Rf feedthroughs and power ground

[coppercone2]

If you have an rf can that needs power typically what i see is that capacitive feedthroughs are used for any power lines and there is a turretthats screwed into the outside of the can for power ground, i think.

Is this always the correct situation? Do you ever put the ground through some kind of feedthrough, even if its not floating in the box? I am fairly certain i saw this done in a 11ghz bias t and fairly sure i saw it done for a spectrum analyzer cans that were hard wired. Just turret to chassis exterior.

or how about like a noise source or something thats connected to a specturm analyzers BNC output DCV? Do you add an additional current carrying wire inside of the BNC connected to the PCB or do you just use the chassis interior? (i.e. use a BNC with a little solder tab you wedge onto it with the provided nut?).

[cdev]

The skin effect applies to RF, not DC!

But you make some good points.

Although this subject interests me too, "don't follow me, I'm lost too", being a non-professional beginner, really.

With a PCB the path for the DC supply voltage may be less prone to serving as a channel RFI staying on (or inside of a multilayer) PCB than exiting via a lead in any other plane than the PCBs own and a feed through cap going off to a side could introduce resonances.

I don't know what they are called but they have (three terminal) two-dimensional (PCB) equivalents to feed throughs. Newer equipment likely uses all of these different kinds of bypassing parts in different applications.

[German_EE]

Just to add something to the discussion, two working principles:

1) A screen will only be effective if it carries no current

2) Any DC current will find the shortest possible path to ground

So, build your equipment as a series of small modules with PCBs inside either tinplate or unpainted diecast boxes. Use RF feedthroughs and turrets for power and control connections and your RF connector of choice for signals. Providing the PCB enclosure is fixed to the chassis at multiple points the circulating currents (if any) will be very small.

As for the PCB inside, you have two choices. Either a) seam solder the PCB ground plane around all four edges to the case so that ground REALLY means ground or b) have one and ONLY one connection between the PCB and the case. Both methods are an attempt to reduce circulating currents inside the case and I prefer the first approach as it's less messing around.

[coppercone2]

would you use a capacitive feed through on the ground thats the same as the power feedthroughs (i assume its a round ceramic cap), or would you use just a isolated one with minimal capacitance? I am not sure how much an effect it would have on common mode.

How come bias T's just use the case?

I am thinking more like a big powerful RF can, say with a power MMIC amplifier in it that needs say 2-3 A, quiescent, or possibly more with a parallel network.

I was thinking about soldering the PCB to the RF can using nanofoils so there is a full connection.

[cdev]

>"How come bias T's just use the case?"

Because the skin effect is useful in RF shielding, because it just happens to RF, not DC. You want DC on your power return (the ground) and ALL the Rf to remain inside the box and on the RF lines.

BTW, I am having a hell of a time right now with RFI on an RPI hat - which I am thinking of abandoning in favor of short ugly construction.

If you want a great way to learn how RF shielding works, I suggest building some LNAs using MMIC chips. Try out different construction methods, cases, RF interconnects, DC interconnects, DC blocks, bias-tees, and via styles. The fact that LNAs amplify amplifies your learning!

Using SMT makes it a lot easier probably because it eliminates a lot of potential complex effects from the picture. Less is more in that respect.

They have some pretty neat feed-throughs now that claim to incorporate a whole RF filter inside. Not just a capacitor.

The standard-shaped feed-through ones work well.

That should be enough in most situations. If ts not maybe add some more bypassing and L outside the box.

But here I am still struggling with an RFI situation that should be addressable. Need to throw a bunch more caps and ferrite beads on there. Bypass the power lines with those 3 terminal 2d feed through caps for SMD. Or put the RPI and GPS in a partitioned screened metal box. (because of the heat) Or dump the hat.

