PCB Coil?

[bugwhiskers]

Low Capacitance diode

In an earlier post I spoke of a low capacitance diode, the following link has the data sheet. Notice it's capacitance is 1pF compared to IN914 which is 4 pF.

http://www.dse.com.au/isroot/dse/support/Z3040.pdf

regards
bugwhiskers

[Tinkerer]

My interpretation of this, is that the decay curve is made up of many different frequencies.
My next goal is to try to find the source of each of these frequencies.


Did you see my post titled "Are we going round in circles ?"

Yes

If the pulse repetition rate is set to say 500Hz (close to the ears most sensitive frequency (human voice)) and "wanted" frequencies (good targets) are boosted then the ear would hear the differences between targets and learn what various targets sound like.

Just need someone with a Spectrum Analyzer to find the frequencies of interest and a few filters (gyrators).

Imagine a machine with a graphic equalizer type control panel where you dial in "wanted targets".

I agree, that if we translate the decay curve that we see on the scope into a corresponding audio frequency, it would give to the trained ear, information on the target.
I have a 50MHz FFT.
Sorry my ignorance, what is a gyrator?


Remember the trick bbsailor told us to test a coils TC, ie putting the test coil near another coil while it is pulsing and monitoring the test coil with a CRO.

I wonder if either your coil is under-damped or if it is responding to a very strong early eddy current causing it to self resonate briefly??

I refer to the coil when it is not connected to anything else than the scope probes.

In an earlier post I spoke of a low capacitance diode, the following link has the data sheet. Notice it's capacitance is 1pF compared to IN914 which is 4 pF.

Thanks for the link to the diodes.

Tinkerer

[bugwhiskers]

Response to Tinkerer

Hi Tinkerer,
Please don't apologise for ignorance, we are all learning and any who say they aren't are very careless with the truth.

A gyrator is simply a filter that doesn't use coils (capacitors instead) smaller and easier to understand.

In the light of your reply about your coils favourite frequency, bbsailor to the rescue !

regards
bugwhiskers

[bugwhiskers]

Further thoughts

I agree, that if we translate the decay curve that we see on the scope into a corresponding audio frequency, it would give to the trained ear, information on the target.




Tinkerer

Not a corresponding audio frequency !

Middle "C" played on a piano sounds very different to Middle "C" on a clarinet or a violin or a set of bagpipes (heaven forbid, an ill wind that nobody blows good).

It's the tonal qualities of the decay curve that the filters need to enhance so that even the most tin eared detectorist can decipher.

regards
bugwhiskers

[bbsailor]

extracting more information from a target

bugwhiskers and tinkerer

What you are trying to do is extract enough additional information from an unknown target to attempt to make it known, using known characteristics of a target and known characteristics of physics. Go to the link below:

http://geotech.thunting.com/pages/metdet/patents/US7075304.pdf

This patent explores the known difference in target deacy time by adjusting the current di/dt time, using a variable damping resistor, to obtain different responses from an unknown target and comparing those responses to known responses stored in a target library all in an attempt to make the target known.

While the patent inventor, Carl V. Nelson suggests using a MOSFET to replace the damping resistor (Rd) across the RX and/or TX coil(s), the general physics issue that is being explored is using a variable damping resistor to change current decay time of the coil as it induces different amounts of eddy currents in targets as well as tuning the RX damping resistor of the coil to optimize it's bandwidth for a known type of target.

This patent has some good background information that would help you with your expriments.

bbsailor

[bugwhiskers]

New coil figures

Re the double sided PCB coil that is shown on page 3 of this forum.

The tracks on each side were originally in series which produced 6.8 ohms DC resistance, 310 uH Inductance and a resonant frequency of 571.4kHz.

