FSP Everest 1010w Power Supply @ [H]

phalanx said:

I guess I'm still not grasping why the unit is considered a failure when it was knowingly run outside of its specification. Derating curves are a fact of life with all power electronics and need to be accounted for. I'm actually surprised the manufacturer even supplied the curve. The 100% load test at 45C should have been at its derated output of about 840 Watts.

Click to expand...

http://www.hardocp.com/article.html?art=MTI4OSwsLGhlbnRodXNpYXN0

Yes derating curves happen but if the unit is supposed to be an 840 power supply at a relevant operating temperature maybe it should be sold as an 840w power supply not a 1010w power supply because you know what the end user doesn't know that the unit is only a 840w power supply at a relevant operating temperatures. I only know it is because I can test the units properly and I have access to FSP's old documentation. Only when I presented that did FSP tell me they knew that the unit could not do its marketed output at anything above 25c and that this was lower temperature than they should be working with.

Now as I said earlier the days this unit was being tested there was NO WAY to test the unit at temperatures lower than ~43.8c because that was my office temperature, and no I was not going to move a refrigerator into my office to nurse a power supply through testing that would fail if it was in my office PC.

Also, if the FSP gets to play by a MUCH lower standard in order to achieve somes set of goals why shouldn't everybody else? And if someone else can do the higher standard but FSP can't that would make the other product better, no? Are we not here to review products to find which are the best and which are not?

If you'd like I could just open each box and go "Ooohhh pretty, shiny, Editors Triple Platinum Award!!!!!!!!!!!!!!!!!1" like the majority of reviews do.

Paul_Johnson said:

http://www.hardocp.com/article.html?art=MTI4OSwsLGhlbnRodXNpYXN0

Only when I presented that did FSP tell me they knew that the unit could not do its marketed output at anything above 25c and that this was lower temperature than they should be working with.

Click to expand...

Did the supply come with temperature derating information in its retail packaging or does this mean that you only had access to the derating curves after sending feedback to FCP about the supply failing?

If it's the latter, then I agree with the failure.

phalanx said:

Did the supply come with temperature derating information in its retail packaging or does this mean that you only had access to the derating curves after sending feedback to FCP about the supply failing?

If it's the latter, then I agree with the failure.

Click to expand...

Like I said:

Yes derating curves happen but if the unit is supposed to be an 840 power supply at a relevant operating temperature maybe it should be sold as an 840w power supply not a 1010w power supply because you know what the end user doesn't know that the unit is only a 840w power supply at a relevant operating temperatures. I only know it is because I can test the units properly and I have access to FSP's old documentation. Only when I presented that did FSP tell me they knew that the unit could not do its marketed output at anything above 25c and that this was lower temperature than they should be working with.

Click to expand...

I knew beforehand what the derating curve was but only because I have prior documentation and I have failed this same design before over a year ago.

http://www.hardocp.com/article.html?art=MTM0NSwsLGhlbnRodXNpYXN0

Providing the derating curve in the documentation would not have kept me from testing this unit at 45c as previous units that have provided this information still got tested at 45c since that is what ALL units get tested at no matter their published or unpublished derating curve.

For instance I have the original documentation on the 700w Epsilons and they share the same derating curve yet:

http://www.hardocp.com/article.html?art=MTUxNywsLGhlbnRodXNpYXN0
http://www.hardocp.com/article.html?art=MTUwMywsLGhlbnRodXNpYXN0

The Andyson built Ultra's are rated at 40c but this is not published to users and so see what happens:
http://www.hardocp.com/article.html?art=MTQzNywsLGhlbnRodXNpYXN0
http://www.hardocp.com/article.html?art=MTM4MywsLGhlbnRodXNpYXN0
http://www.hardocp.com/article.html?art=MTMxMywsLGhlbnRodXNpYXN0
http://www.hardocp.com/article.html?art=MTMxNywsLGhlbnRodXNpYXN0

The Wintech built Ultra here is rated at 25c but isn't published:
http://www.hardocp.com/article.html?art=MTQ2MCwsLGhlbnRodXNpYXN0

As are these Seasonics but this information is published:
http://www.hardocp.com/article.html?art=MTM2OSwsLGhlbnRodXNpYXN0
http://www.hardocp.com/article.html?art=MTM2NCwsLGhlbnRodXNpYXN0

But these Win-Tact's are rated at 50c which is also provided:
http://www.hardocp.com/article.html?art=MTM1OCwsLGhlbnRodXNpYXN0
http://www.hardocp.com/article.html?art=MTQ2MywsLGhlbnRodXNpYXN0

And yes there are more units rated at 40c and 50c than those I just pulled them for examples.

So could you make your PSU provide more power than its rated for with extreme cooling? How well would a unit perform if attached it to phase change, or some other type of sub-zero cooling?

bholstege said:

So could you make your PSU provide more power than its rated for with extreme cooling? How well would a unit perform if attached it to phase change, or some other type of sub-zero cooling?

Click to expand...

The cooler a unit runs the more power it could put out in theory. In reality it depends on the components used, the circuit design, etc and beyond a set design output the DC output quality generally suffers. In reality phase changing your PSU when it was not designed to work at those temperatures would most likely net you no benefit whatsoever.

bholstege said:

So could you make your PSU provide more power than its rated for with extreme cooling? How well would a unit perform if attached it to phase change, or some other type of sub-zero cooling?

Click to expand...

In this case it would be akin to buying a $10000000 saddle for a $1 horse.

phalanx said:

I guess I'm still not grasping why the unit is considered a failure when it was knowingly run outside of its specification. Derating curves are a fact of life with all power electronics and need to be accounted for. I'm actually surprised the manufacturer even supplied the curve. The 100% load test at 45C should have been at its derated output of about 840 Watts.

Click to expand...

For the same reason that Intel doesn't advertise that their Core 2 Duo E8600 can run at 6.25GHz with the help of LN2. Its not a realistic specification, just like 25C/~77F is not a realistic environment. Its what I refer to as an optimal environment.

If a power supply can't perform unless its in a meat locker of a room, then it really doesn't do the person buying it any good, to be honest, unless said person lives in said meat locker. Out of manufacturer specification or not, the damn thing failed in a real world environment test where others in its class have excelled. If that isn't reason enough for it to be considered a failure, I don't know what more to tell you.

Am I the only one who can't find the link to the article in the first post?