[cdev]

The problem is that good RF cases bought are pricey and the connectors most often used for RF are oriented towards PCBs without cases. The area that Ive found to be problematic is the edge of the PCB-to case interface. Aluminum cases are the most available but aluminum doesnt take solder well. Another option is using copper foil inside the aluminum case to extend the edges of the PCB. Where the connectors enter and leave the device are the important grounding spots. Less so the edges but it wouldnt hurt to ground them too. In practice making a c around the pcb with copper foal and using coax connectors that both connect to the PCB and this bendable C works but it only works well if the PCB length matches the case exactly. The problem I've have is that it never quite does. So I've bought the cases I could find and inside where it cant be seen ended up using two different kinds of coax connector in two cases, trying to bridge the gaps with dabs of solder and this copper foil back plane and a short length of coax between a SMA bulkhead and the actual end of the trace at the end of the PCB, observing the best practices I can (grounding on both sides bla bla bla). Its a mess. But it does work better than if I don't do it. I dont have a machine shop so right now that is the best I can do. Any suggestions?

At some point I will post pictures. Its not that bad. Just not photogenic.

Just to add something to the discussion, two working principles:

1) A screen will only be effective if it carries no current

2) Any DC current will find the shortest possible path to ground

So, build your equipment as a series of small modules with PCBs inside either tinplate or unpainted diecast boxes. Use RF feedthroughs and turrets for power and control connections and your RF connector of choice for signals. Providing the PCB enclosure is fixed to the chassis at multiple points the circulating currents (if any) will be very small.

As for the PCB inside, you have two choices. Either a) seam solder the PCB ground plane around all four edges to the case so that ground REALLY means ground or b) have one and ONLY one connection between the PCB and the case. Both methods are an attempt to reduce circulating currents inside the case and I prefer the first approach as it's less messing around.

Most recently Ive had decent luck with a short piece of *thin* copper pipe cut with a copper pipe cutter. This works up to the size my pipe cutter handles which is pretty small. So no large devices. Very small devices only. I used copper flashing (foil-like) for the ends. And it is possible to attach the edges with dabs of solder.

[coppercone2]

if you plate the aluminum you can use a nanofoil to bond the entire PCB directly to the chassis, or if you use copper. I think using conductive epoxy would work too in such a wide surface area.

[David Hess]

I have used a feedthrough for ground when I needed ground to be connected to a specific point inside to prevent a ground loop.

[cdev]

Doh!

Nanofoil = copper tape, right?

Copper tape which I now have a bunch of would work great. Even if it doesn't make perfect connection, the capacitance is likely high enough for it not to matter.

How stupid I was not to think of it.  David, what you said makes a lot of sense too. say anchor your wire right to the edge of the PCB at the closest edge to the device that needs the good ground the most? Then your return signal goes right through the wall to both the inside and the outside grounds too, despite the skin effect. I.e. covering all the possible bases. That makes sense even if my brain isn't working tonight.

[coppercone2]

I have seen the RLC/RFC feedthroughs too.



I want this.

500$ for the starter kit though.

[David Hess]

How stupid I was not to think of it.  David, what you said makes a lot of sense too. say anchor your wire right to the edge of the PCB at the closest edge to the device that needs the good ground the most? Then your return signal goes right through the wall to both the inside and the outside grounds too, despite the skin effect. I.e. covering all the possible bases. That makes sense even if my brain isn't working tonight.

For me it came up when designing low noise voltage controlled oscillators where the ground loop with the chassis was corrupting the power, ground, signal, and common going to the oscillator.  Low jitter digital logic faces the same problem.

An alternative option is to use feedthroughs to bring out points for Kelvin connections where there are a separate force and sense lines.  It seems kind of silly to have a feedthrough connect internally to the chassis ground but sometimes it is the only way.

[German_EE]

If you want to solder to the inside of an aluminum or diecast case then run some conductive adhesive copper tape around the inside and solder to that. I've done this lots of times and providing your PCB is a snug fit the tape will never move.

[cdev]

I have used copper tape successfully in other cases. (lining cardboard ones) I think it will work well with the aluminum ones as you describe.

Good case grounding makes all the difference with a low noise amplifier for VHF/UHF. If you use a case it needs to either be done right or you're better off without it.

If you want to solder to the inside of an aluminum or diecast case then run some conductive adhesive copper tape around the inside and solder to that. I've done this lots of times and providing your PCB is a snug fit the tape will never move.