RF Field-Probes PBS 1 / DC to 9GHz | Hacker News Reader

RF Field-Probes PBS 1 / DC to 9GHz

aaronia.com

These are usually called "near field probes". You can easily make a basic one yourself with a piece of thin coax [1].

As leoedin mentions, they are useless without some instrument to attach them to. Ideally a $100k+ spectrum analyzer, but you can sometimes get some basic readings out of them also with just a cheap oscilloscope + a RF detector [2].

[1] http://www.emcesd.com/tt120100.htm

[2] https://www.aliexpress.com/item/32747667435.html

"You can easily make a basic one yourself with a piece of thin coax [1]."

Fascinating! Did not know this! But in thinking about it, if all of these near field probes are actually small antennas of various characteristics, then technically coax would be an antenna, also... no?

Perhaps they could be used with an rtl-sdr device?

Disclaimer: Not trying to recommend a product or anything here, just never knew that RF probes existed for fields from DC to 9GHz... Of course, this is probably old information for some specialized electrical engineering professions (I mostly do software... EE is just a side hobby for me, so forgive the ignorance...). But anyway, I think the ability to detect fields of any frequency (well, up to 9Ghz!) is unbelievably cool! It gives me ideas for Star-Trek like devices...

These don't do anything on their own. You need to couple them with a spectrum analyser - another $1k+.

I've used these for hunting down problems with EMC on a PCB. They're quite useful, but it's hard to get a good picture of whether the device will actually fail or not based on the results. Generally if you put one of these near a switching power supply it'll go crazy, but that doesn't necessarily mean that the device will fail EMC testing. You really need a far field antenna with some sort of calibrated source to really know how you're doing.

You can still get pretty far with an RTL-SDR attached to them, which can be had for $25. I have successfully debugged radiated EMI issues with the combination of these two:

https://www.amazon.com/seeed-Studio-RF-Explorer-Antenna/dp/B...

https://www.amazon.com/RTL-SDR-Blog-RTL2832U-Software-Define...

What was the culprit? IME noisy switchers are easy to find, Improper signal termination impedance hard to find.

Nice!

"These don't do anything on their own. You need to couple them with a spectrum analyser - another $1k+."

Yes, I knew that you'd have to hook them up to an RF / frequency analyzer of some sort. Each of these probes is really just a small antenna with specific characteristics, is it not?

"Generally if you put one of these near a switching power supply it'll go crazy"

Hmm... interesting information... the physicist in me says that perhaps there's a relationship here between electrical power and all sorts of frequencies at the same time... but that would be only a hypothesis without testing it... is it the electrical line that generates these conditions, or the conversion of power... some power supplies up the voltage, then convert it down (sort of like the way PWM supplies more power to a PC's fan by sending more pulses per unit time)... so maybe this is where all of the electrical noise (aka, many frequencies at the same time) are... I woudn't know, but I would think there are some experiments to be done in this area...

Anyway, super interesting!

I’ve used the Langer probes in so many shops that I almost forgot there were other manufacturers of near field probes.

https://www.langer-emv.de/en/category/rf-passive-30-mhz-3-gh...

Generally these are connected to a high-end spectrum analyzer, but recently I’ve been working on a method to use a WiFi transceiver as sort of poor-man’s spectrum analyzer. This is useful only in applications where you want to snoop power levels in the range the chip works in. In my application we want to measure internal interference issues, within a device, where some non-WiFi subsystem creates noise in WiFi bands. Using the device’s own WiFi front-end to measure power becomes a very effective way to debug these issues.

The cool part is we did a correlation study between a $100k spectrum analyzer and the in-device WiFi chip, and across all noise sources and relevant frequency ranges the two agreed within 1-2dB. Not perfect, but good enough for our application.

"recently I’ve been working on a method to use a WiFi transceiver as sort of poor-man’s spectrum analyzer. This is useful only in applications where you want to snoop power levels in the range the chip works in."

That is insane! (I say that with admiration!) Insanely brilliant!

You'd never think (well, I'd never think!) of a WiFi transceiver as a spectrum analyzer (and yes, before HN goes nuts, we're only talking about the frequencies that the WiFi transceiver works in).

I love it! Brilliant!

If you click through to the datasheet, you can see the frequency response of each of the 5 probes. Each probe is capable of picking up signals down to low frequency, but they only have good gain around a particular frequency band. So it's not as if this is one probe that is equally good at detecting all frequencies.

Hmm... antenna (aka frequency probe) shape vs. gain... someone should make a table of those... actually, even better would be a chart with various combinations of antenna shape, material (e.g., metal/alloy A vs. metal/alloy B), and frequency response... even more interesting might be trying various different meta-materials, fractal shapes, etc., etc... Note to future self: Here might be some interesting experiments...

Reply on news.ycombinator.com