eVGA SuperNova 1200 P2 Power supply hold up time concern

[Wolverine2349]

I just replaced a 1000W PSU in my system trying to just test luck and reduce RTX 3090 coil whine.

 

I wanted to do it ASAP and they only had this 1200W one in stock at local MicroCenter otherwise would ahve went with 1000W one.

 

Now I know these eVGA SuperNova P2 and T2 series are rebranded Super Flower Leadex Platinum PSUs which are very top teir and high qualityfor high performance systems so I went with that.

 

i installed it in my system already and yes it moderately reduced the coil whine on my video card and am very happy

 

However just found out hold up time fails ATX specification after looking furher at reviews:

 

https://www.techpowerup.com/review/evga-supernova-p2-1200/5.html

 

Hold up time is 14.4ms while ATX spec says it should be 16ms or higher

 

The twin of it the LeadEx SuperFlower Platinum 1200W is a little worse at 13.6ms. Seems takes these great SuperFlower PSUs and makes them even better as other specs are a little better as well like voltage regulation and transient response times and such:

 

https://www.techpowerup.com/review/super-flower-sf-1200f-14mp/5.html

 

 

 

Now is it something to worry about.

 

And a very important question that I have googled and have not been able to get an answer on. Is the hold up time specified meaning 14.4ms when under full 1200W load, or is that it regardless of the load?

 

Cause I will never stress this PSU even close to 1200W. Got it just for extra overhead to have quiet operation without high fan speed and great quality.

 

Checks all boxes until I saw hold up time failed.

 

So will holdup time be better under 600W to 700W loads and pass ATX spec and just fails at full load? Or is it the same on this PSU regardless of the load? I have googled this question and no answer found.

 

I just want to make sure I do not fry components if there is a power outage or brownout. I do have an APC Smart UPS 1000Watt which says it has transfer time of 6ms to battery during power failure which is faster than the hold up time of 14m on power supply (but is not true online UPS rather line interactive), so should that provide protection even if hold up time is below ATX spec for all loads? Or is there more to it than that that makes meeting ATX critical regardless of UPS transfer time.

 

Its odd that both the 1000W versions of the Super Flower Leadex Platinum and eVGA SuprNova 1000 Watt P2 pass the hodl up time test and seem to be same units. In fact the manual that came with my eVGA SuperNOVA 1200 P2 also lists 1000 P2 on there so it seems they are4 same maybe 1200W one is better binned and certified for 1200W? So that would get back to my original question on if holdup time would be much better at lower loads and just fails ATX spec at more than 1000W load?

 

 

[akio123008]

29 minutes ago, Wolverine2349 said:

So will holdup time be better under 600W to 700W loads and pass ATX spec and just fails at full load? Or is it the same on this PSU regardless of the load? I have googled this question and no answer found.

Absolutely. Less load means longer holdup time. I suspect under extremely light load it might even hold up for a full second. Hold up time is caused by the capacitors on the input of the power supply; bigger capacitors hold more charge resulting in longer holdup time. Less load will discharge the capacitors more slowly and therefore also results in longer holdup.

 

29 minutes ago, Wolverine2349 said:

I do have an APC Smart UPS 1000Watt which says it has transfer time of 6ms to battery during power failure (but is not true online UPS rather line interactive), so should that provide protection even if hold up time is below ATX spec for all loads?

Well, do the maths; holdup time is 14ms, UPS kicks in within 6ms: you're good.

 

But then regardless of all that, I wouldn't worry about holdup time at all. It's just the time delay between a power outage and the PC shutting off, and a PC shutting off doesn't kill any parts, so this happening a few ms sooner or later isn't really important. Just depends on how big of an input capacitor the manufacturer decided to put in there.

[Wolverine2349]

7 minutes ago, akio123008 said:

Absolutely. Less load means longer holdup time. I suspect under extremely light load it might even hold up for a full second.

 

Well, do the maths; holdup time is 14ms, UPS kicks in within 6ms: you're good.

 

But then regardless of all that, I wouldn't worry about holdup time at all. It's just the time delay between a power outage and the PC shutting off, and a PC shutting off doesn't kill any parts, so this happening a few ms sooner or later isn't really important. Just depends on how big of an input capacitor the manufacturer decided to put in there.

Thank you so much for your answer. That is exactly what I thought and wanted to hear. Just needed to be sure.

 

And yeah i did the math with APC UPS 6ms obviously faster than 14ms so it would transfer fine, but wanted to make sure I was not missing something that maybe being below ATX spec was bad anyways.

 

I wish the reviewers stated for hold up time test it was at 100% load like they did for transient voltages and ripple mv. Both TechPowerUp and Tomshardware have nearly same review format and it is very good. Just wish they stated hold up time at full load.

 

And what is reason they created the ATX spec of 16ms minimum for holdup time if it is not important and your system will just shut off quicker but not damage components? Is it for leeway for lesser UPS systems than mine that do not have a fast 6ms response so they can switch over?

[akio123008]

2 minutes ago, Wolverine2349 said:

I wish the reviewers stated for hold up time test it was at 100% load

They did put that in the text right?

[Wolverine2349]

1 minute ago, akio123008 said:

They did put that in the text right?

Maybe they did and if so I missed it.

[akio123008]

Just now, Wolverine2349 said:

Maybe they did and if so I missed it.

To be fair it was buried in the explanation text under "Hold-up Time":

[Juular]

2 hours ago, Wolverine2349 said:

Both TechPowerUp and Tomshardware have nearly same review format and it is very good. Just wish they stated hold up time at full load.

Because it's the same reviewer, @crmaris. For the record he (and his wife, among other people) also have two YT channels, Hardware Busters, one in greek and one in english, they don't only test PSUs there but sometimes PSU reviews make there before they're published at THW/TPU.

Yes, Aris tests hold-up at 100% load, at lower loads it would be proportionally longer.

[Spotty]

1 hour ago, Juular said:

For the record he (and his wife, among other people) also have two YT channels, Hardware Busters, one in greek and one in english,

Wait, they've split the English and Greek videos in to two channels? No wonder I've only been getting Greek videos in my subscription feed. I thought they were just focusing on making Greek videos.