THE ELUSVE 1us TC TARGET

[Tinkerer]

To be able to detect a target with a TC of 1us (one microsecond time constant), with a PI detector, we need to take a very early first sample.
A TC of 1us means that after 5us the eddy currents are gone.
With a delay of 3us we should have a chance to still detect the 1us target.

This picture shows the response of a square of kitchen aluminium foil of 25x25mm on the coil.
The yellow trace is the Flyback.
The green and white traces are the output of the preamp, no target and with the target.
The cursor does not show on this picture, but we can see from the indicator at the bottom, 3.485us, that we could start the sample at this time.

This is done on a bread board and with a half finished DD coil with a sweep diameter of about 37cm. Please take that into considerations when you see the noise.

I would much appreciate your input for trying to improve the setup. The idea is to start a discussion and to try all good ideas to improve the results.

[Tinkerer]

This is the preamp used for the scope shot above.

[ivconic]

[bbsailor]

To be able to detect a target with a TC of 1us (one microsecond time constant), with a PI detector, we need to take a very early first sample.
A TC of 1us means that after 5us the eddy currents are gone.
With a delay of 3us we should have a chance to still detect the 1us target.

This picture shows the response of a square of kitchen aluminium foil of 25x25mm on the coil.
The yellow trace is the Flyback.
The green and white traces are the output of the preamp, no target and with the target.
The cursor does not show on this picture, but we can see from the indicator at the bottom, 3.485us, that we could start the sample at this time.

This is done on a bread board and with a half finished DD coil with a sweep diameter of about 37cm. Please take that into considerations when you see the noise.

I would much appreciate your input for trying to improve the setup. The idea is to start a discussion and to try all good ideas to improve the results.

Good experiment!

When sampling at low delays, some of these things come into question, that would not at the longer or normal delays.
1. Eddy currents being seen in parts of the coil wire.
2. Solder joints in the coil housing that connect the coil or coils to the coax wire going to the control box.
3. The damping resistor value is determined by the total amount of the coil seen capacitance much or which is parasitic capacitance.

The oscillations that occur after the TX pulse turns off are due to the total coil seen capacitance which needs to be fully damped before the RX circuit can be turned on. Lower damping resistor values are required for higher coil seen capacitance coil designs. Now comes the real design challenge, lowering as much coil seen capacitance as possible to get a higher damping resistor value and a fast coil. A fast coil will approach your design goal to sample very near 3us to detect your small low TC target.

Here are the places in coil design to look at first. Having a good signal generator and scope setup in necessary to measure the coil resonance which reveals the coil seen capacitance. Here are the design places to start considering. Also, the dielectric constant of anything near the coil wire will affect the coil seen capacitance so look for stuff with lower dielectric constant values. Based on my experiments done to do my fast mono coil paper, the coax wire capacitance places about 100 pf of capacitance on the coil. I have played and designed guitars and found that the 10 foot cable between the guitar and amplifier adds about 300pf to the coil inductance to change the resonant frequency in a very audible way. I came across a web site “till.com” where Tillman designed a FET buffer-amplifier that fits in the guitar end of the coax cable plug that puts a multi meg ohm load on the guitar pickups without them seeing the cable capacitance since the FET output is a low impedance. Then I thought how his concept may help a metal detector coil to not see the 100 pf cable capacitance. This means that a creative way to place the minimal amount of active and passive electronic components in or near the coil can eliminate or deeply reduce coil seen cable capacitance and allow a higher damping resistor value.

Here are some ways to lower coil seen capacitance.
1. Teflon insulated wire to make the coils.
1a.Use either fine stranded wire or AWG 30 single strand wire coil wire.
2. Low dielectric constant spacer from coil to shield.
3. Reducing the area of the shield by using Scotch24 wire mesh for a shield.
4. Reducing MOSFET seen capacitance by adding a series diode to drive the coil.
5. Placing active electronics either inside the coil housing or very close outside the coil housing to reduce coil seen cable capacitance.

I hope this post stimulates others to jump in and share more information to sample low TC targets better.

Joseph J. Rogowski

[Gunghouk]

Sounds like a job for a charge coupled amplifier to eliminate cable capacitance.