Changing the windings to parallel has caused the DC resistance to drop to 1.7 Ohms and the resonant frequency in now 1.538 MHz and the Inductance is ~ 70 uH. (my inductance meter shows 0.07mH)

regards

bugwhiskers

[bbsailor]

bugwhiskers,

571.4 KHz for a 310 uH is a very high capacitance coil (as you said).

Here is a good way to plot this coil or any coil to better visualize where on the curve your particular coil falls. This would be helpful to anticipate the self resonance impact of making small changes in capacitance. Try this, as it will be very educational.

Make a graph placing capacitance in pF on the left, vertical axis in 10 pF increments from 0 to 300 pF. On the lower, horizontal axis place self resonance frequencies from 0 to 3 MHz in 100KHz increments. Then go to your favorite on-line LC resonance calculator (or this one http://www3.telus.net/chemelec/Calcu...Calculator.htm) and plot the various resonances for your coil's inductance starting on the upper right side with 300 pF moving down in 10 pF increments. Label that curve the inductance of your coil. If you make multiple coils, you can place then on the same chart in different colors. The curves will all have about the same shape but the curve representing coils with more inductance will move toward the left. What you will see is something that looks like an exponential curve but more importantly it will show you where on the curve your particular coil falls.

My 309 uH coil resonates at 1.25Mhz and has 51 pF so it is below the middle of the curve where the slope is less than 45 degrees. Dropping the capacitance by 10 pF increases the resonance by about 150 KHz. With your 310 uH coil, resonating at 571.4 KHz would be near the top of the curve at 250 pF where the curve is almost vertical so it would take about a 50pF change to alter the resonance by 100KHz. As you will see, capacitance as it applies to coil self resonance, is not linear.

As you design coils that operate farther out toward the right side of the curve, even small reductions in capacitance have a very measurable impact on resonant frequency. Helping to visualize this is very helpful when you want to see where to put your efforts. My own experience is that making reductions in capacitance gets harder as you move away from the center of the curve toward the right.

Thanks for your coil data input.

I hope this helps with your coil experiments?

bbsailor

[bugwhiskers]

Hybrid coil

Hi bbsailor,

Will do that chart and post the results here, obviously I am after the highest inductance with lowest capacitance.

Attached is an idea I had for a hybrid coil. Radial wirewound for transmit and a thin sparse spiral for the receive in the middle.

I can hear you thinking....why not leave space in the middle of the rx coil pcb for a pre-amp !

regards
bugwhiskers

[Tinkerer]

Good idea, if you put the PCB trace on the inside, then the thickness of the PCB serves as the spacer for the shield. I recommend that you measure your coils with shielding included. Without a shield you can not get much sensitivity.
Tinkerer

[bbsailor]

Think Vertical

bugwhiskers

Most coil housings are about 13 to 14 mm thick. Rather that spiral wind a coil where the inner windings contribute less inductance than the outer windings, think vertical where all the windings are the same diameter.

It turns out that 21 turns of AWG 30 Teflon insulated wire, with a .024" OD, will fit 21 turns in 13mm, vertically stacked, and will make a 297uH inductance coil. This coil will have 5.88 ohms of resistance. You can use what ever wire you have on hand. Using magnet wire will give you a coil that is a little shorter, needing only about 19 turns for about 300uH or could have almost 35 turns in the same space for a higher inductance about 1000uH.

You can be creative about how you could make this but here is what I would do to experiment.

Cut a piece of flexible plastic that can be easily bent into a circle that is 275mm OD. Make this piece about 15mm wide by 864mm long plus 4 times the thickness of the plastic.

Cut the ends at a 45 degree angle so the plastic can be epoxied into a ring with a wider surface area than a simple butt joint.

Drill two small holes equal to the wire OD a little less than 1 mm from each edge for the start and end coil winding.

Coat this coil form with a very thin coat of rubber cement. Allow the rubber cememt to get tacky.

Feed about 150mm or wire through the bottom hole.

Place the plastic ring on a flat surface.

Carefully rotate the ring while feeding the wire around the plastic coil form. Here you may need to get a little creative and make a wood center wheel to hold the plastic coil form steady while rotating.

