felezjoo PI(the best pulse induction metal detector that I made until now)

[jladre]

If you were to set up the detector to accept a Euro coin but reject the pliers, I suspect (if you leave the settings unchanged) it will then also reject an aluminium soda can.

Maybe I'm wrong, but perhaps one of you can test this theory and report back.[/QUOTE]

Definately no! The delay adjustment is versatile you can select automatic delay or manual. And about the iron removal it is range in 0-30. 0 means discrimination is off and as 1-30 is your choice which metal you want to reject.. If you set it to just reject iron then you can and still get the aluminum, and if you set it to reject iron and aluminum then you can also. It easy to understand if you just have the machine on hand because theres a plenty of setting you can choose.

[ivconic]

From Ivconic's posts on this subject, and your comments above; it would appear that the Felezjoo achieves it's so-called "iron discrimination" by simply rejecting large high conductivity targets. This is discrimination by target conductivity, not ferrous discrimination per se.

Due to the large preamp gain, the opamp easily rails out and takes some time to recover. Ivconic mentioned this delay earlier, and is probably a combination of opamp recovery time and software processing. If the coil parameters and the opamp gain are set correctly, you can a get a situation where a Euro coin (for example) will not saturate the preamp when close to the coil. If the micro is sampling at a number of points along the decay curve (instead of taking one main sample) then the reported results would make sense. The detector appears to be set in an "auto" mode, and it's not possible to determine the main sample delay since it's hidden in the software.

We can see in the videos that the audio tone changes in discrete steps as the coin gets closer to the coil. If the decay from the coin does not extend into the period(s) where later samples are taken, then the target is accepted. With a larger high conductivity target (such as a pair of pliers) the decay time is longer, and this gets rejected. Hence, it makes the detector look like it can reject iron targets of all sizes and shapes, when in fact it is actually rejecting all large high conductivity targets over a certain size.

If you were to set up the detector to accept a Euro coin but reject the pliers, I suspect (if you leave the settings unchanged) it will then also reject an aluminium soda can.

Maybe I'm wrong, but perhaps one of you can test this theory and report back.

It appears so. There are 3 steps in audio. As you noticed correctly; 1 euro coin is causing only 2 of those to appear as it gets closer to coil. Third which is missing at 1 euro coin is probably the point of saturation.
I will most certainly test it on large low conductive mass too.

[ivconic]

"...I will most certainly test it on large low conductive mass too..."

I tried to to that right now, just to check and clear up from any suspicions and doubts.
Although right now i am working on something quite different, my workshop is in complete mess.
I had difficulties to place and position coil correctly because it is surrounded with horrible noise from various sources.
I got several "coil error" messages until i founded position of coil where it could be balanced correctly.
Once it was balanced; the test could be done.
As "low conductive mass" i used aluminum plate. It was correctly detected as "Non-Ferro" and displayed with proper VDI number.
Damn... i didn't recorded very interesting GEB sequence! I used a decent piece of ancient Roman brick which is rich with mineralization.
After the first power On; detector reacted on that brick. Than with a brick at 10cm from coil surface i pushed "Down" button (Ground balance) and it was balanced well on that brick, no more reactions on it.
Obviously it is having GEB and it is working correctly.
Next time i will record also a video on that.

[Speedy_G]

[QUOTE=ivconic;214477]"...I will most certainly test it on large low conductive mass too..."

"...I tried to to that right now, just to check and clear up from any suspicions and doubts.
Although right now i am working on something quite different, my workshop is in complete mess.
I had difficulties to place and position coil correctly because it is surrounded with horrible noise from various sources.
I got several "coil error" messages until i founded position of coil where it could be balanced correctly.
Once it was balanced; the test could be done.
As "low conductive mass" i used aluminum plate. It was correctly detected as "Non-Ferro" and displayed with proper VDI number..."

