Electrical shielding newbie question

That is one purpose for shielding. Shielding deals with local magnetic and electric fields and emission of radiated waves.

Sometimes you contain the source of noise with a shield, and sometimes you are not in control of the source of the field, so you contain the circuits subject to disturbance from fields. Sometimes you do both.

Changing electric fields cause displacement currents that connect the object to the voltage source. Changing magnetic fields cause local eddy currents in conducting loops. Either cause of current will cause voltage drops in conductive paths.

The way I see it, the shield is an extension of the box that contains the converter, so that it also contains the sensor and its wiring. It should be separated from all wiring, except for a connection at one point. In this case, I would pick a point very close to he converter. That way, all currents that slosh around in the shield conductor in response to electric or magnetic fields produce little voltage drop in any of the signal conductors, including the signal common.

I don't totally understand your last paragraph. In our case, our box has signal conditioning for a 0-3V analog sensor input. We provide with the ground. In other words, the A/D converter is in our plastic box and the sensor is in the other one. If I understand well, you would make one single connection of the shield with the ground, close to the converter, in our box? Every book I have read mentions it has to be close to the source of voltage, hence the sensor....

I know that you either shield from electric and/or magnetic disturbances. In our case, I'm more concerned about electric disturbances, as the signal frequencies are in the order of 1 kHz. The sensor will have little disturbances from external sources as there won't be transformers and such close by.

In that perpective, I don't understand how such a shield would protect against electrical disturbances if I have a plastic box. How would it change the electrical fields then? Where are currents flowing with and without that shield? I wish I could draw what I see in my head... ;)

Thank you,

Yvan

What, electrically, do you mean by, "we provide with the ground"?

I generally connect the shield to the signal common at the point where I care about the signal integrity the most. I would wrap that shield around as much of the wiring and system as needed for it to be the capacitor plate that receives change displacement form external electric fields. If capacitive currents are dumped through the shield, it will pass these currents to signal common at the connection point, and that may bounce the potential at that point around (and all the signals, with it). But if all signals and signal common share the same noise at that point, the circuits that responds to the difference between signal potential and signal common potential at that point all share the same noise addition, the differences between them will show little of that noise.

Another source of magnetic interference is power lines running parallel to your signal lines.

Feel free to email any sketch you may have to me. If I have any useful comments, I will post them along with your graphic on the newsgroup alt.binaries.schematics.electronic, as well as by return email.

Ok thanx, I will come back to you shoftly.

Yvan

Ok thanx, I will come back to you shortly.

Yvan