PI coils- Shielded or Unshielded?

[waltr]

The problem i have is that the coil shell i have (ebay from bulgaria) does not allow me to use thicker insulation, as the bundle become to thick and is not fitting inside the housing. I'm not making coil for prospecting, i'm aiming only gold rings targets.

The purpose of the detector is important. Gold rings being never 24K actually have quite a low conductivity and are easily detected with sampling delays in excess of 25usec. No need for a "fast" coil.
Small Silver coins are the bigger challenge but are easier to detect than small gold nuggets.

For beach detecting, park, etc a coil from enamel wire works perfectly fine. Check the coils and target responses in my HH2 thread:
http://www.geotech1.com/forums/showt...ake-on-the-HH2

[ionut_mtb]

The purpose of the detector is important. Gold rings being never 24K actually have quite a low conductivity and are easily detected with sampling delays in excess of 25usec. No need for a "fast" coil.
Small Silver coins are the bigger challenge but are easier to detect than small gold nuggets.

For beach detecting, park, etc a coil from enamel wire works perfectly fine. Check the coils and target responses in my HH2 thread:
http://www.geotech1.com/forums/showt...ake-on-the-HH2

Thanks for answering Waltr. In this thread i ask about those deep signals of gold rings, what is your opinion? As the rings are not shallow in the sand. Do we need fast coil for those deep signals? I'm not so much into electronics theory so numbers and formulas are not my thing.

http://www.geotech1.com/forums/showt...&highlight=srf

[waltr]

Thanks for answering Waltr. In this thread i ask about those deep signals of gold rings, what is your opinion? As the rings are not shallow in the sand. Do we need fast coil for those deep signals? I'm not so much into electronics theory so numbers and formulas are not my thing.

http://www.geotech1.com/forums/showt...&highlight=srf

First a disclaimer: I have never been out detecting (just started into this this winter) and only have experimented in my shop. I have read just about everything and have also done much research.
So I use my 14K wedding ring as a quick test target and found it is the easiest object to detect and it also gives the biggest signal and the greatest distance. See my HH II thread- posts #14 & #24
http://www.geotech1.com/forums/showt...ake-on-the-HH2
There are two reasons for this:
1- 14K gold (58% Au, the rest mostly Cu) has a low conductivity. In my experiments I can set the delay way out- in excess of 30usec, and still get a strong response to this ring.
2- it is a RING which will have the Eddy currents going around the ring just like a coil (see the book "inside the metal detector" for this experiment). this coil seems to keep the eddy currents going.

Both of these seem to cause a long and strong response so a fast coil for beaches is not needed.
I have read (but no experience) that due to salts in sand at beaches one may need to increase the delay a bit. What I've read is even with a fast coil that can sample with an 8usec delay this doesn't work well on salt water beaches and recommended to increase delay to ~15usec.

The hardest target I have experimented with, shortest distance, is a US Silver Dime. This has a fairly high silver content, high conductivity, and is fairly small, low weight/mass.
For this target I find a faster coil with shorter sampling delay works best. Check the two post in my HH2 thread for charts of target responses.
I don't have a Silver ring to experiment with but am interested in how one would respond.

[waltr]

Update on some experimenting:

have had a few warmer days with unfrozen ground so out to detect around my house. Used the 8 inch concentric coil (see details in my thread).
This works very well and having NO response to the ground or even the red bricks in my front walk. I have found a lot of roofing nails and other small bits.

Bought a few small W. Australian gold nuggets (0.3 to 1.1 gram) for testing. These are seen by my detector but at reduced distance with the smaller sizes. This is expected since they are quite pyre, high conductivity, and small. I have a 10 Ohm resistor in series with the MOSFET switch and coil to shorten the coils Tau. As an experiment I decreased the value to 5 Ohm and tested to see if signal and range would increase which all targets test did. No response to mu Hand, Ferrite or a brick so this looked good.
Next day went into my yard and was getting lots of Falsing. One place in when sweeping the coil from the grass to the red brick walk. Then did the 'pumping' the coil toward and away from the ground. Got response when the coil approached the ground. Tried adjusting the GEB without effect.
Finally gave up and went to the shop and changes the series resistor back to 10 Ohm. All these problems went away and the detector worked nicely.

