PI metal detector for really small nuggets

[darov922]

I would say, just try and compare. I'm planning to produce my own cores for flat spiral coils but it will likely take few months before I can post results as my time is limited at the moment.

OK LUCIFER... Do you think that in your project this type of coils used to transfer energy could be used in detectors? maybe you already get a coil on magnetic material. Dario

[lucifer]

OK LUCIFER... Do you think that in your project this type of coils used to transfer energy could be used in detectors? maybe you already get a coil on magnetic material. Dario

They may work but it's possible that a ferrite makes a worse performance. Note that the magnetic flux lines are concentrated in the ferrite and for a TX coil we need the magnetic field to spread as further in space/ground as possible. So with a ferrite you may get less depth as magnetic field tends to go towards the ferrite.

On the other hand when used for an RX coil it performs better with a ferrite.

I've done just few initial experiments with flexible ferrite sheets and it is exactly what happens, TX performance gets worse unless a hole is present in the center.

[mickstv]

Asking is still the best way I know of for getting answers...

An obvious solution is to use a separate RX coil. It does not have to be induction balanced and offers the flexibility of optimizing the TX and RX inductances separately.

If you want to use a mono coil then options are limited. The circuit below works pretty well though I have never tried to push it into the realm of aggressive performance.



[ATTACH]55204[/ATTACH]

Hi Carl, I know this thread is a bit old but any way. I haven't been on the forum for a while and haven't messed with detectors in years.

Anyway I saw your circuit posted and was wondering what the timing to control the fets was ?

Seeing as I'm rusty at this now I don't want to start nuking fets for no reason....

Cheers
Michael

[Tinkerer]

Hi Michael,

Maybe you should look at the AMX project: https://www.geotech1.com/forums/forum/projects/active-projects/amx-project

This is a bi-polar, square current wave PI, very well suited for very small gold nuggets, as well as for extra deep large nuggets.
Very flexible.
Very advanced.

I am about to post an excellent timing program from PICON, for the Tinkerer TX version.

[Carl-NC]

Sorry, was traveling and then forgot to get back to this. Timing is simple:


Tinkerer is right, there are better ways. But this will give you a basic bipolar TX.

[mickstv]

Thanks for the replies and answer....

I remember years ago I played with CCPI style TX but could never find a way to use it with just a mono coil (which i prefer). Then on a DD coil I could never get ground balance.to work well on highly reactive ground. But maybe its worth another look.

[Tinkerer]

Yes, it needs a separate RX coil. DD, or Concentric. The concentric can be induction balnced, but does not need to be.

[mickstv]

Yes, it needs a separate RX coil. DD, or Concentric. The concentric can be induction balnced, but does not need to be.

Hi Tinkerer, how would you deal with coil imbalance over highly reactive (mineralized) ground ?

When I was experimenting with the idea I had issues keeping the coils inbalance, maybe thats why ML use a DOD type coil ?

[mickstv]

Yes, it needs a separate RX coil. DD, or Concentric. The concentric can be induction balnced, but does not need to be.

Ok I decided to have a look at the AMX, I was reading its going to use so.e sort of tilt comp to keep the current level using samples etc, how would this work, does it just look at the current used during samples then adjust tv voltage to change tilt. All in the tx section.

Or is the tilt comp done from the RX side samples then fed back to the TX Side ?

[Tinkerer]

Hi Tinkerer, how would you deal with coil imbalance over highly reactive (mineralized) ground ?

When I was experimenting with the idea I had issues keeping the coils inbalance, maybe thats why ML use a DOD type coil ?

I made several concentric coils with a separate RX coil, without any induction balance. I tried with a small diameter RX coil, like about 30% diameter of the TX coil. Also with several other diameter RX up to 70% of the diameter of the TX coil.
All worked electronically, but the detection field shape varied.

Now I am trying DD coils. Many variations are possible. Like always one needs to make a compromise between the advantages and disadvantages.
The DD coil that I am testing now, aim at a wide sweep field for small nuggets.

The DD coil gives some degree of differentiation for FE targets.

I consider the smallest gold nugget of interest to be about 1us TC.

[Tinkerer]

Ok I decided to have a look at the AMX, I was reading its going to use so.e sort of tilt comp to keep the current level using samples etc, how would this work, does it just look at the current used during samples then adjust tv voltage to change tilt. All in the tx section.

Or is the tilt comp done from the RX side samples then fed back to the TX Side ?

The tilt manifests itself by an offset in the RX. This limits the possible gain in the preamp.
Up to some degree this can be compensated by a simple RX offset compensation.

The tilt is caused by the losses in the TX power. Circuit losses as well as losses caused by the absorbed energy by the targets. As the ground is a very large target, it can cause large losses and therefore large RX offsets.

These TX losses can be compensated in many different ways in the TX itself. Several ways of compensation have already been posted. I have been working and testing on a different one.

Like in every development, different ways of achieving the goals have to be tested and the best way be integrated in the project.

[moodz]

The tilt manifests itself by an offset in the RX. This limits the possible gain in the preamp.
Up to some degree this can be compensated by a simple RX offset compensation.

The tilt is caused by the losses in the TX power. Circuit losses as well as losses caused by the absorbed energy by the targets. As the ground is a very large target, it can cause large losses and therefore large RX offsets.

These TX losses can be compensated in many different ways in the TX itself. Several ways of compensation have already been posted. I have been working and testing on a different one.

Like in every development, different ways of achieving the goals have to be tested and the best way be integrated in the project.

I dunno that is strictly correct ... if you are using a preamp that was designed for a damped PI using a decay timing then that is most likely wrong a wrong application as a bipolar pulsing pi does not have a "decay time" .. .it only has a "flyback time" during which energy is conserved to the next pulse.

The bipolar pulsing PI is really a frequency domain beast.

[Tinkerer]

I dunno that is strictly correct ... if you are using a preamp that was designed for a damped PI using a decay timing then that is most likely wrong a wrong application as a bipolar pulsing pi does not have a "decay time" .. .it only has a "flyback time" during which energy is conserved to the next pulse.

The bipolar pulsing PI is really a frequency domain beast.

You make an interesting observation. What kind of preamp do you propose?

Up to now I have only used a PI type preamp with the bi-polar TX. (differential) I obtain classical decay times that are clearly recognizable in their typical TC decays.

Below is a link to my working preamp (breadboard) and a scope shot. I would be interested to try your version of preamp before I make the PCB.

https://www.geotech1.com/forums/foru...521#post419521

[Detectorist#1]

Hi lucifer,
My observations (with many experiments with pinpointer's variants) also points that flat spiral coil is very good solution. Using of ferrite screen on one side of the TX coil will direct the energy only on one side but this will increase the energy only two times. The distance of the recognizing of the targets will increase very moderate. Maybe the efforts have to be directed to improvement of the RX channel.

[lucifer]

Hi lucifer,
My observations (with many experiments with pinpointer's variants) also points that flat spiral coil is very good solution. Using of ferrite screen on one side of the TX coil will direct the energy only on one side but this will increase the energy only two times. The distance of the recognizing of the targets will increase very moderate. Maybe the efforts have to be directed to improvement of the RX channel.

Flat spiral coils can be very fast but not as deep, especially if printed on a PCB as the copper resistance is too high.
I don't see how ferrite screen would direct the magnetic field only on one side. This is wishful thinking but in reality it concentrates the magnetic field from both sides.
Working for a better RX is always a good idea.