Coil shielding problems

[Polymer]

Finding this old post from 2004 solved my warbling problem with the Surf PI.

THANK YOU!

Thought it might be useful for others, so "bumping it up" or how you call it to the year 2017!

Here's the thread with my original question:

http://www.geotech1.com/forums/showt...867#post218867

Hi Jackdetect and All,

It is perhaps incorrect to describe shielding as cancelling noise. It is better to refer to shielding as attenuating noise. We also need to consider what frequency noise we hope to deal with by this method. As Reg mentioned, the front end of a PI has to be broad band. Not necessarily down to d.c., but it could be from a few hundred Hz, up to 100kHz or more, depending on how fast a decay curve we want to look at. This means that any coil shielding must be transparent within that frequency band. The aluminium shielding referred to in the first post, obviously was attenuating frequencies at 50kHz as it was increasing the useable delay to 20uS or more. In this case it was impeding the higher frequency components of the TX pulse from getting out. If the shielding starts attenuating at 100kHz, then for higher frequencies, there will be ever greater levels of attenuation, depending on the skin depth characteristic of the particular material used, and how thick it is. Aluminium or copper tape could be used, but it would have to be very thin, more like the aluminium film deposited on Mylar, for decorative material.

Even when the shielding is giving high levels of attenuation, it depends how near the source of the interference is, as to whether there is any effect. i.e if the shielding reduces the interference amplitude by a factor of 100, but the source is brought nearer, so that there is 100 fold increase in signal, you are no better off. R.f. interference has two effects on a PI detector. One is a beat note effect, such as experienced in here in England with the high power 200kHz transmitter at Rugby. A PI detector running at a pulse frequency which is a close sub multiple of 200kHz will experience beat interference. e.g. 200Hz, 2kHz, 20kHz. A low frequency warble of the audio will be the result. That is why a TX frequency control is very useful, as a small adjustment can increase the frequency of the beat note to the point where the integrator smooths out any residual effect. Obviously any reduction in the interference amplitude as seen by the receiver input, helps enormously. You can certainly see the reduction in the 200kHz signal when a coil is shielded, as compared with an unshielded one.

The other type of r.f. interference results from much higher frequencies, even to 100’s of MHz. This is usually from very close range devices such as mobile phones, VHF transceivers etc. This results from the small amount of signal that gets through the shield and which is modulated with audio. The input circuitry of the detector, with its protection diodes, acts as a demodulator and presents the receiver with audio frequencies which are within its bandwidth. Even just switching a transceiver on without modulation can cause a dc offset at the receiver with a resulting bleep in the detector audio before the differential integrator cancels it. Again, any reduction rf pickup by shielding the coil is valuable here. Additional rf filtering is often used on industrial PI detectors in the form of ferrite sleeves and LC filters.

The effects of coil shielding is best observed on a spectrum analyser, and believe me, it does make a difference, particularly if you want to comply with EU rf emissions and immunity standards.

Low frequency interference within the passband of the detector is not reduced by coil shielding. It is then that you have to resort to figure of 8 or differential receiver coils in some situations.

Eric.