Hi Ivconic

Thanks for you excellent video on the aluminium plate with felezjoo PI - but can you please explain to me why you are referring to the aluminium plate as "...low conductive mass..." ?
According to information that I have, aluminium has a conductivity of 59% of that of copper. This means that aluminium is of HIGH conductivity, and furthermore, in relation to other metals aluminium has quite a low resistivity.
Am I missing something here? I think that this confirms what Qiaozhi has been saying that "...it would appear that the Felezjoo achieves it's so-called "iron discrimination" by simply rejecting large high conductivity targets. This is discrimination by target conductivity, not ferrous discrimination per se...". So what you have here is a HIGH conductivity target of LOW mass (your aluminium plate).

The real test to determine whether Qiaozhi is correct is to do your same experiment using two different metal targets (e.g. iron and aluminium) of equal weight (if at all possible) and repeat that experiment on each target separately, using the same felezjoo PI settings. Please let us know what your new readings now show us.

Thanks.

Ciao

Speedy_G

[Qiaozhi]

The real test to determine whether Qiaozhi is correct is to do your same experiment using two different metal targets (e.g. iron and aluminium) of equal weight (if at all possible) and repeat that experiment on each target separately, using the same felezjoo PI settings. Please let us know what your new readings now show us.

Thanks Speedy_G ... I was about to ask the same question.

[Qiaozhi]

Here's the problem:

Is it just me, or does there seem to be something amiss between what we're seeing in Jiadre's videos and what we observe in the published schematic?
The Felezjoo PI has a single-stage preamp with a gain of about 2000x (twice that of many other PIs, such as Hammerhead, Surf-PI, Baracuda, etc.). The coil inductance is approximately 320uH, and the damping resistor is quite low in value. All of these factors indicate that this design is not suitable for detecting low conductivity targets like thin gold chains and gold nuggets. However, Jiadre shows the Felezjoo detecting a small signal diode (presumably something like a 1N4148 ) as a test for a low conductivity target. Something just doesn't add up here.

So ... I decided to conduct a small experiment.
I took a Laser Hawkeye, a VLF detector known to be sensitive to very thin targets such as hammered silver coins. The sensitivity was set to 10, all metal mode, and the test was done in air using a 9"x8" concentric coil. Here's what I discovered:

A 1N4148 diode was detectable (when presented flat to the coil) at a distance of 1".
But then I remembered that Jiadre presented the diode to the coil end on. In this case I found the diode was detectable at 4".
As a true representation of a low conductivity target, I then did the same test with a 5mm x 5mm section of soda can. It was detected at 1".
Next I carried out the same tests with a Crossbow Classic PI (which is a modified version of Hammerhead). This detector is designed for maximum depth on coins, and is aimed at beach detection and hoard hunting. The main sample delay was set at 15us, and the coil was a 10" mono. It was not expected that the CCPI would be able to detect the 5mm x 5mm square of aluminium, so I wasn't disappointed when there was no response. However, I was surprised to discover that the 1N4148 diode could be detected 1" away when flat, and 2" away when end on.

Clearly, the diode cannot be considered as an example of a very low conductivity target, and the video demonstration is misleading. It does in fact respond more like a reasonable sized piece of aluminium foil.

What I would like to see is a scope image showing the TX oscillator pulse on channel 1 versus the preamp output on channel 2. This will show whether the detector is really capable of early sampling, which I find unlikely.

[Qiaozhi]

Here's some more info to chew on ...

The attached simulation includes the Felezjoo transmitter circuit, plus the preamp. If you run the simulation you can readily see that (even with a 380uH coil having zero parasitic capacitance) the preamp doesn't start to come out of saturation until about 15us, resulting in a possible main sample delay of around 20us. This is exactly as would be expected by just looking at the component values in the schematic.

In that case you would have to ask the question as to why it would even require a ground balance feature?
From Ivconic's tests it is clear that there is a movable hole in the response, which is capable of eliminating a Roman brick. In which case it's conceivable that the so-called iron discrimination is a combination of conductivity discrimination (mentioned in an earlier post) and the movable hole. Personally I can envisage that this is a recipe for missing lots of valuable targets, and I don't expect we will eventually reveal some magic secret in this circuit. If I had the circuit here for analysis, it would be possible to figure out the truth, but I would only be interested to do so in the interests of solving a mini-mystery

[ivconic]

[QUOTE=Speedy_G;214482]

"...I will most certainly test it on large low conductive mass too..."

