4
\$\begingroup\$

I have been having ringing spike issues at the output of SMPS power supplies. I don't know what is this phenomenon is about but it is definitely related to switching devices on the same electrical network. When a switch or switching device is on the same electrical network. The switching device upsets the SMPS output. This is directly related to my previous question. Bu that time I didn't figure out the issue starts with SMPS output getting upset.

To demonstrate that I used a grinder as shown below:

enter image description here

And next to the grinder from the same power strip a 5V SMPS is powered.

Here is the output of the 5V power supply and the spikes "during" the short time the grinder is swithcing ON:

enter image description here

And here looks like high frequency ringing when AC coupled:

enter image description here

The problem is this power supply noise propagates through all sensitive devices it powers.

Does anybody have experience with such phenomenon? I even tried with EMI power line filter and isolation transformer but it didn't help. I guess SMPS already have filters at their input. Is there a way to filter out this at the power supply output? Would that issue happen with a linear supply?

\$\endgroup\$
  • 3
    \$\begingroup\$ How long is the ground clip lead on your oscilloscopes probe? \$\endgroup\$ – Marla Nov 16 '18 at 14:46
  • 1
    \$\begingroup\$ are you sure this is real? as @Marla is alluding to \$\endgroup\$ – JonRB Nov 16 '18 at 14:49
  • \$\begingroup\$ @Marla 15 cm. But I don't think this is a measurement error due to the wrong probing. I'm saying that because when there is no scope this SMPS upsets the sensitive devices it powers. There is a direct relation between the switching device upsetting the SMPS output. Scope measurement came later after the issue. \$\endgroup\$ – atmnt Nov 16 '18 at 14:49
  • 1
    \$\begingroup\$ To draw conclusions from your oscilloscope measurements, you need to know the limitations of your setup first. Yes, you will still have a problem and you need to show layout and schematic to get help past that. \$\endgroup\$ – winny Nov 16 '18 at 15:14
  • 1
    \$\begingroup\$ Seems normal, but a block diagram would help here. If your problem exists even without scope, we still need schematic and layout to help you. Probably poor layour and insufficient decoupling. \$\endgroup\$ – winny Nov 16 '18 at 15:48
0
\$\begingroup\$

SMPS as a unit has high bandwidth therefore any not filtered high frequency signal at its input will appear at its output. DC-DC linear chips are also have the same frequency response and any high frequency signal at their input could be appeared at their output. just in case that you are using isolated transformer with iron core for an AC-DC convertor you have an intrinsically filter for some high frequency signals.

If you cascade proper Common Mode Filter at the DC/AC input and then proper low pass differential mode filter at the SMPS input, you could be sure about bypassing the injected noise at the input of SMPS from its output. Remember for measurement such phenomena it is recommended to use probes in their X10 mode and do not use their attached long 3-4 inch Ground cable. you have to use the very short ground clips that is companion with probe's purchased package.

Ringing is also companion with an SMPS that is not implemented with good PCB. Please look at the related TI document (SLPA005). In addition to conducted emission the Grinder has strong radiated emission too. using proper casing and filter (Common mode) at the input and output of the SMPS could be helpful. Please look at the following links: https://www.giee.org/wp-content/uploads/2016/08/ferrite_bead_181_hex_open_and_closed.jpg

https://www.radiodan.com/Henry/parts/RF_chokes.htm

\$\endgroup\$

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.