Hardware Insights Forum

http://hardwareinsights.com/change-o ... e-changes/

Let's discuss the article…

Cool. In future, I would suggest PM-ing Longrunner before publishing articles and asking him to proofread them. There were at least a dozen typos in that. I've fixed up the spelling mistakes this time. It's quite unprofessional to publish an article with mistakes and only getting someone to fix them afterward.

Sure, I've been quite tired so it was bad, but I wanted to publish it immediatelly so there was no time…

What are the differences between your testing equipment and the one used by Callum or other sites?

( I still haven't figured out why there was so much deviation on the review of Silverstone's ST60F-ESB 600, between your testing and Techpoerup's review http://www.techpowerup.com/reviews/Silv ... ST60F-ESB/ )

What do you mean, efficiency? Check this http://www.hardwareinsights.com/powe ... thodology/. I measure current with clamp meter. This si now much better, I got the new UNI-T UT210E instead of UT203 for some time already. I have compared them in first review with the new UT210E and there is up to almost 3 % difference. The UT210E now measures much more believable numbers which correspond with other reviewers. Also if you ploted a chart, the usuall course of efficiency depending on load woudl appear (whiel with UT203 it was often increasing till the very end).

Next step is hopefully some reasonable true wattmeter like chinese PF9811. These cheap wall meters I use (and Callum did) can measure strange numbers soemtimes and you cannot really believe them. Plans for much better loader are going for two years now…mybe I will build make it one day

All future reviews will be with UT210E. There will be some major changes soon, but in the end, I am fully switching for HWI with reviews in native english first, and than I will resell czech translations franchise-style, probably to some other site than DiiT.

I mean the large deviations about ripple measurements.
In Techpowerup's review the ripple gets out of limits near 100% load, while in Hardwareinsights's review, we see the ripple getting out of specs much sooner, near 500 watt load, and on multiple rails as well ( 5VSB, 3.3V, 5V )
What i want to say is how a customer can ever feel safe about a PSU's quality, if there are so large deviations between testing equipments ??
One thing is to have , let's say, 0.5mv deviation between reviews, but here we are talking about completely different ripple measurements !!

EDIT: Oh, by the way, since this is a thread about Hardwareinsights's ownership change, i would like to thank Callum, for all the work he has done in the PSU area!!
I hope for even better results under the new owner

Ripple measurement is a story itself…according to Intel's guidelines, you should even use a differential probe. Guess how many sites do? And also, you actually do not measure the ripple which is coming from the unit. You measure ripple on decoupling capacitors which simulate motherboard. So every site has somewhat different results depending on how exactly they implement Intel's guidelines, what decoupling caps they have, O-scope etc. TBH I am even glad I was not the only one who find this unti is out of spec so in the end, this is good result for me showing how close we are

Customer should get extremely cautious each time the ripple gets even just close to the limit. That means there is high probability it is not good. Especially when you have chinese units performing in 20 mV area on +12 V and sub-20 mV on other rails, taking anything which barely mades it through spec is not wise. (Cannot tell more, this is private measurement for potential import so I am under NDA. But if they opt for importing them, I will have final samples.)