Since there has been different answers to the original question of this thread I think I now have another answer.
It seems if the Coil current is low and/or the coil Tau is short then coil shielding may not be needed. However, if coil current is higher then the detector may not work without a shield.

I need to continue this experiment, lower series R, and try a shielded verse unshielded coil.

[6666]

I need to continue this experiment, lower series R, and try a shielded verse unshielded coil.

Interesting results thanks for posting, especially with the Aussie nuggets, I just love finding those small bits,
with regards to your shielded verse unshielded coil. there is a very quick and dirty trick you can try
to shield your coil, simply run a ground wire down to the coil then wrap some very thin wire (.125 MM)
around the coil say 15-20 MM wide wraps hold on with a couple of bits of sticky tape and test.
Leave gap between start and finish.

There are also a few interesting remarks about adding resistors to coils in this old thread
http://www.geotech1.com/forums/showt...ighlight=cs6pi

[kosacid]

agreed i do the same thing rap a coil round the coil i used the same wire as the coil is made of, its that quiet when im close to metal objects i hear the change on the head phones and zero in on it, its the best shielding i have found

[waltr]

I will try a wire wrapped around the coil as a 'drain' (shield) on the next round of tests.
thanks for the pics and suggestion.
This should work well on the Mono and DD coils but will need something like Graphite on the Concentric coil.

Oh, I did try both the Concentric and Mono coil with the lower series R and they both acted the same.

The worse part is that I did not see issues in the shop but only when testing with Real Ground. Then again, a lot of people have stated that one must test on real ground.

6666,
Yea, would love to find nuggets but there are none in the USA North East.
I bought the ones I now have and they are from WA near Kalgoorlie.
Maybe some day I can get out to Nevada and find a few.

[dfbowers]

Update on some experimenting:

have had a few warmer days with unfrozen ground so out to detect around my house. Used the 8 inch concentric coil (see details in my thread).
This works very well and having NO response to the ground or even the red bricks in my front walk. I have found a lot of roofing nails and other small bits.

Bought a few small W. Australian gold nuggets (0.3 to 1.1 gram) for testing. These are seen by my detector but at reduced distance with the smaller sizes. This is expected since they are quite pyre, high conductivity, and small. I have a 10 Ohm resistor in series with the MOSFET switch and coil to shorten the coils Tau. As an experiment I decreased the value to 5 Ohm and tested to see if signal and range would increase which all targets test did. No response to mu Hand, Ferrite or a brick so this looked good.
Next day went into my yard and was getting lots of Falsing. One place in when sweeping the coil from the grass to the red brick walk. Then did the 'pumping' the coil toward and away from the ground. Got response when the coil approached the ground. Tried adjusting the GEB without effect.
Finally gave up and went to the shop and changes the series resistor back to 10 Ohm. All these problems went away and the detector worked nicely.

Since there has been different answers to the original question of this thread I think I now have another answer.
It seems if the Coil current is low and/or the coil Tau is short then coil shielding may not be needed. However, if coil current is higher then the detector may not work without a shield.

I need to continue this experiment, lower series R, and try a shielded verse unshielded coil.

Hi Walter. I have been following your posts for some time. In reference to your series coil resistor.. My Geotech Baracuda for example works perfect in salty beach environments but is not very useful in Pennsylvania dirt because of ground response, even at a 15uS delay, but the addition on a 10 or even 15 ohm series resistor really takes the edge off of ground response with minimal loss of depth on most all targets in different conductivity ranges. This is something I cannot accomplish with decreasing threashold (because of major loss of depth) or increasing sample delay (loss of low conductors). At least for the Baracuda project this seems to be a very simple but effective solution to ground resposne. I cannot however eliminate fired red brick.
Thanks.
Don

[waltr]

Thanks for info on your experience.

I am suspecting the issue is due to the coil/switch Tau (increasing R decreases the Tau) allowing the coil current to level out before switching off.
Without the series resistor the Tau is high and coil current is still increasing when switched off. Taking a wild guess that this is causing the greater Ground response and causing electrostatic issues.