[Carl-NC]

Detecting a 1us target with an IB coil is not nearly as difficult as with a mono coil, and the things you do to get there might be quite different. For example, fighting for every scrap on the TX side is not as beneficial.

In both cases, the clamp diode recovery time can be a significant contributor to the overall preamp recovery; in an IB system it depends on the high frequency null. I would place all of the resistance in front of the diodes, and probably increase it to, say, 3k or so. Leave the first stage gain at 5, it will help recovery time. The NE5532 might need to be replaced by a faster recovery opamp.

You might find that adding some capacitance in parallel with the RX coil (along with a damping R) will slow down the turn-off transient and spread it out in time. The goal is to prevent the preamp from saturating in the first place. This will also spread out the target response, making it easier to see with a later delay than you would expect. Another option is to increase the feedback cap on the preamps with the same effect. I've done this in PI designs to effectively see faster targets than the sample delay should allow.

[Tinkerer]

Sounds like a job for a charge coupled amplifier to eliminate cable capacitance.

Thank you for your feedback. Could you tell us more about your idea? Maybe show a schematic?

[moodz]

ZP damping easily does 1 us detection. As I have said in other threads you have to move away from the damping resistor.

The red trace is coil damped with ZP.
The blue trace is damping resistor ( old school ).
The green trace is the preamp on ZP.

The coil with the old school damping resistor is still sitting a 50ma residual whilst the ZP is sampling happily at 2 us after turn off.

[Tinkerer]

ivica,
Great, I dug the 5532 out of the parts bin, where it was lying for near 40 years, especially for you.

Could I ask you for a favor?
A checklist of all the factors that need to be taken into account for a PI frontend would be very useful.

I suspect that the checklist will end up with more than 100 items. Each item has its importance. Each item as a part of the frontend and the whole PI detector has to be designed and fit into the total design. Many compromises have to be made for a good fit.

For example: look at the coil. This one is a DD coil. Why DD? How big? Why? Purpose? How deep should it find what target? How heavy can it be? Waterproof? etc., etc. Then we can decide on how many turns of wire, how many ampere turns? How much induction? How many Ohms? What kind of wire?

Could you start the checklist now? Then add every new item as we discuss and define it?

I am sure this would be much appreciated by many forum members.

[ivconic]

Tony I don't think I'm the best address for something like that.
I have been dealing with detectors since 1987 as a detectorist only.
And PI detectors didn't interest me in the slightest until a couple of years ago.
I've made some proven PI projects before, but only mechanically made them and didn't delve into the technology itself.
For the last few years, I've been trying to catch up.
And mostly everything I could say on that topic has already been said many times on the forum. It would be a mere repetition of someone else's words.
(otherwise a frequent occurrence on the forum, which does not make much sense)
...
Why does the coil have to be DD? Why not concentric coplanar?
And as for the ease of making and even easier balancing in the DIY variety; "Omega" coil is the easiest to make.
If our target is with the shortest possible TC, it logically follows that we cannot expect greater detection depths.
In that case, the "Omega" coil becomes even more interesting. Because it is the easiest in terms of balancing and making.
As for RX frontend opamps, the NE5534/2 is a good choice for those on a budget.
But I also have several pieces of OP37 and I think that opamp is also suitable.
When it comes to JLPcb and their "assembly" option, i.e. that they make the pcb but also "glue" some small SMD components on the pcb;
in their list of components there are all those "old" components in SMD packaging as well.
A two-stage differential RX is a good approach. Small gain too. The smallest possible offset or the possibility of additional offset adjustment.
We want the best possible S/N ratio and very fast RX with a wider range until the saturation.
LV/LP type opamp is preferred here.
I think (need to check) that both NE5534 and OP37 also exist in LV/LP variants.
If not, then surely there are adequate substitutes.
From everything written, I hope that some of my suggestions can be seen.

