE port! Having to go out and buy extra cable was annoying.
2 - There custom back plate for the connectors, come on, they couldn't just use the standard. Literally 1 audio jack did not line up with the case back plate.
Video Card issues:
All through it is a great card it didn't mention that it needed a PCIe power connector. So I ended up having to buy a new power supply just because of that. Even in manual it does not tell you what power connector it is. Only when you open the it up and look at the card. FAIL. Ok yes I'm a NOOB and PCIe connectors are prob pretty common with high end cards, but proper documentation would have been nice.
Positives:
The case was really nice to work with.
The Intel processor came with a fan. I'm use to getting CPU's with no fan.
I can notice a real difference between my old PC and this one.
TF2 look REALLY sweet. I put all option to MAX setting and it runs great.
Never had a PC that could do that.
Only got 3GB RAM recognition, but I'm only using a 32bit OS. Once windows 7 come out (for a reasonable price) I'll be all set.
During my build I did run into some annoying problems.
Gigabyte issues:
1 - Gigabyte needs to supply the correct amount of cables for there motherboards. They have 4 SATA ports but just 2 cables? Come on...really? 2 SATA cables and they give you 2 IDE cables. There is one one freakn' IDE port! Having to go out and buy extra cable was annoying.
2 - There custom back plate for the connectors, come on, they couldn't just use the standard. Literally 1 audio jack did not line up with the case back plate.
Video Card issues:
All through it is a great card it didn't mention that it needed a PCIe power connector. So I ended up having to buy a new power supply just because of that. Even in manual it does not tell you what power connector it is. Only when you open the it up and look at the card. FAIL. Ok yes I'm a NOOB and PCIe connectors are prob pretty common with high end cards, but proper documentation would have been nice.
Positives:
The case was really nice to work with.
The Intel processor came with a fan. I'm use to getting CPU's with no fan.
I can notice a real difference between my old PC and this one.
TF2 look REALLY sweet. I put all option to MAX setting and it runs great.
Never had a PC that could do that.
Only got 3GB RAM recognition, but I'm only using a 32bit OS. Once windows 7 come out (for a reasonable price) I'll be all set.
0
Comments
//edit: There is no PC backplane standard. That's why they're interchangeable.
I love how you try to make it sound like I you caught me. :tongue2:
Had the case power supply came with a PCIe connector it would have been the same as a overprices name brand.
away with YOUR moonspeak
compare apple to apples in analogy please.
600W is 600W on any supply
I question the test method- no detail on what type of load it was.
Should we put this to a test. Get a name brand and cheapo and see how they compare.
Anyone want to donate a name brand for a few days.
I can show you half a dozen 600W units that will never, ever produce 600w. They'll die at 300 or less. In fact, Peter just linked 2 units that fall short of their rating because they were made with shitty components.
Your comments about load and heat are all well and good ceteris paribus, but it doesn't acknowledge the reality of PSU manufacturing.
Some are bad. Some are not. Not all are equal.
Anything else is just putting your head in the sand.
//EDIT: JohnnyGuru uses a SunMoon load tester on the 12v, the primary rail for today's systems.
It also should be noted that wattage doesn't mean much compared to amperage on the individual lines. You can have a Eleventybillion Watt PSU but it doesn't mean anything if the 12v line only supports 20 amps.
I looked at his tables and his math does not add up. He prob has no idea what he is doing
JohnnyGuru is the most respected reviewer of power supplies on the internet.
I have personally exploded PSU's with the unit, it is a lot of "fun".
The tester that Guru's site uses.. is equally as good and believe me, they do know what they are doing. That site is one of the deFacto sites on the net for accurate and unbiased PSU testing.
Sure.. I have a "name brand" PSU you can use, lets put it up against your no name.
I will collect the no name from you when I am in Warren in 2 weeks, then I will send it along with one of my OCZ's to PCPower facility, so they can be hooked up to the chroma.
Your 600W (or whatever it is) will fail at about 55 - 68% capacity, just a prediction.
um his result tables are provide by the manufacture?
Check out Results from Hairong ATX-480W COLD load tests
Test 1 he gets 198WDC. I add up 184.34WDC.
"is the most respected reviewer of power supplies on the internet" .
I forgot that if your on the internet that means good.
I read though his test methods. Not at all scientific, He let the machine do everything. Anybody can press buttons. What Waveforms, duty cycle, timing. ??? where are they. Inrush currents can screw up the testing as well, no mention on how he took care of that. He didn't test each rail on it's own. Come on if this hack is respected, I could start my own gig and rule.
Lets really do a test. A real test. not some hack pressing buttons on a tester. Be a great write up for Icrontic.
I would say:
1- rails have to be tested individually at first, then all together.
2- Have to measure the MOSFET temperature.
3- Constant and pulsed loads(varying duty cycles).
4- input power and voltage to be measure
5- peak to peak ripple voltages for all rails.
6- tested supplies either to have UL, CE or some other mark that tells it has been tested.
I would expect this.
The name brand will be able to hold the high power for longer periods of time, but both can provide the rated power.
Alright, now were talking. Send me the specs on your PSU and I'll go buy the cheapo equivalent. Oooo this is going to be fun
I was going for this
Sunbeam PSU-BKS580-US 580W ATX Power Supply - OEM
it's $20 for 580W. (newegg) Does anyone have ant objections...Thrax?
/me sits back a watches
If you are good with a 20 over or 30 under or whatever.. then we are good.
ok how about this one
COOLMAX V-500 500W ATX Power Supply
We also need to lay the ground rules of exactly what the challenge is.
Just to put out 500W for 30 sec? 1 min? If your going that short.. forget it.
Minimum 30min at max wattage or something.
I'll layout a test plan before we get started, but 30min at max load won't be in it. 30min will definitely see the thermal runaway and I already said better PSU's can handle this better.
I guess it's a good time restate my augument.
Argument:
For average users (like me) who don't push the limits of their PSU for long periods of time, one PSU is the same as any other. ie 500W OEM is the same as getting a 500W brand name.
Reasoning(Warning! moonspeak ahead):
When MOSFETS heat up their internal resistance (RDSon) goes up. As the resistance goes up the power dropped across the MOSFET goes up which results in more heat. Thermal run away occurs and the PSU will fail. For intermittent operation there is a chance for the MOSFETS to cool down and thus cheaper PSU's selection is not a big deal.
Ok there you have it. This is what I'm setting out to prove.
Counter points:
1- the cheapo power supplies can not provide rated power even for short durations. They will fail way before their rating.