Re: FunCube Dongle ProPlus noise and RFI improvements. Tips please.


jdow
 

There is a very basic point regarding ground to get straight before you go further. Ground is not ground. It is NOT what comes to your mind unbidden. It is not a giant unipotential plane of unobtainium perfect conductor within your shack. Even a perfect conductor does not make a ground pane. You might as well consider everything is floating and one point is picked as a reference for all other voltages including that ground on the schematic that is 2" away from the one you picked as a reference. The speed of light sees to that. Look at all the fancy loops in your system from the small one around the edges of your enclosure. Some are small. Some are large. ALL pick up energy one way or another, which segues us to another consideration.

Another thing to consider is that wires do not carry the energy in the system. Fields are everything in electronics. The wires properties guide the fields. That is why loops pick radiate and pick up energy. This is as true at 20 Hz as at cosmic ray frequencies and all in between. So you want to make the loops you can find in your system as small as possible and, where appropriate, as high impedance as possible. This latter point is critical with the giant loop that forms your "ground" from the electrical circuit breaker cabinet through your ham gear to your grounded antenna (for lightning reasons at least) and back through the earth, if nothing else. Clamp-on ferrite beads are some of your best friends. They raise the losses within your big loops, such as the power cords and antenna leads in your system. Model it however well or crudely as you choose. Look for places that will increase the impedance to reduce the circulating currents. And you might also look for ways to provide a low impedance from end to end of your box. (And pick the ferrites to be optimum for the frequency range(s) of interest. A bead that looks like 200 ohms at 200 MHz is probably worth less than nothing at mid-HF and down.)

Addressing a specific issue - USB and SMA sockets - are you doing a relay out on the funcube? If not you are kinda stuck there. You MAY do better with the USB shield grounded to the RFI enclosure as is the SMA socket. Small is likely better than a nice big 10" cube for this. Ferrite beads on both ends go a long way towards making the box as close to unipotential as possible. This means you have a better chance that the ground conductor in the USB cable is closer to the chassis potential than without the ferrite bead. Less unwanted energy couples into the funcube board. That is good. Short cables sound nice. A longer cable can be wound around larger beads, such as needed at say 3 MHz, for better isolation. Shielded is good. Drain wire - it may help bridge gaps. I am not sure it does much more for a good cable shield. For a make it yourself box use double sided copper clad PCB material and solder the sides. Laptop running on battery can go a long way towards breaking that big ground loop that includes the antenna feed. There is still some capacitance to ground. But, it is a small effect. Soldering PCB copper to PCB copper does not require a wrist strap unless you are into the oxygen free cable nonsense.

{o.o}

On 20220924 02:07:51, David Coles wrote:

  • Seek to create an RFI enclosure.
  • Might both USB and SMA sockets be commonly earthed?
  • Acknowledge heat sinking by both sockets.
  • Upgrading connection with 4 metres (max) shielded and drained USB cable.
  • Fitted with 2 clip-on ferrites.
  • Laptop running on battery.
  • Considering adhesive copper foil as a shield.
  • Should I use an anti-static wrist-band for this job?
  • Praise to all correspondents!
  • 73 Dave G7GZC
  •  

Join main@SDR-Radio.groups.io to automatically receive all group messages.