Near Field Probes Usage on Relative Measurements

[OneGeekGuy]

I open this post here since I would say is more Testing related than other topics.

I wanted to have experience from other people regarding usage of Near Field Probes, I have been using them several times to locate the "exact" place where radiation is happening after doing component level EMC testing. I think we can agree on the fact that they are not fully valid to perform a pre-compliance if we don't have any reference since we use to do these measurements at open air bench level. But I was thinking, if they could be use to perform relative measurements, lets say that I design in same PCB two similar designs for a DCDC, with some small changes in the way the layout is done, some more ideal, some less. If we place the PCB or PCBs in a Test fixture in a way we ensure the position from the Probe to the PCB is exactly the same in both, I would consider that could be possible to evaluate which of the two designs perform better relative to each other right?

They could be use for testing shielding capabilities was well, with a fixed setup...

Just asking this, because I would like to justify the need/usage of them at user level, also know experiences from others.

Thanks and regards!

[nctnico]

This is a valid approach just as long as you make sure to position the probe in the exact same place.

Another option is to use a lo-Z probe with a high attenuation (say 20x to) and probe signals on the board directly (for as long as they are within limits of the lo-Z probe). In the end the noise is radiated due to the shape of an electric signal so why not measure the signal itself? This leads to measurements with a better repeatability and also (when searching for the source signal) for an exact determination of the cause of excessive emissions.

[Conrad Hoffman]

Yes, they can certainly be used for comparative measurements. Be careful with direct probing as spectrum analyzer inputs are quite fragile. A protective device like Tekbox TBFL1 from Saelig is highly recommended.

FWIW, I figure that if I can't pick up much of any significant signal at near contact with a near field probe, compliance isn't going to be an issue.

[TheoB]

A near field H probe helps to identify what radiation is present. Then zoom in to the board and see if the current loops are as small as possible. I would not use a contact probe as that will also create ground loops. Relative measurements are of course fine to judge which of several solutions is the best, but at the end you still need a shielded room with antennes and rotation stand to achieve a reasonable pre-compliance check (at least that's how we do it)

[OneGeekGuy]

This is a valid approach just as long as you make sure to position the probe in the exact same place.

Another option is to use a lo-Z probe with a high attenuation (say 20x to) and probe signals on the board directly (for as long as they are within limits of the lo-Z probe). In the end the noise is radiated due to the shape of an electric signal so why not measure the signal itself? This leads to measurements with a better repeatability and also (when searching for the source signal) for an exact determination of the cause of excessive emissions.

Yes that would be the key factor, ensure the probe it is in exactly same place, also ideally same "environmental" conditions if possible.

What is the advantage of using lo-Z probe? I could understand you get better frequency response and lower capacitive load, but on the other hand, it depends on what you want to measure could be problematic. Also, might require a high bandwidth oscilloscope with it is "less trivial" than a SA right?

Thanks!

Yes, they can certainly be used for comparative measurements. Be careful with direct probing as spectrum analyzer inputs are quite fragile. A protective device like Tekbox TBFL1 from Saelig is highly recommended.

FWIW, I figure that if I can't pick up much of any significant signal at near contact with a near field probe, compliance isn't going to be an issue.

Thanks! I knew about attenuators but not specifically for this Transient limiter, something to take in consideration. Thanks!

A near field H probe helps to identify what radiation is present. Then zoom in to the board and see if the current loops are as small as possible. I would not use a contact probe as that will also create ground loops. Relative measurements are of course fine to judge which of several solutions is the best, but at the end you still need a shielded room with antennes and rotation stand to achieve a reasonable pre-compliance check (at least that's how we do it)

Agree, I would not go either with contact probes for this approach. Also, it is clear that the only way to get accurate pre-compliance is to go to EMC lab (I did-do the same). But this method might help trying to get or tune at bench level certain critical aspects into layout for example.

Thanks!