Power Supply Testing Method Update - Protections

AUTHOR:Lucas N.
Published:

What is Our Problem?


In the recent LTT video discussing our power supply testing methodology we shared a relatively dramatic failure rate of 23%. This isn't the rate of power supplies that have failed to meet an efficiency tier, it is the rate of power supplies where two samples have fully stopped functioning during testing.

While our tests don't involve any outrageous abuse and I consider them to be quite fair, they are certainly more onerous - creating a larger testing load/burden - than typical use in a gaming rig or workstation. There are pros to more onerous testing and strict requirements like finding only the best of the best and ensuring that anything we "recommend" is essentially bulletproof. However, our primary goal is to provide useful information to the audience, and sometimes that is best done with testing that is more representative of typical use. There is a balance to be struck.

While the "deaths" of the power supplies are relatively distributed across the tests that we run, there are a significant number that fail during the testing of the protections, specifically Over Current Protection(OCP) and Over Power Protection(OPP). These are incredibly important features to test, but most users won't be operating their power supplies above the maximum rated load. A failure of the power supply to gracefully recover from a shutdown may not be as important to them. In this case, it doesn't feel right for a failure during the OCP/OPP tests to stop us from getting test results that might matter more to certain people. Many of the "failing" power supplies could be used for years without issue; an inability to push the limits may not be the most important factor to someone trying to stay within a budget.

Of course, over 75% of power supplies we've tested have had no issue with the protections testing and that is valuable information if you're looking for our top recommendations. We'll continue to conduct all of the tests and find out which power supplies should be able to withstand whatever weird things you have planned.

Moving forward, our plan is to treat survival and recovery of protections testing as a separate part of the results.

Protections Testing as a Separate Objective


We will conduct all of the same tests as a part of our gauntlet, except that the protections testing will be moved to the end of the sequence. Assuming that the power supply doesn't fail during standard[1] testing(which would earn a "Fail"), we can collect almost all of the test data[2], and then do the more onerous protections testing afterward. The four results will be defined as below:

Pass: Passed all of the standard[1] testing, as well as the protections testing, without failure.

Protections Fail: Passed all of the standard[1] testing, but died during the protections testing.

Partial Fail: Failed during the standard[1] testing, a second unit was tested, but passed without issue.

Fail: Failed during the standard[1] testing, a second unit was tested and also failed.

I want to be really clear that we are not de-emphasizing the importance of protections. We will still conduct the protections testing on the power supplies and present the results, but we will also be providing more nuanced results.

I don't think most people are making their purchasing decisions primarily based on the OPP limit of the power supply.[3] It is useful information to have, especially for enthusiasts who are searching for the limits, but it has the potential to obscure otherwise good test results and prevent collection of more relevant data.[2]

TLDR


We're planning on changing the order of the testing, and adding a fourth possible result to the power supply testing to better communicate when/how a power supply failed. We will go back through our previous results to update some to "Protections Fail", and we'll begin using the updated test order going forward. The PSU Circuit videos will remain unchanged, but at some point we may introduce a third video template for this new result.

If this is a terrible idea then we'd love to hear about it! There is an accompanying LTT Forum post where we hope to collect feedback on this change.[4] You can let us know if you believe our testing/presentation should stay the same, or if there is a better option that we're missing!

Lucas

[1] 'Standard' testing includes efficiency sweeps, load regulation, noise, excursion, brownout, etc. We've already written an article to explain these tests and results.

[2] Previously/currently we test at 0°C, 20°C, and 40°C ambient temperatures, in that order, with the protections testing at the end of 20°C testing. Currently, a failure during protections testing leads to no 40°C ambient results. Part of our problem would have been solved by just testing protections at the end of the sequence, but we also wanted to create some differentiation in the severity of failures.

[3] Next, we create an organization called 100PLUS, where power supplies earn badges based on protections testing.

[4] While we're at it, if you would like to provide feedback about the website, that can be done using this form, while feedback about Labs in general, that can be done in this thread.