Shielding Woes

[Dai]

Hello all,
First, a thank you to Carl for making this site possable and to all who have contributed to this wealth of knowlage...

I have built the Surf PI kit from Silverdog, works fine, its a nice little kit. I have made a coil and have had it sampling reliably to 5.5uS and detecting a 1 euro coin at 40cm in the air. Slight noise detection, as expected in the garage with energy saving lights nearby. Hand not really detected in proximity to coil but slighty noticeable when I touch coil. All good and well untill I put shielding in place. Im using a graphite/varnish mix, 8mm gap on coil shield, spacing between coil and shield is approx 8-10mm. Shield to coil capacitance is 80pF, damping resistor 402 ohms. Surface of shield at the moment is about 4K ohm measured about an inch anywere on coil. The problem I have is the shielding has made things worse, my hand is now detected as if it were a lump of metal and if I put my hand on coil and then remove hand there is a huge signal as if a chunk of metal is rapidly moved towards coil. On the plus side I can open the threshold a lilttle more, but not much. Also if i reduce the delay to about 15uS (I havent measured this to confirm exactly), the effect of touching the coil is drasticaly reduced but my hand is still detected when waved in front of coil.

Any one have an idea why this is happening? Is it a case that I need to reduce the shield resistance quite a bit or have I just killed my perfectly good coil by wanting that bit more from it?

Dave

[baum7154]

Sounds to me like you have followed good shielding procedure. Did you get 4K ohms in just one coat of graphite/varnish? If multiple coats did you lightly sand with very fine sand paper between coats? Are you sure you grounded the shield? Have you also re-done critical damping or is the 402 ohms the value from the unshielded coil?

Regards,

Dan

[baum7154]

Please make sure you attached ground to only one side of the gap.

Dan

[Dai]

Hi Dan,
I applied three coats, sanding between coats, although the surface resistance was pretty much the same before and after sanding, the resistance did drop with the second and third coat,however I would have expected it to drop more than it did,a test on paper (sample of shield compound brushed on paper and allowed to dry) gave a surface resistance of around 500 ohms. The paper does have a much smoother flatter surface though.
Yes the ground wire (unfluxed desolder braid, slightlysplayed) is attached one side only. Yes the 402 ohm damping resistance was before shielding, I have adjusted this with my decade box but have not yet measured it and forgot to mention it in previous post.

Regards
Dave

[baum7154]

It sounds like you may have a bad ground connection. I would measure from the end of your coax if you used it to the contact point on the shield and then around the shield. Is your coil a bunch wound coil or some other type? What sampling delay are you running now?
Your damping resistor value should have gone lower with the shield can you measure it?

[Dai]

Oh,and just a side note, before shielding, the coil would easly detect the titanium plate in my wrist at 35-40cm, now it is detected at amlost half that. The coil is still better than my first attempt-this wouldnt detect my wrist plate at all.

[baum7154]

Yes, I have seen shielding reduce the sensitivity of the coil at values in the range of 1500 ohms per inch of spacing. The shielding value will also change somewhat as your varnish cures over the next week or weeks. How long ago was it applied? With the added capacitance of the shield it is likely that your coil will no longer operate at a delay of 5.5us. Do you know the self resonant frequency and inductance of your coil?

Dan

[Dai]

Thats what I thought and was one of the first things i tested,yes I am using coax, the ground connetcion seems fine, less than 300 mOhm from pcb to drain wire on coil, contact point is about 100mm long, embeded in shield compound, resistance increaseing from ground side wire to end of shield (other side of gap) as would be expected with a linear resistance track. Coil is bunch wound. I can still sample down to 5.5uS but at this level it detects any movment of my hand/body within almost half a meter of it. It seems like the shield is acting like a tunned antenna... Yes the damping resistor value will be lower, have not measured it yet, I have a fixed resistor on pcb, and the balance in the coil plug, I use the decade box to tweak to suit then substitute accordingly.

[Dai]

Yes, I didnt want to drop the shield resistance too much to start with and thought that 4K would be a good time to try it. The varnish was applied over a week ago and has been sun bathing the last few days... The coil (windings) is about 250mm internal Dia, 305uH 1.3 Ohm and WAS about 800kHz before shield, havent tested since, Im waiting for a new sweep gen, my old ancient sig gen is to unreliable and unstable...

[baum7154]

If you have 4" of copper desolder braid in your shield I suspect that your detector will see that slow decay time copper @ a setting of 5.5us. You will have to set the delay further out in order to ignore the copper. Did you try the 4" copper braid to see if the detector sees it before your put it in the shield? Rule of thumb " if your det sees the shield or contacts, don't use it if you need to run at the lowest delays"

Dan

[Dai]

I have witnessed the coil cut of point on my first test when I built the kit,dead-no response. This can still sample down to 5.5uS although the hand effect is reduced if i increase delay, say, 8uS upwards, but will still detect my hand with out problem. Enclosing my hand around coil with delay approching around 20uS just reduces sensitivity of coil (simmilar to puting coil near large metal plate). I dont get the large sudden signal spike at the higher delay like i do below 12-15uS.

Dave

[Dai]

Yes, it wasnt detected, neither was the spool of it, infact it seems to have a reduced sensitivity to a 500g reel of copper...

[Dai]

What I dont understand is why now, with shield in place, it can detect my hand (withouth the plate in) as if it were a chunk of steel,whereas before I could rub,shake touch the coil with hardly a murmer other than electrical noise from the lights etc.

[Dai]

Im just off for a ciggarette,Il put the megger on the coil, just to check insulation between coil and shield..make sure there no solvent abuse been occuring...

[Dai]

Insulation is fine..