I am designing a Raspberry Pi shield and it needs to have two separate supplies: one for the RPi itself, and one for the rest of the circuitry (consisting of energy metering ADE7758 chip). I plan on using Mean Well switching mode power supplies. The ADE7758 part of the circuit needs 5V/100mA. The RPi supply needs to provide 2A at 5V.

The concern I have is about the power supply ripple and I am looking for advice on how to calculate the value for the smoothing capacitor or any other method for smoothing the output of the supply. I don't want to plug in a LDO regulator after the SMPS as I don't have enough space on the board. The datasheets of the power supplies specify 150mV and 200mV ripple but do not specify the frequency, so I assume it is 50Hz.

Using the equation for capacitor discharging enter image description here is giving me unreasonable values if I use Vc = 5.000, starting voltage 5.075V and t = 0.01s: 13.43mF. That is if I want to reduce the 150mV ripple in half.

Edit 1 (as suggested by the comments): Acceptable ripple for the RPi supply is 120mVp-p, and for the ADE7758 power supply is 50mVp-p.

Edit 2: Adding a schematic.

enter image description here

  • 1
    \$\begingroup\$ The concern I have is about the power supply ripple Why? Before assuming it is an issue you need to quantify it. What amount of supply ripple is acceptable? Just "having a concern" is OK but before you can fix it you need to know what you're up against. Suppose you decrease the ripple by a factor 10, is that enough? Or do you need a factor 100? \$\endgroup\$ Nov 28, 2019 at 9:40
  • 4
    \$\begingroup\$ ...ripple but do not specify the frequency, so I assume it is 50Hz. Why? Because the mains frequency is 50 Hz? But the PSU is an SMPS, it rectifies the mains frequency and then uses switching and a high frequency transformer (usually 100 kHz or so) to convert the power to a lower voltage. So 50 Hz isn't an issue, the 100 kHz might be. For that you need low ESR capacitors, not capacitors that work at 50 Hz. \$\endgroup\$ Nov 28, 2019 at 9:43
  • \$\begingroup\$ According to datasheet (meanwell.com/Upload/PDF/IRM-02/IRM-02-SPEC.PDF) switching mode PS operates at 130kHz. It also says: "Ripple & noise are measured at 20MHz of bandwidth by using a 12" twisted pair-wire terminated with a 0.1uf & 47uf parallel capacitor". With a reasonable output decoupling strategy, like @Bimpelrekkie mentioned, you could potentially get lower output ripple. Have you looked for similar drop-in packages with lower ripple specs? \$\endgroup\$
    – eeintech
    Nov 28, 2019 at 13:03
  • \$\begingroup\$ I am stuck with the Mean Well supplies, but I do not have on at the moment for testing. The clarification about the output ripple frequency is really useful, that means I can use relatively large electorlytic capacitor with small ESR. \$\endgroup\$
    – Slav
    Nov 28, 2019 at 14:02
  • \$\begingroup\$ @Slav I wouldn't recommend just electrolytic caps, but a combination of an electrolytic and multiple MLCCs (MLCCs have much much lower ESR). Also looking again at your circuit, I would recommend moving the fuses downstream (MeanWell module > decoupling caps > fuse) and make sure it is rated to handle the inrush current of your downstream devices. For the LED circuit, you may want to use a higher value for the resistor, with only 2-3mA most green LEDs are already well visible in daylight. What is your intention behind adding inline ferrites and parallel fuse at the power input? \$\endgroup\$
    – eeintech
    Nov 29, 2019 at 16:39


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.