Antec TruePower Quattro 850W Review

[burebista]

@Hardware Secrets.

The only problem with Antec TruePower Quattro 850 W was that voltages weren’t “clean” when delivering 850 W at 50º C: noise level was above the maximum admissible. That was the only problem we had with this unit. With temperature below that noise level was just fine.

Hmm, it seems that everything is fine but not at full load and 50ºC internal temperature.
A bit strange though. Gabe said

But after temperature increased to 50º C noise level started to increase after just one minute, jumping to 88 mV at the +12V1 input from our load tester (i.e. at the +12V2 rail from the power supply) and to 62.8 mV at the +5 V output, making the power supply to work out specs, as the maximum admissible noise level for this output is of 50 mV.

but jonny said

I must admit that I am absolutely amazed with the lack of ripple and noise on this power supply. In most cases, the ripple and noise measured on the +12V rail is actually less than that on the +3.3V or +5V!

Pretty much, the +12V doesn't even beg for attention until test 7, and even then we're talking about ripple and noise that is less than 10mV p-p. 10mV! At it's absolute worse, we only measure 20mV p-p on the +12V rail during tests 9 and10.

SKYMTL likes it too:

The ripple / noise suppression displayed by both of these power supplies is nothing short of astonishing. They are both among the best we have ever tested and for them to be performing like this speaks volumes about the high quality components used in both designs.

But Paul see something similar like Gabe:

The DC output quality is the one stage where the Quattro went from 4 stars to 1 star. While it was good throughout testing on most rails the 3.3v rail inexplicably skyrocketed during Test #3 and #4 with only a 5A load. What made this really worrisome is that the unit had 40mV of noise/ripple at 75% total load (5A 3.3v load) while at 100% load (5A 3.3v load) it had ~50mV.

So we have a tie here. What's the catch? Different samples, or test conditions?

[jonnyGURU]

Either they changed something in the unit or the different balance of loads resulted in different ripple/noise characteristics.

Not sure. Gabe is checlking out one of the ES-800's right now and he dropped me an email saying that the +3.3V and +5V ripple was way above spec. on the unit he was using. When I tested it, it was right within spec. Assuming that maybe he got a bad unit, I went ahead and pre-tested a second one and sent it off to him. Let's see what he finds out. I'd hate to think there was something weird in his test environment that was causing these results, but I know it's possible. Matt had something funky going on with one room of his place that was causing added noise to the line.

[Spectre]

Originally Posted by burebista

@Hardware Secrets.

Hmm, it seems that everything is fine but not at full load and 50ºC internal temperature.
A bit strange though. Gabe said

but jonny said

SKYMTL likes it too:

But Paul see something similar like Gabe:


So we have a tie here. What's the catch? Different samples, or test conditions?

I would vote both. My Quattro did run right smack into the spec limit. Had I run the unit hotter I may well have pushed it over the edge. Also, Gabe runs his in a lot less controlled manner as his temperatures fluctuate a good bit so the thermal cycling *may* have an affect as well.

SKYMTL's results are different becasue he is looking at it in a set system and not on the load tester so his loads change. Doesn't mean he is wrong or inaccurate just different results under different conditions. Hell his could be closer to what users will see while Gabe's and mine are closer to the worst case scenario.

And then there is the sample to sample variation. This can be huge! Boy the stories of those problems.

[burebista]

Originally Posted by Spectre

Hell his could be closer to what users will see while Gabe's and mine are closer to the worst case scenario.

And then there is the sample to sample variation. This can be huge!

Seems quite right. Worst case scenario vs. real world vs. sample variation.
Pfuui, looks like a lottery now.

[SKYMTL]

I think this shows us why different testing methods by reviewers and reading more than one review by readers is exactly what should be done. As Paul said: different load conditions means different results. That is why I ALWAYS read Paul's, Gabe's or Okla's review after finishing mine since I love seeing a worst case scenario versus the varying loads I put on the units.

Plus, it means I don't have to pony up the money for a load tester.

[Super Nade]

It should not be a lottery if QC is top notch. The only true lotteries I know of are Enermax. Topower does not count as they have failed to impress me consistently.

[Gabriel Torres]

Hello guys,

Just dropping a line to say that the noise level only increased after a while at 50º C. We usually push more on temperature than other websites. Our environment isn't as controlled as HardOCP's (we don't have a thermal chamber we use a er... "I made myself with a cardboard box" box).

Different methodologies can lead to different results. And that is so great to have several different talented people reviewing the same product with different methologies and opinions!

Anyway, in my opinion Cooler Master Real Power Pro 850 W achieved better results. But both products are great. This noise level over specification should not affect many users, but we should follow the same procedure for all reviewed units, right?

I am also facing a huge noise level problem with BFG ES-800 at its +3.3 V output. I am seeing this with Jonny before posting the review, to make sure that we made everything we can to solve this issue. Paul aka Spectre also saw this output reaching the limit (50 mV), but here it is far worse than that.

Cheers,
Gabriel.

[Travis]

Originally Posted by Gabriel Torres

Hello guys,

Just dropping a line to say that the noise level only increased after a while at 50º C. We usually push more on temperature than other websites. Our environment isn't as controlled as HardOCP's (we don't have a thermal chamber we use a er... "I made myself with a cardboard box" box).

Different methodologies can lead to different results. And that is so great to have several different talented people reviewing the same product with different methologies and opinions!

Anyway, in my opinion Cooler Master Real Power Pro 850 W achieved better results. But both products are great. This noise level over specification should not affect many users, but we should follow the same procedure for all reviewed units, right?

I am also facing a huge noise level problem with BFG ES-800 at its +3.3 V output. I am seeing this with Jonny before posting the review, to make sure that we made everything we can to solve this issue. Paul aka Spectre also saw this output reaching the limit (50 mV), but here it is far worse than that.

Cheers,
Gabriel.

Oh, maybe Mike's ripple readings higher than others are a consequence of high temp.
We may study this ripple issue.

[Gabriel Torres]

Who's Mike?

[Travis]

SPCR's Mike Chin.