@Crusty1 , Agree with your point on electrical contractors, VFD's qualify as big outhouses! I always keep some 40mm E-cores in my kit so I can make CMT's on sensitive wiring (like encoders). Then I get them to put in equipotential conductors, while I rip up and reroute the signal wires and fix the grounding.
Apart from the leaky-feeder tea-strainer coupling mode which is basically electrostatic in origin, you have an additionally coupling mode due to the non-zero resistance of the shield. Assume a modest 100mohm for a length of coax, then 1 amp of RF coupled into the shield generates 100mV of potential that gets capacitively coupled to the centre conductor and adds 100mV of noise. That's +100dBuV compared to the GPS -30dBuV. So you really need to knock down the RF shield current to a miniscule 1uA.
How do you get RF current in the shield? Easy: make a short circuited loop by grounding both ends of a short piece of coax, and put a changing magnetic field inside the loop. Even if the ends are not physically connected, the capacitance between the GPS antenna groundplane and the chassis will close the loop at 1600MHz.
Apart from cooling everything to liquid helium temperatures, you can manage this mode by
adding thick equipotential ground cables, so the noise currents flow through the path of least resistance, NOT your shield
increasing the inductance of the coax (ferrite beads, clamshell ferrites or Common Mode Transformer (My favorite))
use a double screen cable or Triax so that noise current doesn't flow through the inner shield
judicious cable routing to avoid doing a lap around the outhouse
generous groundplanes and power planes to minimise the magnetic field generating volume
Use differential signalling if possible
Just because a conductor is at ground potential doesn't mean it can't radiate! Circulating ground currents easily couple into adjacent signal conductors.
@Sanjib.A : I tested a lot of leakeay feeders for use in underground tunnels (London Underground Tube), the shield was a continious copper sheath with repeating slots cut at the correct length for the RF frequency that would be transmitted and recieved by it, some even had multifrequency slots of different length.
So a coax or shielded cable, with a very loose weave, is in my opinion about as good as a tea strainer is for holding water and for the same reason. I always used a continious shield on instrumentation cable for this reason.
I have one bit of information to impart concerning installing screening, be aware that the Electrical installation engineer (for power) has a different idea about screeening(earthing) to that of the instrumentation engineer. I found this out after a lot of field work tracking down noise in systems that had been installed by electrical contractors.
@Crusty1: Yes, you are correct...I was trying to understand the technical reason why a solid shield cable would be better than a braided cable; the first thought that occurred to me is that for high frequency signal a poor quality braided cable would be a "leaky feeder"...would give rise to emission issues....what is your opinion on this?
@Daryl: Yes, together we will learn...and this is such a subject, one could continue to learn life long...or as you have correctly said in your other article - we would learn as we get more "battle scars" (failures) in EMC...I liked that phrase you used "battle scar"...very well said...it has become one of my favorites. :-)
I had a graphic illustration of this a few years ago. I was troubleshooting a customer's Sat TV installation. The dish (LNB) was connected to the receiver by about 30m of coax. The usual coax for this has a braid plus a foil shield, Water had got into the cable and, though the braid was continuous (a bit higher resistance than I would have liked though) the foil had been corroded into nothing. Replaced the cable and sealed it well at the LNB and all was well again. LNBs send to the Rx at around 1 GHz - and the power for the LNB goes the other way - so the cable needs to be good.
Agree with both of you! Even the best shield (coax or general cable shield) can be rendered ineffective at high frequencies by poor grounding/terminations. As an analogy, even the best garden hose can still leak at the faucets.
As I am fond of saying, cable terminations are such a common EMI problem that I put two kids through college repeatedly solving this problem!
@zeeglen: True!! The solid shield cables provides better performance over the braided shield cables at higher frequencies. At higher frequencies (smaller wavelengths) the braided shield does not provide effective shielding. Is it because the braided shield does not remain as a "low impedance" return path to ground for the high frequency signals? Also may be, because of smaller wavelength of the high frequency signal the braided shield becomes like a perforated shield for high frequency EM radiations and hence ineffective to contain all EM radiations? Or both?
Even if the coaxial cable is perfect, something that does not get perfect most of the time is "grounding"; An improper grounding would always give rise to ineffective shielding even if a perfect coaxial cable or any other shielded cables are used.
NASA's Orion Flight Software Production Systems Manager Darrel G. Raines joins Planet Analog Editor Steve Taranovich and Embedded.com Editor Max Maxfield to talk about embedded flight software used in Orion Spacecraft, part of NASA's Mars mission. Live radio show and live chat. Get your questions ready.
Brought to you by