[Tinkerer]

Tony I don't think I'm the best address for something like that.
I have been dealing with detectors since 1987 as a detectorist only.
And PI detectors didn't interest me in the slightest until a couple of years ago.
I've made some proven PI projects before, but only mechanically made them and didn't delve into the technology itself.
For the last few years, I've been trying to catch up.
And mostly everything I could say on that topic has already been said many times on the forum. It would be a mere repetition of someone else's words.
(otherwise a frequent occurrence on the forum, which does not make much sense)
...
Why does the coil have to be DD? Why not concentric coplanar?
And as for the ease of making and even easier balancing in the DIY variety; "Omega" coil is the easiest to make.
If our target is with the shortest possible TC, it logically follows that we cannot expect greater detection depths.
In that case, the "Omega" coil becomes even more interesting. Because it is the easiest in terms of balancing and making.
As for RX frontend opamps, the NE5534/2 is a good choice for those on a budget.
But I also have several pieces of OP37 and I think that opamp is also suitable.
When it comes to JLPcb and their "assembly" option, i.e. that they make the pcb but also "glue" some small SMD components on the pcb;
in their list of components there are all those "old" components in SMD packaging as well.
A two-stage differential RX is a good approach. Small gain too. The smallest possible offset or the possibility of additional offset adjustment.
We want the best possible S/N ratio and very fast RX with a wider range until the saturation.
LV/LP type opamp is preferred here.
I think (need to check) that both NE5534 and OP37 also exist in LV/LP variants.
If not, then surely there are adequate substitutes.
From everything written, I hope that some of my suggestions can be seen.

ivica,
Sorry, I did not explain well.
All I wanted from you is a list of any questions you might have about PI. Starting with the frontend.
Bbsailor has made some excellent suggestions.
Carl has given very good advice. Actually, I will use Carls AMX schematic for the next frontend test. There I can show on a real circuit, on the scope, How the various experiments with damping behave and what the results are.
Moodz has offered an advanced dynamic damping as a next possibility.
However, there are so many things to try to achieve the best results, that I often loose track and get lost on some insignificant details.
A checklist of all the questions and recommendations would help me keep on track.

Most of us on this forum are here to learn something, compare results and get help or suggestions how to solve problems or challenges. Myself, I have learned very much and received a lot of help during my 25 years on the forum and am very grateful to Carl for having maintained this forum all these years.
As I am developing this PI, I think it might be interesting and useful for others to see and learn of the various possibilities and experiments in the development of a PI.
In the frontend alone, there are many factors that have a significant influence in the end result.
I am trying to explain and show on a real circuit and with scope pictures what the function of every single part is and how it influences the whole.
This additional work would be my pay-forward for all the benefit I received from the forum over the past 25 years.

[pito]

any questions you might have about = can you increase gain of U2 and remove U1 a/b ?

[Detectorist#1]

Hi Carl,

Please, comment this text from the book for engineers:
So, Schottky diodes do not have reverse recovery time because they do not have anything to recover from. However, the vacuum is effectively acting as a dielectric in one direction, so there is some small amount of parasitic capacitance. The reverse current seen in Schottky diodes is not actually reverse conduction, but merely a capacitive discharge. This is why Schottky's are said to have 'soft' recovery, as the curve is really just a capacitor discharge curve, and that takes time. But it is not 'on' and allowing reverse current flow. All the current flowing in reverse is due to energy stored capacitively from the diode itself.

[Detectorist#1]

This is the reason I use low signal BAT48RL Schottky diode with 7A peek current in my PI projects

[ivconic]

ivica,
Sorry, I did not explain well.
All I wanted from you is a list of any questions you might have about PI. Starting with the frontend.
Bbsailor has made some excellent suggestions.
Carl has given very good advice. Actually, I will use Carls AMX schematic for the next frontend test. There I can show on a real circuit, on the scope, How the various experiments with damping behave and what the results are.
Moodz has offered an advanced dynamic damping as a next possibility.
However, there are so many things to try to achieve the best results, that I often loose track and get lost on some insignificant details.
A checklist of all the questions and recommendations would help me keep on track.

Most of us on this forum are here to learn something, compare results and get help or suggestions how to solve problems or challenges. Myself, I have learned very much and received a lot of help during my 25 years on the forum and am very grateful to Carl for having maintained this forum all these years.
As I am developing this PI, I think it might be interesting and useful for others to see and learn of the various possibilities and experiments in the development of a PI.
In the frontend alone, there are many factors that have a significant influence in the end result.
I am trying to explain and show on a real circuit and with scope pictures what the function of every single part is and how it influences the whole.
This additional work would be my pay-forward for all the benefit I received from the forum over the past 25 years.

