Als ik zo gedacht had toen ik in 2014 de AX1200i kocht met een i7 3930K systeem en 780Ti dan had ik wellicht afgekund met een HX750, maar dat ding zou regelt warm lopen want ook dan was een piek van 700W wel haalbaar.
Natuurlijk weet ik dat veel van deze degelijke voedingen boven hun maximum capaciteit specced zijn, maar of je daarop wilt gokken voor de lange-termijn levensduur denk ik niet. De oude
HX750 HX850 voedingen waren ook getest in het verleden en daaruit bleek dat ze behoorlijk boven hun geadverteerde vermogens onder piekbelasting met 50°C konden leveren.
De layout is kapot van de nieuwe website van overclock3d.net is kapot dus hier betere links:
Voor de Corsair HX750:
https://web.archive.org/w...air_hx750w_750w_atx_psu/3Finally in test 7 the HX750w was pushed as far as it could go without OCP (Over Current Protection) kicking in and switching the unit off. This is the one test that could actually tell us if the HX750w is simply a re-badged HX850w, or if indeed there are some minor differences inside the unit that give it its 100w lower power output. With a whopping 21A load on both the +3.3v and +5v rails and 59A on the +12v rail, the HX750w topped out at a total of 913w. This is of course WELL above its rated output, but still a little below that of the HX850w which managed an amazing 1084w before showing signs of weakness. However, even at these extremely high loads, the HX750w managed 90.12% efficiency along with some rock solid voltage outputs.
Bumping up the ambient temperature to a toasty 50°C and re-running the room temperature results, hardy anything at all changed. In fact, aside from the odd +/-0.2v fluctuation on a couple of the rails and a few dips in efficiency, the only real thing worth talking about is the MAX load test which topped out at 887w - 26w lower than in the room temperature tests. This is by no means anything to get upset about as the HX750w is still operating at 137w above its rated capacity, and with 88% efficiency too.
Finishing up the testing, snapshots of the HX750w's +3.3, +5 and +12v rails were taken using a Rigol 25Mhz 400MSa/s oscilloscope while installed inside the 50°C hot box. Unfortunately there's not enough room on the page to put these side-by-side with the results from the HX850w, but if you're interested, the 850's results can be found here.
Once again the HX750w closely resembles that of the HX850w giving stunning results with the +3.3v and +5v rails staying at or below 10mV ripple in tests 1-6 and only hitting 12mV ripple in test 7 on the +5v rail. The +12v rail was also extremely clean with ripple starting at 12mV in test 1 and only increasing to 40mV by test 4. Test 7 saw the ripple hit 52mV, which is 10mV higher than that of the HX850w, but still way below the maximum of 120mV as specified by ATX standards.
Voor de Corsair HX850:
https://web.archive.org/w...air_hx850w_850w_atx_psu/4Finally in test 7 the HX850W was hooked up to both of our load testing machines to try and bring the unit to its knee's. Did I manage it?...did I hell! With 21A on both the +3.3v and +5v rails and 73A on the +12v rail the PSU was still laughing at me while delivering rock solid voltages. That's 1084W by the way, 234W more than what it says on the box!
In one last off-the-record attempt to get one over on the HX850W I hit it with everything it could take on the +12v rail. 91 AMPS was the read-out on the load tester before the PSU finally decided safely power off. Even then the voltage readout was 12.12v right up to the end. Corsair - 1, Me - 0!
Bumping up the ambient temperature to a toasty 50°C and re-running the room temperature results, hardy anything at all changed. In fact, aside from the odd +/-0.2v fluctuation on a couple of the rails and a few dips in efficiency, the results might as well be classed as identical. It really is beginning to look like nothing I can throw at the HX850W can phase it in any way, shape or form.
For the last part of the testing I analysed the ripple on the +3.3, +5 and +12v rails using a Rigol 25Mhz 400MSa/s oscilloscope. All readings were taken while the HX850W was installed inside the 50°C hot box to provide us with worst case scenario results.
Yet again the HX850W gave stunning results with the +3.3v and +5v rails staying below 10mV ripple in tests 1-6 and only hitting 12mV ripple in test 7. The +12v rail was also extremely clean with ripple starting at 10mV in test 1 and only increasing to 32mV by test 4. Test 7 saw the ripple hit 42mV, but this is still way below the maximum of 120mV as specified by ATX standards.
Voor de Corsair AX1200: (helaas geen ladende plaatjes)
https://web.archive.org/w...ax_1200w_atx_psu_review/5OK. Before we even talk about anything else, I think you should take a look at the DC Watts being pulled in test TMax1. Yes, you did read that right. That's a whopping 1522W that the AX1200 can output continuously. At one point I even got it up to 1600W, but unfortunately it turned off after a few minutes. This is also testament to the OCP functioning properly on the unit. No matter how much I overloaded it by, or how many times I switched it straight back on after it had cut out, the AX1200 WOULD NOT GO BANG!
The rest of the results speak for themselves to be honest. Good voltages (especially on +12v) throught all of the tests, no weird voltage see-saw results during the cross-load tests, no exhaust temperatures suitable for stripping paint from your walls and efficiency levels that would give Captain Planet a reason to retire.
Ripple.....what bloody ripple. I almost felt like that idiot who stands up on the stage saying "is this thing on" into the microphone. Tests 1-4 are flat, flat, flat and flat with no more than 24mV being recorded at any stage. The cross-load tests are also equally as uninteresting, and it's only when we get to the max-load test and seriously give the AX1200 a kicking that it puts out a still quite measly 42mV on the +12v rail. Awesome job Corsair!
En tja, hoewel de
HX750 de helft kostte, was die van mij na ~8 jaar (Q4 2010 - Q4 2018) defect door te grote afwijkingen buiten ATX specificatie waardoor de PC met i7 950 en 780Ti uitviel tijdens normale game-loads.
Verder is de
HX850 in die tijd ook niet enorm veel goedkoper, daarbij komt er nog bij dat de AX1200i toen de mogelijkheid gaf tot het uitlezen van alle voltages via Corsair link en zelfs het aanpassen van de rails naar wens of de fan snelheid aansturen, dat ding was ook volledig digitaal geregeld terwijl die oude voedingen nog analoog geregeld werden zonder uitleesmogelijkheden. Plus die HX750/HX850 waren volgens mij origineel niet passief qua fan.
De huidige
AX1200i gaat gewoon zijn geld wel opbrengen in levensduur en is de bedoeling dat hij straks weer dienst doet in de gedemonteerde i7 3930K met de huidige 1080Ti voor vele jaartjes, de voeding is nu inmiddels ruwweg 10 jaar oud en presteert nog net zo strak als fabrieksnieuw, totaal geen omkijken naar.
Inmiddels zijn er twee HX750i's gekomen die zowel de HX750 vervangt als een merkloze 580W in een oude Q9550 PC die wellicht niet eens verder komt dan 300W maar tegelijkertijd dient als back-up voeding omdat de Q9550 zijn levensduur aardig op zijn eind is qua compatibiliteit en snelheid in een modern OS.
Dus nee, in mijn geval is in één keer een super voeding kopen de juiste geweest.