Power Supply Seasonic Focus SPX Series 750 W SFX PSU Review

[riba2233]

So, how did we miss this one, I was wondering when will Seasonic release true SFX psu:

Seasonic Focus SPX Series 750 W SFX PSU Review

The Seasonic Focus SPX-750 offers good overall performance, but is struggling to compete with the Corsair SF750 Platinum in most areas. Its fan-speed profile and over-temperature protection also require further tuning.

www.techpowerup.com

Too bad it is more expensive than corsair, but I guess price will drop a bit over time.

 

[prayogahs]

note that in Aris's testing, 2 of the samples died when testing for OTP.

 

[riba2233]

note that in Aris's testing, 2 of the samples died when testing for OTP.

Lol, that is not good but I hope it was just a fluke. I generally respect seasonic the most out of all psu manufacturers

 

[Overtek]

hmm, that's interesting. I'll see what the 'vanilla' non consumer version comes in at tomorrow.

OTP? Who's Aris?, Is it a recent test as I think the Platinum version is newish.

 

[prayogahs]

hmm, that's interesting. I'll see what the 'vanilla' non consumer version comes in at tomorrow.

OTP? Who's Aris?, Is it a recent test as I think the Platinum version is newish.

OTP is Over Temperature Protection. Aris is the reviewer of the above article, which is one of the most well known PSU reviewer and also runs the Cybenetics certification and Hardware Busters YouTube channel. He did mention in the conclusion that he cannot recommend this PSU because two samples died when testing for OTP.

Just for context, he is also the first one who found out the issue about Gigabyte P750GM exploding which Gamers Nexus followed up with.

 

[REVOCCASES]

just my 2c concerning the "OTP issue" but the way this was tested doesn't really meet any industry standards and isn't even close to a real world scenario where OTP is supposed to trigger (e.g. the fan fails / is blocked).



usually - at least for industrial SMPS - this is tested in a controlled environment / testing chamber and not by heating up some components locally with a heat gun.

 

[riba2233]

just my 2c concerning the "OTP issue" but the way this was tested doesn't really meet any industry standards and isn't even close to a real world scenario where OTP is supposed to trigger (e.g. the fan fails / is blocked).

View attachment 1693

usually - at least for industrial SMPS - this is tested in a controlled environment / testing chamber and not by heating up some components locally with a heat gun.

Agreed, this seems a bit extreme

 

[Overtek]

heat gun, If it's like mine it chucks out temps that are beyond most component spec limits ie 400degrees higher. I stick to using that for cooking fingers whilst holding wire with heat shrink. Prefer to just test on an over power protection basis, over loading in ambient till it shuts down.

Disappointed not to see a price this morning for the vanilla one , but will enquire to see if its going to be about in future or not.

 

[thelaughingman]

just my 2c concerning the "OTP issue" but the way this was tested doesn't really meet any industry standards and isn't even close to a real world scenario where OTP is supposed to trigger (e.g. the fan fails / is blocked).

View attachment 1693

usually - at least for industrial SMPS - this is tested in a controlled environment / testing chamber and not by heating up some components locally with a heat gun.

Understand the concern about testing methodology but it could still be valid as long as it is applied consistently in all of his reviews.

If assuming the heat gun method is applied consistently in all of his reviews, and that other PSUs did "pass" and only SPX750 "fail", then such a fail is still significant. Just means it requires more deep-dive / further validation.

 

[REVOCCASES]

Understand the concern about testing methodology but it could still be valid as long as it is applied consistently in all of his reviews.

If assuming the heat gun method is applied consistently in all of his reviews, and that other PSUs did "pass" and only SPX750 "fail", then such a fail is still significant. Just means it requires more deep-dive / further validation.

well, I'm not in the position to say if there really is an issue with the OTP or not - I haven't tested the PSU myself and he did not provide further details / photos how he did this test exactly (e.g what was the temperature settings of the heat gun? at which components did he point it exactly and how long? how can he control the conditions consistently over all of his tests? etc...). All I can say is that I find this test methodology questionable from an engineers point of view. There are industry standards in place how power supplies and their safety functions are tested and they do not involve a heat gun. I showed this to a friend who worked in R&D for EMERSON/ARTESYN and he couldn't help stop laughing. The problem is, if you only apply heat locally you may risk that some components are already overheating and getting damaged without the OTP circuit noticing it early enough. I mean he obviously has a "hot box" for thermal testing, so why not using it for the OTP test?

Overall I found his review quite helpful and well written but IMHO he should ditch the heat gun with a more professional approach. Reviewers have a great influence and help to point out issues related to the manufacturers but they also do make mistakes sometimes and should adapt / improve when necessary.

 

[tinyitx]

well, I'm not in the position to say if there really is an issue with the OTP or not - I haven't tested the PSU myself and he did not provide further details / photos how he did this test exactly (e.g what was the temperature settings of the heat gun? at which components did he point it exactly and how long? how can he control the conditions consistently over all of his tests? etc...). All I can say is that I find this test methodology questionable from an engineers point of view. There are industry standards in place how power supplies and their safety functions are tested and they do not involve a heat gun. I showed this to a friend who worked in R&D for EMERSON/ARTESYN and he couldn't help stop laughing. The problem is, if you only apply heat locally you may risk that some components are already overheating and getting damaged without the OTP circuit noticing it early enough. I mean he obviously has a "hot box" for thermal testing, so why not using it for the OTP test?

Overall I found his review quite helpful and well written but IMHO he should ditch the heat gun with a more professional approach. Reviewers have a great influence and help to point out issues related to the manufacturers but they also do make mistakes sometimes and should adapt / improve when necessary.

I myself am an engineer and agree that industry testing standards should be used and, of course, if a reviewer wishes to supplement with his own methods, he might choose so.

On one hand, I also agree that consistency over all reviews is important to make comparison meaningful. On the other hand, I also think the reviewer's own testing method should be reasonable enough to yield practically useful result. This is probably even more important.

Say, a reviewer tests the enclosure integrity of a bunch of PSUs by dropping them from a 20-storey building. 9 PSUs survive and 1 does not. How useful is this result? Very little. So, what kind of real world setting is being simulated/replicated by using a heat gun? Maybe, if an end user has his system on a test bench and somehow he puts his system right next to his room heater?