Rotate 21 turns of Teflon insulated wire (or 35 turns of magnet wire) on this coil form and the feed the other end of the coil into the other hole. The rubber cement should hold the wire in place for a while.

Wind electrical tape over this coil by going around the coil form holding the tape on an angle to minimize the overlap.

You will get a coil that resonates somewhere between about 1.5MHz and 2MHz (depending on the wire type and number or turns) and one that is very sensitive to the dielectric constant of the plastic material chosen to make the coil form and the tape. Polyethylene has a lower dielectric constant than PVC. This unshielded coil will be very sensitive to the capacitance of things around it.

The real benefit of this coil winding techniques is that you can make coils with more turns than the conventional bundle wound coils. It is these extra turns that makes the coil more sensitive without raising the coil capacitance too much. However, when you add shielding and the coil lead, and hook it up the the PI circuit, the final performance of the coil starts moving back to the center of the curve that I mentioned in my previous post.

Experiment with making coils that operate on both ends of the curve so you can learn where the coil performance tradeoffs are located. You will discover a practical spot near the center zone of the curve that will be where most of you finished coils will fall and where you will get a predictable performance.

Going from a coaxial cable (coil lead) that is 7 ft long, to a coaxial cable that is 2.7 ft long, will have a greater impact than the most exotic coil winding technique.

The real benefit of attempting to make a fast coil lies in the TC of the current fall time where L/Rd defines this time constant. So if you want to detect small gold targets that have a decay time of 5us, you need a coil with about 1us fall time TC when the coil is connected to the PI circuit. Here is where the speed of the coil contributes with other circuit factors that governs the sensitivity of you coil to the desired targets.

As the damping resistor value is chosen to crtitically dampen all the capacitance in the coil circuit, you will find that Rd is functionally in parallel with the op amp input resistor (Rin) even in a DD design, while the clamping diodes are conducting. 300uH/1000 ohms = .3us but if Rin is 1000 ohms then the formula becomes 300uH/500 ohms = .6us. Look at the circuit component values that could slightly extend the coil current fall time TC. If the combined effective value of Rd and Rin were 300 ohms, you can see how easily you may have a circumstances where a desired 5us decay target may not be detected. When you start to want to detect targets with a 3us decay time, you can see how things can become a real balancing act when Rd is 1000 ohms and Rin is 1000 ohms; when 5 times the current decay TC time of .6us is 3us.

It is just my opining that its this area of detecting small things, with low decays, where the real exotic coil winding techniques may become useful. I am mainly an ocean beach hunter so I must defer to those who seek small nuggets to do the reverse engineering necessary to build the type of coil they need for their targets in their soil environment.

I would be interested to hear the results of bugwhiskers and tinkerer coil designs to add to my knowledge.


bbsailor

[Tinkerer]

Bbsailor,

Now how do you name this vertical wound coil? I can see some definite advantages there. The surface is small so the coil to shield capacitance will be small also. Easy to fit into a standard coil shell too.
A spiral coil is definitely at a disadvantage there.
A spiral coil seems to be very good for pinpointing. I wonder how different the magnetic fields of the two types of coils would look if we could see them?
About short first delays: I see that Eric is testing a 5 micro second Pi on the beach. Thin gold chains and other jewelry are often lost on the beach. I can see the advantage there. Now here are my problems:
I spent a long time to get a coil that lets me sample at 5uS. Once that I finally got there, I found that I can detect a fast decaying sample (0.6gram gold nugget) at 25 uS, if I give the coil amp a high gain.
So why I am still looking for a fast coil? Because the signal is so much stronger at the short sampling time. That makes the detector more sensitive. Thus it should be able to find deeper stuff better too.
OK, I hear you, the high gain will saturate the circuit with any significant mineralization or even salt.
Yeah, I found that out 20 years ago when I used the detectors on ocean shipwrecks.
So it took me a while tinkering around this problem. But I now seem to have it solved. It seems that my ground balance takes care of the worst mineralization. So I can use high gain without saturating.
But, still another problem: my health does not let me go to the beach anymore to try it out. So here is my question: Would you build a detector if I send you my schematic; and test it on the beach for me? It is not a very complicated design, but it is quite different.
Tinkerer

[bugwhiskers]

Coke can test

Hi bbsailor,

It would seem the best combination for a hybrid coil would be a Helical coil (best inductance/capacitance ratio) coupled with a spiral PCB coil in the middle (best for pinpointing). I hope to make one within the next week.

