1
\$\begingroup\$

This 2.2 kg 1000W BeQuiet BN213 ATX power supply is able to provide 12V 83A (no typo. Eighty-three.), which is around 1000 W. enter image description here

This huge PS900W (60V15A 30V30A 15V60A) bench power supply weighs 5 kg already, reaches 900 W, but is significantly larger. enter image description here

This ~2.4kg PS-3010DF variable laboratory SMPS has 0–30V with up to 10A, which maxes out at 300W. It's housing is significantly larger than the 1000W ATX power supply. enter image description here

This 230W 19.5V 11.8A laptop power supply is significantly more compact, weighs lighter and does not require any external passive/active cooling (heatsink or fan): enter image description here

All of them are switching-mode power supplies.

How come the PC and laptop power supplies deliver much higher performances per same volume? What are the technical reasons?

Is it because the bench power supplies have a variable, adjustable output voltage while the PC/laptop power supplies deliver a fixed voltage?
But most PC/laptop power supplies usually have a current control mode when voltage controlled is no longer possible due to overcurrent. (also known as “current limiting”)

Is it maybe because of voltage ripple filtering or something similar?

\$\endgroup\$
4
  • \$\begingroup\$ The question is answered already, but I encourage other users to answer as well. Maybe there is something that answer does not contain. \$\endgroup\$
    – neverMind9
    Jun 8, 2019 at 22:49
  • 1
    \$\begingroup\$ you don't need to say that :) that's a given. However, if you want to encourage more answers, because you don't think my answer is comprehensive, maybe simply wait a few hours before clicking the "accept" checkmark (you should be able to remove it for now!). I don't mind! \$\endgroup\$
    – Marcus Müller
    Jun 8, 2019 at 22:50
  • \$\begingroup\$ @MarcusMüller No, I consider your answer complete, very comprehensive and very accurate. I just wanted to say it should not stop others to answer as well, just in case. \$\endgroup\$
    – neverMind9
    Jun 8, 2019 at 22:54
  • 1
    \$\begingroup\$ Variable SMPS over a wide Voltage and current range require many more circuits and magnetics to handle this dynamic range in a low noise output. Fixed regulators are very efficient in size with forward converters all tightly regulated by high mutual coupling and winding ratios. \$\endgroup\$
    – Tony Stewart EE75
    Jun 8, 2019 at 23:02

1 Answer 1

Reset to default
3
\$\begingroup\$

Because lab supplies will actually be tested by people with unpredictable loads and high-quality measurement equipment.

Thus, the quality of voltage stability and accuracy needs to be significantly higher for lab supplies.

Do a test: get a 2 Ω power resistor and attach it to your notebook supply to draw ca. 10 A; measure the voltage with that load, and without a 100 Ω load. Compare. Do the same with a lab supply. If you have access to that, also observe the ripple with an oscilloscope.

Noise is also much more of an issue for lab supplies – and that is something that very much is noticeable in the size of components, because SMPS lab supplies will a) tend to overdimension their inductors, and b) do a significant amount of voltage drop in a linear voltage regulator following the SMPS, since that doesn't introduce switching noise, but needs larger heat sinks.

Also, you can optimize a PC supply for the very limited set of voltages it needs to generate; lab supplies are usually adjustable and hence harder to size-optimize for a given load and voltage scenario.

\$\endgroup\$
11
  • \$\begingroup\$ “and without a 100 Ω load.” – Did you mean “with a 100 Ω load”? \$\endgroup\$
    – neverMind9
    Jun 8, 2019 at 22:54
  • \$\begingroup\$ “Noise is also much more of an issue for lab supplies” – Isn't that one of the purposes of linear power supplies already? \$\endgroup\$
    – neverMind9
    Jun 8, 2019 at 23:00
  • 1
    \$\begingroup\$ linear supplies will not operate fast enough, to remove input ripple/trash at high frequencies. I've seen LDOs operating on 1uA, that barely handle 40dB reduction of 60hz; any switchreg ringing will blast right on thru. \$\endgroup\$
    – analogsystemsrf
    Jun 9, 2019 at 2:24
  • 1
    \$\begingroup\$ @analogsystemsrf hm, supplies for sensitive (as in: radio receiver) circuitry I've seen basically all look like: SMPS step down -> passive filtering -> LDO -> more filtering; that makes a lot of sense to me. But in these cases, the switchers often were even synchronized to other clocks in the system to do as little harm as possible. \$\endgroup\$
    – Marcus Müller
    Jun 9, 2019 at 6:03
  • 1
    \$\begingroup\$ Also, while it's definitely true that SMPSes inevitably cause noise, your radio usually should not be picking that up so strongly. The fault is not only with the SMPSes – these probably have reasonably low EMI – but with your radio, being more sensitive to noise on different frequencies than to actual radio transmissions. \$\endgroup\$
    – Marcus Müller
    Jun 9, 2019 at 21:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.