PSU calculators

[avsquare]

I've recently changed from my single R9 290 to a pair of GTX 970 and adjusting stuff here and there.

Still using the entry level CX750M but I just got a freeze earlier.

I used this PSU calc: http://powersupplycalculator.net/

It says that the min recommended wattage is 613W considering all the devices and drives that I have (refer to the right side of this post, it shows what I have now).

But the calc is under the assumption of single monitor, which I have 3.

The CX750M runs at 85% power efficiency under load... so in actual fact I think I am running on thin borderline which explain the freezes right?

 

[Jian19]

freezes in ? i believe 750w is sufficient for ur setup

 

[kandinsky]

The CX750M runs at 85% power efficiency under load... so in actual fact I think I am running on thin borderline which explain the freezes right?

Efficiency rating doesn't mean it's 'running on thin borderline'. Your PSU should be able to deliver it's rated 750W output. If it is 85% efficient, my layman's understanding is that it means your PSU will need to draw 882W (750/0.85) to output 750W. Assuming your components are drawing the full 750W.

 

[Elijahonli]

dont be confused... your psu is able to output 750w max safely to your component regardless of the efficiency

 

[Maeda_Toshiie]

I've recently changed from my single R9 290 to a pair of GTX 970 and adjusting stuff here and there.

Still using the entry level CX750M but I just got a freeze earlier.

I used this PSU calc: http://powersupplycalculator.net/

It says that the min recommended wattage is 613W considering all the devices and drives that I have (refer to the right side of this post, it shows what I have now).

But the calc is under the assumption of single monitor, which I have 3.

The CX750M runs at 85% power efficiency under load... so in actual fact I think I am running on thin borderline which explain the freezes right?

750W maximum output (the efficiency issue is covered above by other posters) but consider this:

1. How much in total on the +12V rails? (IIRC a single rail design, CWT made PUQ-B). THIS IS VERY IMPORTANT!

2. The 750W is rated at what temperature? -10% rated wattage per increase of 10C.

3. Component aging, especially capacitors. Smaller than what most PSU calculators claim, but still something to consider in the face of Corsair's cap choices (see point 4).

4. It is a CX, with the cheaper caps among the Taiwanese caps (not the better Taiwanese OSTs) for the secondary side (primary cap may be Japanese).

-------------------------------------

Note the following:

1. Drives and fans consume the most power when spinning up, ie at system start up.

2. CPU/GPU do not necessarily hit their states TDP during typical usage, but power viruses can possibly drive GPU power consumption beyond the nominal TDP rating, which is what that is stated by the manufacturer (see point 4 as well).

3. OCing can drastically drive up power consumption. Power is directly proportional to frequency and to the square of voltage!

4. Note that Nvidia and AMD have different definitions of TDP. Same goes for AMD and Intel (too lazy to dig up the references on this).

5. Rated TDP != maximum power consumption at any one time.

 

[avsquare]

Hmm I haven't done any OC'ing to the GPU yet cos I am still monitoring stock temperatures overall; I've reapplied thermal paste to the CPU and made sure that the fan's heatsink isn't choked but the idle temp is still 45-50c and around 88c during gaming, which I think is a bit high on air cooling.

Tech specs of the PSU: http://www.corsair.com/en/cx-series...750-watt-80-plus-bronze-certified-modular-psu

It says continuous output rated temp: 30c (can't be this right lol)

And thanks for the explanation on the wattage efficiency, I've always thought the % is take it times the rated wattage lol, didn't know that the rated wattage is actually already at the rated efficiency %

I'll monitor further maybe since it only freezes once when I was churning photos on Lightroom.

Not sure if it could be just the CPU because recently I've got a few BSoDs with the error: CLOCK_WATCHDOG_TIMEOUT

 

[Jian19]

that pretty high, ideally if u plan to sli u should get really agressive cooler cards or ref cards with blower design which will exhaust heat air out from ur case

 

[Maeda_Toshiie]

Hmm I haven't done any OC'ing to the GPU yet cos I am still monitoring stock temperatures overall; I've reapplied thermal paste to the CPU and made sure that the fan's heatsink isn't choked but the idle temp is still 45-50c and around 88c during gaming, which I think is a bit high on air cooling.

Tech specs of the PSU: http://www.corsair.com/en/cx-series...750-watt-80-plus-bronze-certified-modular-psu

It says continuous output rated temp: 30c (can't be this right lol)

And thanks for the explanation on the wattage efficiency, I've always thought the % is take it times the rated wattage lol, didn't know that the rated wattage is actually already at the rated efficiency %

I'll monitor further maybe since it only freezes once when I was churning photos on Lightroom.

Not sure if it could be just the CPU because recently I've got a few BSoDs with the error: CLOCK_WATCHDOG_TIMEOUT

It's a budget unit, so rating at 30C is expected. High end units are usually rated at 50C, a typical operating temperature for PSUs (especially since this is the tropics, not the arctic region). That's why Jonnyguru's PSU reviews always feature PSU testing in both cold and hotbox conditions.

For reference: http://forums.anandtech.com/showthread.php?t=157636

Finally, the PSU efficiency certifications are done at 23C (go figure...). Jonnyguru makes it a point to test if PSUs can meet those efficiency ratings at hot conditions...

 

[avsquare]

that pretty high, ideally if u plan to sli u should get really agressive cooler cards or ref cards with blower design which will exhaust heat air out from ur case

I don't really think it is an issue of whether I SLI or not, even with the single 290 (Gigabyte Windforce) the temperature range is also the same, so I don't think the 970 SLI affected the CPU temp.

And my GPU always run at a temp lower than the CPU at any point in time. Even the bottom card of the SLI (which receives the hot air blasted by the top card) is still lower than the CPU man

This have been puzzling me for quite long.

It's a budget unit, so rating at 30C is expected. High end units are usually rated at 50C, a typical operating temperature for PSUs (especially since this is the tropics, not the arctic region). That's why Jonnyguru's PSU reviews always feature PSU testing in both cold and hotbox conditions.

For reference: http://forums.anandtech.com/showthread.php?t=157636

Finally, the PSU efficiency certifications are done at 23C (go figure...). Jonnyguru makes it a point to test if PSUs can meet those efficiency ratings at hot conditions...

I never knew this



In any case I still can't figure whether it is cos PSU can't handle load causing system to free or that CLOCK_WATCHDOG_TIMEOUT (cores of the CPU failed to sync) causing the freezes (and that shitty CPU temp too).