0
\$\begingroup\$

I have a buck converter and I am measuring the voltage ripple at the output capacitor along with that I am measuring the switching frequency at the cathode of the diode. Please refer the below image. I am using 500MHz keysight scope.

enter image description here

Probe 1: Output Voltage across output capacitor (AC coupling mode is used with short ground tip)

Probe 2: Switching frequency measured at cathode of diode (DC coupling with ground of the probe connected to scope ground)

While measuring these two simultaneously, my output ripple voltage is picking up high noise. If I don't measure the switching frequency along with the output voltage ripple, I have no problem. While measuring together, my output ripple voltage probe, picks up noise.

What might be the problem?

\$\endgroup\$
  • \$\begingroup\$ Show your probe setup! \$\endgroup\$ – winny Nov 21 '19 at 11:06
2
\$\begingroup\$

The output of your switching converter is a low-impedance node (at high frequency, it is basically a capacitor) so you don't have to worry about the probe loading the probed signal.

It is a hostile environment though:

  • Coil and traces emit AC magnetic fields...

This means the Magnetic EMI Loop Antenna Sensor in your scope probe (also known as a ground wire with alligator clip) will happily pick up any stray fields and turn them into spurious signal on your scope. It also adds inductance in the ground connection which will do all sorts of bad things to high speed signals (spurious ringing, lowered bandwidth, etc). You need to get rid of that loop antenna, and/or make its loop area as small as possible. Most likely, if you got a 500MHz Keysight scope, the probes came with these ground springs, so use them. It will look a lot cleaner.

enter image description here

  • There is HF AC current going through "GND"

Thus "GND" will not be at the same potential everywhere, and you should put the probe's GND spring tip on the correct spot, which probably won't be too close to where the switching happens. It is also a bad idea to probe on a capacitor. Since it shorts AC, ripple will be lower... but since it has AC current running through it which creates voltage in the pin inductance, HF ripple might also be higher... or just wrong.

If you're interested in ripple at the load, then... probe there. Or on vias on your ground/power plane. Or any place downstream of the DC-DC output caps. But not directly on the caps.

While measuring these two simultaneously, my output ripple voltage is picking up high noise. If I don't measure the switching frequency along with the output voltage ripple, I have no problem.

I'm going to guess your two probes' Grounds are at different places on your PCB, and there is a voltage difference between these two "GNDs" on your PCB, which creates a current inside the probes, cable shields, and scope internal ground. This can be a problem. Try putting the probe grounds in the same place, and a bit away from the switching.

You can clip a ferrite sleeve on one of the probe cables to increase its common mode impedance, too.

If you want a high-quality measurement of output ripple, but you're just interested in the switching signal on the diode to measure frequency or trigger the scope, but you don't care about signal integrity or accurate measurements, then simply not connecting the ground of this probe at all may be a solution. Use the ground clip on the probe which measures ripple, and just use the tip on the other probe.

You can also solder a 51R resistor and a 50R coax, then connect that to your scope. Make sure to use 50R termination on the scope, and double check if the termination resistor inside the scope is specified to take the voltage you're using. You can AC-couple it at the source by adding a cap in series with the 50R resistor too.

\$\endgroup\$
  • \$\begingroup\$ Thanks for accepting! So how did you solve the problem in the end? \$\endgroup\$ – peufeu Nov 21 '19 at 18:01
  • \$\begingroup\$ Hi. I haven't solved my problem yet. My two probes have the ground wire with alligator clips. One more question : I am using the 500MHz keysight scope. Does the scope ground (which is present on the front side before the probe slots) is connected to mains earth? \$\endgroup\$ – Newbie Nov 22 '19 at 8:18
  • \$\begingroup\$ Please check this -> I have created a new question with my waveforms -> electronics.stackexchange.com/questions/468525/… \$\endgroup\$ – Newbie Nov 22 '19 at 8:32
  • \$\begingroup\$ Usullally scope ground is earthed... you can check with multimeter. \$\endgroup\$ – peufeu Nov 22 '19 at 10:25
  • \$\begingroup\$ I hadn't thought that both measurements would be taken with different scope horizontal sweep rates since the question doesn't mention it and you didn't include screenshots... Bimplerekkie's comment in the other question makes sense! \$\endgroup\$ – peufeu Nov 22 '19 at 11:40

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.