Electronics > RF, Microwave, Ham Radio

Grounding in a RF High-ESD Risk Mobile Application

(1/2) > >>

selkathguy:
I've done some searching on here and google and SE, and I haven't been able to find a straightforward response.

I have a battery-powered device in a small 100%aluminum enclosure (non-anodized).  There is an (RP)SMA connector through the case which is to the antenna.  The environment in which the device can be used I expect to be capable of generating very high electrostatic noise (think fabric, polyester, clothes).  How should I (or not) connect the chassis to the circuit ground to prevent the circuit from being damaged, while retaining the highest signal fidelity during transmission?  Will the frequent charging and discharging from the nearby clothing/etc to the case have an adverse effect on the transmission fidelity as the entire circuit's reference is constantly moving around?  All transmitters and receivers in this system are mobile and isolated from earth.

It is a single board that slides into rails in the enclosure.  Currently I have both the side edges of the board with an exposed copper plane and stitched vias to the internal circuit ground plane for a low impedance path between circuit and chassis ground.  All other connections that might lead outside the chassis have ferrites and TVS diode clamps.

Is this setup naive/improper?  Please explain.  Thank you!

German_EE:
It sounds like you have done everything right regarding grounding, especially the non-anodized case and the exposed copper planes with lots of vias. Most of the connections are dealt with using ferrites and voltage clamps but your one remaining point of entry is that antenna connection. A 1M ohm resistor across the SMA connector will be enough to bleed off any static charge and will not affect the 50 ohm impedance signal by a measurable amount.

There are two ways of connecting the PCB to the chassis ground and one or both of them will work, only testing will find out which

Option One
Isolate your PCB from the chassis using insulated pillars (including the antenna connector) and have a single ground strap from the PCB to chassis. Maintaining this isolation will be a pain in the ass but you are guaranteed that there will be no circulating currents.

Option Two
Have a ground plane on the top of the PCB and a copper pour on the other side giving as much ground area as possible (especially around the edges of the PCB) Run multiple vias between the ground layers and ensure that when the PCB is slid into the enclosure there is a good solid contact at as many points as possible. The idea here is that ground is at the same potential no matter where you are in the assembly. This is my favoured solution and it has worked well in my various RF projects over the years.

selkathguy:
Glad to hear I'm on the right track at least.  I'll give it a shot with option two as that seems to be what I've got now.  If I have issues in testing I will try the single-point chassis bond.  Thank you for the tip about the 1meg across the SMA!  I might post pictures here when I get it all assembled.

XFDDesign:
I might be blind, but what frequency and power level are you running at? Depending on how high up you go, there are ultra-low capacitance 'suppression' diodes available that serve for input protection on RF Devices.

selkathguy:
~900Mhz, 250mW
It's a module.

Navigation

[0] Message Index

[#] Next page

There was an error while thanking
Thanking...
Go to full version