There are two different possible issues. I have not determined which one or combination of these two issue I was seeing with a lower series R value:
1- coil shielding is to prevent electrostatic potential from causing false response. But why this does not cause an issue if a series R is added without a shield?
2- GEB- not having a response to mineralized ground.
Maybe these are more related than I know but it seems from reading these forums that they are not the same and have different fixes.

I did find when originally testing the that target distance response (air testing in the shop) did not decrease noticeably when I added the 10 Ohm series resistor and actually helped the circuits stability.
Also, I did measure the Coil's current rise and with 10 Ohms could get 80% of maximum current with a 100usec pulse. Whereas without the series R (400uH/2Ohm) the TX pulse needed to be over 600usec to obtain 80% of maximum current.

I still have not been able to prove the three sample GEB scheme works to eliminate ground. It does however allow for a Low tone on high conductive targets and a High tone on iron and low conductive targets like the White's TDI in All Metal mode.

[Chiv]

.

There are two different possible issues. I have not determined which one or combination of these two issue I was seeing with a lower series R value:
1- coil shielding is to prevent electrostatic potential from causing false response. But why this does not cause an issue if a series R is added without a shield?
2- GEB- not having a response to mineralized ground. .

I have added a 10ohm resister in series to a coil tested it on the beach same depth but coil still falsed when changed swing direction.
so back to the drawing board and added a shield to the outer edge of the coil (but not to top side or bottom of coil) but in side of the housing using some type of ally tape, and retested on the beach last night and got no falsing at all very happy, still getting same depth and even hit some iron nodules that I did not even picked up before(very faint sig at first nearly did not check it out).and have some of the hardest beaches for testing on, full of mineral sands
now just need to build a faster coil with same set up.

[surfdetector]

agreed i do the same thing rap a coil round the coil i used the same wire as the coil is made of, its that quiet when im close to metal objects i hear the change on the head phones and zero in on it, its the best shielding i have found
[ATTACH]42315[/ATTACH]

I had always understood that it is best practice to install a spacer in between the coil and the shield, I am wondering if your method of using a shield consisting of a wire wrap does not require the spacer or perhaps if it would work even better with a spacer?
Cheers

[Elliot]

Eric is right on with his comment about shielding the coil I would like to add some additional information based on my own coil design research.

1. As the delay gets lower, between 5uS and 10uS, shielding material and shielding techniques become more important. Just choose a shielding material that is not detected or very minimally detected at your lowest delay setting. This includes how you wrap the shield around the coil. Try to avoid making conductive shielding metal to metal contact around the circumference of the coil cross section wire bundle. If you must overlap put a tape insulation to keep from forming a conductive loop.

2. Use wire insulation that has the lowest dielectric constant to minimize coil turn-to-turn capacitance. Thicker insulation causes less capacitance but lowers the coil inductance slightly so you may need to add an additional coil turn or two to get the desired coil inductance.

3. Use a low dielectric spacer between the coil wire and the shield to minimize the coil-to-shield capacitance.

4. When operating at very low delays the wire thickness may be holding eddy currents. Try to use single strands no thicker than AWG 30 or use marine tin plated stranded wire. The tin plating puts a slight resistance between strands and tends to minimize the formation of eddy currents in the wire itself.

5. Use the lowest capacitance coax wire to connect the coil.

6. Make sure that the peak fly back voltage is below the rated MOSFET voltage to prevent clamping and speed up the potential receive time.

7. Use a two stage low gain first and second amplification stage to minimize the amplifier lock up time.

Here is a good point to help you form a simple mental model about PI detectors and delay time. It takes 5 time constants for a targets fully stimulated eddy currents to decline to near 0%. If you have a 2uS small target and attempt to detect it at 10 uS, there will be no signal left to detect, as during the delay, all the energy in the target has dissipated. If you have a PI detector that can go down to 7.5 uS delay you may be able to pick up the tail end of the dissipating eddy currents. Thus smaller targets need lower delays to detect. But there is more.... lower delays pick up more ground effects also. Good shielding is more critical at lower delays!!!

