The Story Behind Cybenetics: Past, Present & Future

LAMBDA Noise Certification

Let’s start first with the obvious question. What does the LAMBDA name stand for?

Nothing in particular. I just wanted to use a Greek letter to name the Cybenetics sound level certifications, and I chose Lambda (uppercase “Λ,” lowercase “λ”; Greek: Λάμ(β)δα lam(b)da), which is the 11th letter of the Greek alphabet.

Noise Levels (115V/230V Input) Noise Requirements (1)
A++ <15 dB(A)
A+ ≥15 dB(A) & <20 dB(A)
A ≥20 dB(A) & <25 dB(A)
A- ≥25 dB(A) & <30 dB(A)
Standard ++ ≥30 dB(A) & <35 dB(A)
Standard + ≥35 dB(A) & <40 dB(A)
Standard ≥40 dB(A) & <45 dB(A)

(1) If the DUT is close to the lower limits, with a maximum margin of 0.25 dB(A), we will provide a second chance to the manufacturer which will have to provide us with another sample. In case the second sample registers a significant overall difference compared to the first one [>2 dB(A)], we will ask for a third one, and the average noise measurements of all three samples will deliver the final result. 

When we started the Cybenetics test programs in 2017, there was no certification for noise output in IT parts, and currently, only the Cybenetics LAMBDA exists. This is because it is complicated to find a way to measure a PSU’s output noise accurately and have a single number to describe it.

I will quote again the PSU Test Protocol listed on the Cybenetics site to describe the whole procedure of how we take PSU noise measurements:

As the efficiency measurements procedure mentioned, we apply at least 1450 different load combinations on the DUT rails while monitoring all vital data, including the fan speed. With the fan speed range data in hand, we take noise measurements with as small intervals as possible, with high accuracy, and covering the entire range. For example, if the fan speed range is 400-2000 RPM, we take noise measurements per 50 RPM intervals.

The noise measurements are taken in a hemi-anechoic chamber with the DUT switched off and its fan connected to an external power supply, applying the voltage required to achieve the desired fan speeds. Moreover, the fan speed is continuously monitored by a tachometer. This way, we can eliminate third-party noises, including the noise of electronic loads.

We make a table with the fan speed in RPM and the corresponding noise at that speed. Our software then uses this table to analyze all data gathered during the load tests and assigns a decibel value to each fan speed value.

Once we have a dBA value for each of the tests we conducted with multiple load combinations, we convert the dBA values to SPL to average them, and once this is done, we reverse the outcome to dBA again. This procedure allows us to have a single number describing the DUT’s average noise output with at least 1450 load combinations. According to this number, we tax the DUT into one of the LAMBDA categories.

Besides the fan’s noise, we also check for electronic noise (coil whine) during the PSU’s load testing. The human ear is the best instrument to check for coil whine, especially if its owner is young! Thankfully, most of the Cybenetics engineers are below 30.

Pages ( 6 of 7 ): « Previous12345 6 7Next »

Related Posts

4 thoughts on “The Story Behind Cybenetics: Past, Present & Future

  1. Από το thelab?!

    Με πηγές πολλά χρόνια πίσω. Διάβαζα τότε τα ποστ (στο worklog?) και το θαύμαζα.

  2. Hi Aris and team,
    As a PC building lover, cybenetics and hwbusters have been a great help for me and other people I have recommended to, especially on your PSU reviews and Fan Case reviews.
    I just know this website and channel not long ago. Looking up for a proper, standardized, and comparable data sheet from a 3rd party is not something you can find easily for a PSUs. Since I already dive deep into this hwbusters and cybenetics I’ll try to spread the words to some other community I know.

Leave a Reply

Your email address will not be published. Required fields are marked *