1 of 16

Basics of RF Emissions Debugging

How to fix a failed test

Presented by Alex Whittemore for the Hackaday Remoticon 2020 workshop series�https://hackaday.io/project/175080-remoticon-basics-of-rf-emissions-debugging�alex@alexw.io https://www.alexwhittemore.com

These slides: https://docs.google.com/presentation/d/1QJ2Pu3y-F7zJg2ZN3TOdezDzjAVtaXPsHZbvoPqMlZ8/edit?usp=sharing

​

2 of 16

Background info

Software setup instructions/background are at https://hackaday.io/project/175080-remoticon-basics-of-rf-emissions-debugging/log/185818-software-setup-notes

We’ll need to install/set up:

  • Workbench materials
  • SDR toolchain
  • TinyFPGA toolchain

3 of 16

Workbench Materials

Your workbench should include the following:

  • RTL-SDR
  • Low Noise Amplifier (LNA) and some way to power it (probably the SDR)
  • Magnet wire and SMA connectors (unless you’ve already built probes)
  • Soldering iron
  • SMA connectors
  • TinyFPGA BX
  • Cables to attach everything together (including Micro USB)
  • Maybe some other boards to play around with measuring.

4 of 16

Software setup: SDR

We’ll make use of the following SDR software:

  • QSpectrumAnalyzer (Critical)
  • SDRAngel (Strongly preferable, but if you have something else set up to capture a spectrogram, that’s probably good enough)

DragonOS LTS includes these out of the box; here’s a VirtualBox VM �ready to go.

VirtualBox performance is marginal: If you can, import this VM into�Parallels Desktop (on Mac) or VMWare Workstation.

5 of 16

Software setup: TinyFPGA

TinyFPGA tooling setup is (relatively) painless.

  • Install Atom
  • Follow the setup instructions for TinyFPGA
  • Clone the repository. No peeking!

6 of 16

What is RF Compliance?

Electronic devices emit signals, intentionally or unintentionally, that may interfere with other devices:

FCC part 15 Subpart B governs “Unintentional Radiators” - Electronic devices that may cause accidental interference. ��FCC part 15 Subpart C governs “Intentional Radiators” - devices that �specifically implement wireless radios to broadcast RF energy.

7 of 16

How does testing work?

  • Put device inside a Farady cage
  • Point a big expensive antenna at it
  • Turn it on and spin it around
  • Measure the antenna with a big expensive spectrum analyzer.

8 of 16

And then?

Whoops, you failed.

9 of 16

And then?

Whoops, you failed.

10 of 16

And then?

… Or did you?

11 of 16

And then?

Well, not this mode, anyway.

12 of 16

So what do we do about it?

  • Fiddle with firmware
  • Look at the layout and correct any obvious deficiencies
    • But, like, you already did that… didn’t you?
  • Sniff around with a magic wand (near field probing)
  • Near field? Far field?
    • << 1 wavelength or >> 1 wavelength
    • Also about impedance far vs near
    • Magnetic field dominates in near-field.

13 of 16

Near Field Probing (magic wand method)

  • Get a spectrum analyzer
  • Get near field probes
  • Wave them around your board to try to pinpoint the issue

(This is what we’ll be doing)

14 of 16

Near Field Probing (Expensive toys method)

15 of 16

Demo: Using expensive gear

  • Tek RSA507A
  • R&S HZ-15 Probes

16 of 16

Demo: Building a probe and using cheap gear

  • Let’s make our own probe!
  • And use it with realllly cheap receiver hardware!