2
\$\begingroup\$

I'm using a step-up DC/DC converter (https://www.torexsemi.com/file/xcl101/XCL101.pdf) in my circuit to boost a 1.5V AAA battery voltage to 3.3V. The 3.3V is powering an audio circuit, and I heard a lot of noise in the audio signal. I suspected the issue was a noisy power supply, so measured the 3.3V voltage on the scope, AC coupled: enter image description here

Disregarding the spikes, the peak to peak ripple is about 40 mV. Would you expect this level of noise ripple at the output of a DC/DC step up converter?

This is the scope shot when I measured the AC coupled voltage across the battery by itself: enter image description here

I'm not sure of the cause of the large spike, but it looks unrelated to the constant 40 mV noise on the 3.3V voltage.

This is the configuration I'm using for the 3.3V step up converter (from the datasheet):

enter image description here

The only thing I can think of that might help is using different capacitors. Right now I'm using ceramic 0603 capacitors (https://www.digikey.com/product-detail/en/murata-electronics-north-america/GRM188C81C106MA73D/490-7198-1-ND/3900421). The 3.3V step up converter's datasheet also recommends ceramic 0603 capacitors, however I noticed the specs for them indicate "Low ESL" (https://www.digikey.com/product-detail/en/tdk-corporation/C1608JB1A106K080AC/445-11201-2-ND/3948437). I'm not sure if using this recommended capacitor will make that much of a difference...could this be the problem?

\$\endgroup\$
  • \$\begingroup\$ Is it actually noise, or is it periodic? \$\endgroup\$ – vofa Jan 25 '18 at 2:51
  • \$\begingroup\$ Please note that probing is important. Also, IMO, 40mVpp or 14.4mVrms for a 3.3V output is not a problematic value because it equals to 0.4% ripple. Most likely the audible noise comes from the built-in inductor because the inductors tend to "vibrate". \$\endgroup\$ – Rohat Kılıç Jan 25 '18 at 3:51
  • \$\begingroup\$ I just measured on the scope again with a 10 us time scale, and the ripple is actually periodic at 16 us, I'll add this scope shot to the question. \$\endgroup\$ – donut Jan 25 '18 at 5:40
3
\$\begingroup\$

Place RC lowpass filter on output of your power supply: 100 ohms and 100uF. This is a 0.01 second time constant, -3dB at 16 Hz, -20dB at 160Hz, -40dB at 1,600Hz, -60dB at 16,000Hz.

So use your scop probe across the 100 uF capacitor. CLip the scope probe GND lead to -side of the 100uF, and clip the probe tip to +side of the 100uF.

=============================================== Note the ESR and ESL will limit the far-out attenuation.

=================================================

That first screen shot shows 14 uS periodic ripple, or 70KHz.

A 0.01 second TAU RC LPF, at 15Hz F3dB, is 5,000X slower, thus might ultimately produce 5,000:1 cleaner voltage.

\$\endgroup\$
  • \$\begingroup\$ Why would you use a lowpass filter, and how did you choose a 0.01 second time constant? \$\endgroup\$ – donut Jan 25 '18 at 4:56
  • \$\begingroup\$ The other options for reducing high frequency ripple I have seen are using lower ESL capacitors (not sure of the impact to this issue), and "what you need to do is add a capacitor that has a self-resonant frequency that is slightly higher or at this [noise ringing] frequency."(training.ti.com/capacitor-selection-high-frequency-ripple) Why don't they just suggest a simple low pass filter? \$\endgroup\$ – donut Jan 25 '18 at 7:11

Your Answer

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

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