To keep the story as simple as possible, I will explain why I am interested in and need a good PI detector.
That is, why, despite the planned purchase, I will not buy Deus 2, but I will direct the money to some good PI.
(Most recently I was impressed with Yotube demonstration of AKA Intronik by MikronBg, but I can't be sure, relying only on one video)
On my sites, described many times (but few people read those posts), the situation is as follows:
1) Pointed conical hills
2) Very dense small vegetation, mostly hornbeam-type wood, small stunted plants with greatly expanded branches that stick out in all directions.
Often no higher than 2-3 meters.A couple of times I had severe eye injuries in those places.
3) Soil is composed of material from the collapsed towers and fortresses that were/are on top of such hills.
4) A lot of crushed ceramic particles, a lot of pieces and smaller and larger same ceramic.
5) All this mixed with earth but "interspersed" with smaller and larger stones that resemble a mica stone, and the detector sees it in the mildest way described as "hot rock".
6) In such places (the ones I am most interested in) all the VLF I/B I have had so far behaved the same; more or less worse.
7) XP Deus, which I have had since 2012, "sang its own" in such places too. No more use of it there.

But why am I most interested in such places?
Because only surface layers up to 10-15 cm depth were investigated. All the coins and precious metals were collected from those depths a long time ago.
And these are places with a rich history and very valuable finds.
In the same location, there was a fortress before Christ, during the republic, then until the 5th century, later the Byzantine period, later the Middle Ages.
Civilization by civilization; layer by layer.
There are more valuable finds left there than have been found so far.
I'm talking about silver and gold coins of great numismatic value. As well as silver and gold jewelry of great value.
I'm talking about depths from 10-15cm to 50cm and in some places up to 100cm.
Minelab SD, GP and GPX type detectors cannot approach such places. It is impossible to work with such robust detectors and such robust coils there.
The XP Deus type detector is ideal for such places. By construction, coil without cable, very small control unit and very resistant and simple rod.
And above all very small coils. Which can approach.
But XP Deus is ideal only in "packaging". According to the way it works, it is no longer capable to find anything. Because is VLF I/B.
Gold coins are often at a depth of up to 10cm. VLF I/B detectors have no response to them at all despite the crazy shallow depths.
No reaction, not even a threshold change. Completely mute.
There are no gold nuggets in such places.
By the way, I am not interested in anything smaller than a gold or silver coin, sizes starting from 6-7 mm in diameter and more.
The coil should be as small and "deep" as possible.
GEB and DISC are mandatory!
DISC because there are many particles and pieces of rusty iron in such places, of various shapes and sizes.
Sometimes there are 50 "signals" per square meter.
So, in the described way, I hope I have explained what kind of PI detector I am interested in.
Instead of listing nebulous specifications, parameters and who knows what other nonsense; it is much clearer when I describe in this way what I need for
continuous and successful work on my sites. I have about 50 such sites in the area in a radius of 20 km around my house.
Tony, I hope you had the patience to read this far with understanding.
Because what I wrote here applies to over 90% of detectorists in this part of Eastern Europe, the "Balkans", etc.
"1uS... 5uS...TC this.. TC that... oscilloscope curve like this... that.. flyback this ... that... simulation shows this... that..." really
me totally don't care about all such forum stories.
From the point of view of a real detectorist, I am not at all interested in such quibbles. Completely empty and useless ramblings.
If the end consumer is a detectorist; and if you are some kind of engineer who designs such electronics; then here are your real requirements.
And I'm sure that 99% of the members and visitors of this forum want the same thing.
A solution that will give them practical results on the fields. Success in pursuit in their fields, etc.
All the stories and theories that remain in the abstract domain, all the mental self-centered diarrhea... this forum is full of; it is of no use to the vast majority of people who come here.
Gentlemen engineers, gentlemen "experts", gentlemen "geniuses"; here's a guide on how to put your brilliant minds to work.