As for the flexible plastic, I am going to find an offcut of 12" OD PVC pipe to be used as the former for winding only (later potted with the PCB coil in epoxy.

Today I did the coke can test (aluminium) and my PIMD with the 300mm*150mm PCB coil detects at just over 500mm. A 2 Euro coin is picked up at just over 250mm.

regards
bugwhiskers

[bbsailor]

Tinkerer,

Now how do you name this vertical wound coil?
It is called a helical coil.

A spiral coil seems to be very good for pinpointing.
A ferrite core probe is also good for pinpointing, espicially in a deep hole.

I wonder how different the magnetic fields of the two types of coils would look if we could see them?
I believe that a helical coil will go a little deeper because the windings are in line and do not tend to cancel as the spirals get closer to the center.

About short first delays: I see that Eric is testing a 5 micro second Pi on the beach. Thin gold chains and other jewelry are often lost on the beach. I can see the advantage there.
That post series is from last year, 2005. His 5 us machine is operating at 20,000 PPS.

Now here are my problems:
I spent a long time to get a coil that lets me sample at 5uS. Once that I finally got there, I found that I can detect a fast decaying sample (0.6gram gold nugget) at 25 uS, if I give the coil amp a high gain.

I would think that the signal for that 0.06 gram nugget would have a much shorter decay time.

So why I am still looking for a fast coil? Because the signal is so much stronger at the short sampling time. That makes the detector more sensitive. Thus it should be able to find deeper stuff better too.

I agree. But it is not just a fast coil, it is a combination of a fast coil, wide bandwidth, lower gain first RX amplifier and lower capacitance MOSFET operating at about 10 to 12 Volts. His 5 us coil is 344 uH, 3.8 ohms and has a 555 KHz self resonance (including the coax coil lead) with a 20 us TX pulse width.

OK, I hear you, the high gain will saturate the circuit with any significant mineralization or even salt.

If the machine were adjustable between 5 to 15 us it could accomodate a wide range of bach conditions.

Yeah, I found that out 20 years ago when I used the detectors on ocean shipwrecks.
So it took me a while tinkering around this problem. But I now seem to have it solved. It seems that my ground balance takes care of the worst mineralization. So I can use high gain without saturating.

Can you post your schematic on this forum?

So I can use high gain without saturating.

Sounds like the Pulse Devil which processes part of the signal during the TX pulse and uses a special balanced coil.

But, still another problem: my health does not let me go to the beach anymore to try it out. So here is my question: Would you build a detector if I send you my schematic; and test it on the beach for me? It is not a very complicated design, but it is quite different.

I would like to see your design but my plate is full until the Spring of 2007.

bbsailor

[bugwhiskers]

Tinkerer,

Sounds like the Pulse Devil which processes part of the signal during the TX pulse and uses a special balanced coil.

bbsailor

Could you please explain how this is done bbsailor. I tried monitoring the current curve for the TX pulse with a .39 ohms resistor in series but even with a large steel target up close I didn't discern any change.

regards
bugwhiskers

[bbsailor]

Bugwhiskers,

I wish I could explain it, but I am repeating what I rember from Dave Emery's posts either on the Geotech forums or on Eric Fosters PI forum. Dave just made some updates about his new design on http://groups.yahoo.com/group/ProspectinginOz/

What I do remember is that he used a balanced coil to implement his design. It seems that he is about ready to release it for early user testing. He may even be close to early first run production units.

Sorry, I can't give you more details.

bbsailor