Now here is is where it gets more complex. Theory tells us that a target must be stimulated 5 times faster than its own time constant.

Let’s use our 2uS example from above that needs to have a 0.4uS coil discharge pulse time constant. The coil discharge time constant is calculated by dividing the damping resistor value into the coil inductance. A typical 300uH coil would need a 750 ohm damping resistor to fully stimulate that 2uS target. Now here is where it all comes together. Less coil, cable and TX circuit capacitance allows higher damping resistor value to be used. Thus the term “fast coil”.

Eric would you please add anything that I have missed or rank the importance of the points I have attempted to explain.

Thanks

Joseph J. Rogowski

Hi Joseph, I have read in your various posts (example above) your explanation about the need to stimulate a target five times faster that the target's TC in order to fully charge it. You then refer to the need of the TC of the coil decay cycle to meet this requirement. May I suggest that it is the coil charge cycle and peak back-EMF change in polarity that does the target charging, not the coil decay cycle? Any thoughts?

[ZED]

Hello Elliot

I may have miss-understood your post but if your suggesting the high voltage back EMF is what is inducing eddy currents in a target then your mistaken,its the coil currents rise and fall and its rate of change that is creating the eddy current,the fly back is merely an artifact within the confines of the coil and the TX cct.

[Elliot]

Hi Zed


As I understand it ,the current in the coil gradually ramps up (coil charge cycle) during the pulse-on time (say 25-65uS). As soon as the current source is switched off at around coil Imax (i.e. when the coil voltage is low to zero), the back EMF rising voltage in the coil response kicks in and reaches a peak (say 400V for example). It is at the point of the quick change in coil voltage polarity at the peak of the back EMF when the eddy current in created in the target, which affects the decay time of the coil. From what I understand reading several of Joseph's text (and I may have read it wrongly), he is inferring that it is the decay part of the coil's signal that is charging the target, which I don't agree with. Now, if I am wrong, at what other point in time is the target (not the coil) charged, and how can you charge the target five times faster than the target's TC? Don't you just love science!

[ZED]

G'day Elliot

Eddys build on a target during the TX period,after the TX period is terminated the current ramps down,now depending on the TX cct and coil the eddys created during
the ramp down will cancel eddys from the ontime and then build eddys again with the remainder of the coil current as it continues to ramps down,it all depends on the
TX cct as a whole and the target.
When the coil current stops increasing and flattens out the ontime eddys will decay away because there is no longer a rate of change with the coil current,its static "aka DC",so
there will be no eddys to cancel as the TX terminates and the current begins to ramp down,on the other hand if the TX period terminates while the coil current is still rising then the ontime eddys
will in part be canceled out by the eddys that are created as the coil current ramps down,and if your going to ask why it will be because the ontime eddys have a opposite polarity to the eddys that are
generated during the current ramp down period,and so there will be some cancellation of eddys during the ramp down period ,now this is just a basic description at best,there are a lot of other factors at play as well.
Its not uncommon to see many references to the coil current ramp down period as "the flyback",typically they both happen at the same time,i even use the phrase as well,its convenient,but !, the very loose
use of the phrase can be very misleading,and over a period of time it becomes the norm.

Now you said quote "how can you charge the target five times faster than the target's TC? "thats a slippery question,what does it mean,are we talking time,or amplitude,or both.A targets eddys will decay
at the same rate regardless of how strong the EMF and rate of change is,and why do you want to charge it faster,whats the gain ?
I dont see the point of setting a TX CCT (including coil) up for a particular target TC when like here in VIC targets can be as small as lead shot or as big as a can,its a science for sure but no one shoe
fits all.Coil current thats flat top at switch off has its place and coil current still climbing at switch off has its place as well,neither one dominates over the other.


I have attached 3 pics of scope shots showing the TX waveform and the target signal as seen by the receive coil,using a DD 11" coil,the TX winding is 0.4 ohms,on one scope shot i have used a 10 ohm series
resistor in circuit so the coil current flat tops before switch off.
If you have a keen eye you will notice the 1.5 ounce gold nugget still has on time eddys quit late in its decay

I hope im not being a nuisance,i kinda feel like im hijacking the thread.