XPG Core Reactor 850 W

[crmaris]

The ADATA XPG Core Reactor 850 W is a fine power supply, offering high efficiency, good overall performance, and excellent build quality. The warranty is also long enough, and pricing is reasonable, but could be a bit lower.

Show full review

 

[mechtech]

I thought the EMC Pre-Compliance Testing - Average and Quasi-Peak EMI Detector Results was an interesting addition to the testing/review.

Was that testing done at idle or 50% or full load? Is it standardized testing to start at 150kHz? Can that device test all the way down to 50Hz? Was there a test done with the device off or unplugged to determine if there was a reference for background EMR?

Thanks

 

[cucker tarlson]

The ADATA XPG Core Reactor 850 W is a fine power supply

 

[crmaris]

I thought the EMC Pre-Compliance Testing - Average and Quasi-Peak EMI Detector Results was an interesting addition to the testing/review.

Was that testing done at idle or 50% or full load? Is it standardized testing to start at 150kHz? Can that device test all the way down to 50Hz? Was there a test done with the device off or unplugged to determine if there was a reference for background EMR?

Thanks

Actually I have those for quite some time now. I run them at a very light load for two reasons: this test can run for up to 2 hours and at light loads I noticed that the majority of PSUS featuring LLC resonant converters usually have the highest EMI spikes. So I apply a load of 28W. The most significant part is though, that I don't use a passive load (e.g. resistors) but an active load which simulates the response of a real system, meaning that it also generates EMI noise which the PSU's corresponding filter has to deal with, else it passes to the spectrum analyzer.

The standard dictates the 150KHz - 30MHz range, yes. The background noise is minimal, since I have a LISN and a 1:1 transformer behind the LISN and a dedicated AC line.

 

[mechtech]

Actually I have those for quite some time now. I run them at a very light load for two reasons: this test can run for up to 2 hours and at light loads I noticed that the majority of PSUS featuring LLC resonant converters usually have the highest EMI spikes. So I apply a load of 28W. The most significant part is though, that I don't use a passive load (e.g. resistors) but an active load which simulates the response of a real system, meaning that it also generates EMI noise which the PSU's corresponding filter has to deal with, else it passes to the spectrum analyzer.

The standard dictates the 150KHz - 30MHz range, yes. The background noise is minimal, since I have a LISN and a 1:1 transformer behind the LISN and a dedicated AC line.

Interesting, thanks for the feedback.

And I suppose the EMR limits are another standard where the standard depends on the country? How low and how high can the tool measure to?

Do you test in a pc case or an open bench.

Sorry for all the questions, just interesting stuff, even if for just becuase reasons and not practical ones.

 

[crmaris]

The EMC is defined by standards. You can see more here, in an older review: https://www.techpowerup.com/review/mistel-mx650/11.html

I only take conducted EMI readings and not radiated, which demand a proper and isolated environment.

 

[mrmrmr]

Thanks for the elaborate test!

Would it be possible to mention results of the 2% load and less at 230V? Even though the ATX 2.52 specification has requirements for loads at 10W @115V, half of the world uses 230V.

This is especially important for systems that remain on without suspend/hibernate and use less than 10W in idle. Many non-gaming Intel desktops use less than 10W in idle, combined with a Pentium Gold or i3-9100 processor and just a SSD, it will easily remain below 90W total DC power consumption. Even servers can go below 10W in idle.

Another question: which ATX power supply has the current best power consumption at 10W @230V?

 

[crmaris]

Hi! I don't get any 2% load measurements with 230V since this is not required by the ATX spec. Expect it to be notably lower with 230V, though.

As for the second question, the RM units are among the best in this perspective.

 

[jslade]

Aris, I noticed in the relative performance rating of this review that the Super Flower Leadex III 850W (non-RGB) scores higher than it did when you originally reviewed it in July 2019. Did Super Flower make some sort of silent revision to improve their load regulation, or is this due to sample variation? It appears to be around a 2% relative performance improvement which is significant! And if it's down to sample variation, then I wonder about their manufacturing consistency and quality control.

 

[Ronnychan]

Si las 2 tuevieran el mismo valor con cual de las 2 se quedarían y porque este xpg de 850 o la rx850 de corsair