"...I tried to to that right now, just to check and clear up from any suspicions and doubts.
Although right now i am working on something quite different, my workshop is in complete mess.
I had difficulties to place and position coil correctly because it is surrounded with horrible noise from various sources.
I got several "coil error" messages until i founded position of coil where it could be balanced correctly.
Once it was balanced; the test could be done.
As "low conductive mass" i used aluminum plate. It was correctly detected as "Non-Ferro" and displayed with proper VDI number..."

Hi Ivconic

Thanks for you excellent video on the aluminium plate with felezjoo PI - but can you please explain to me why you are referring to the aluminium plate as "...low conductive mass..." ?
According to information that I have, aluminium has a conductivity of 59% of that of copper. This means that aluminium is of HIGH conductivity, and furthermore, in relation to other metals aluminium has quite a low resistivity.
Am I missing something here? I think that this confirms what Qiaozhi has been saying that "...it would appear that the Felezjoo achieves it's so-called "iron discrimination" by simply rejecting large high conductivity targets. This is discrimination by target conductivity, not ferrous discrimination per se...". So what you have here is a HIGH conductivity target of LOW mass (your aluminium plate).

The real test to determine whether Qiaozhi is correct is to do your same experiment using two different metal targets (e.g. iron and aluminium) of equal weight (if at all possible) and repeat that experiment on each target separately, using the same felezjoo PI settings. Please let us know what your new readings now show us.

Thanks.

Ciao

Speedy_G

Ok... if you followed all the posts here; most probably you saw this:

Qiaozhi - "...If you were to set up the detector to accept a Euro coin but reject the pliers, I suspect (if you leave the settings unchanged) it will then also reject an aluminium soda can..."

aluminium soda can

[ivconic]

Thanks Speedy_G ... I was about to ask the same question.

aluminium soda can

[ivconic]

I just did what i was suggested to do.
I don't have aluminium soda can right now and that's why i used Al plate... similar "mass" and same metal type.
And now both of you asking me why i did that?
BECAUSE I WAS SUGGESTED TO TO THAT BY QIAOZHI!
...
Somebody here is crazy... we only have to find who?


People... FelezJoo PI is very simple to build.
If you are interested in details; please build it and perform all the tests you want.
But don't come here and make empty claims before you are 100% sure in those.
Claims like: "do not have GB... don't discriminate...this... that..."
As i said; without practice you can use your "profound" theory only to wipe your... you know what.

[Qiaozhi]

[QUOTE=ivconic;214489]

Ok... if you followed all the posts here; most probably you saw this:

Qiaozhi - "...If you were to set up the detector to accept a Euro coin but reject the pliers, I suspect (if you leave the settings unchanged) it will then also reject an aluminium soda can..."

aluminium soda can

In your post you said: "Once it was balanced; the test could be done.
As "low conductive mass" i used aluminum plate. It was correctly detected as "Non-Ferro" and displayed with proper VDI number..."

You did not state whether you first set up the detector to accept a Euro coin and reject a pair of pliers before testing with the aluminium plate.

Of course, it might be that you were using aluminum (American spelling) instead of aluminium (English spelling).

A bit of history for you -> http://www.worldwidewords.org/articles/aluminium.htm

[ivconic]

In your post you said: "Once it was balanced; the test could be done.
As "low conductive mass" i used aluminum plate. It was correctly detected as "Non-Ferro" and displayed with proper VDI number..."

You did not state whether you first set up the detector to accept a Euro coin and reject a pair of pliers before testing with the aluminium plate.

Of course, it might be that you were using aluminum (American spelling) instead of aluminium (English spelling).

A bit of history for you -> http://www.worldwidewords.org/articles/aluminium.htm

It is assumed that under the very same setup it is accepting 1 e coin and Al plate and rejecting the iron.
I didn't emphasized that because it is obvious.
So, to repeat; yes under the very same adjustments and setup; it is accepting 1e coin and Al plate and rejecting iron.
In Serbian it is also "Aluminium" and i get constant spell-checker alerts because of that (red underline), than i started to use word "Aluminum" instead.
Oh boy! Don't you tell me nothing about Aluminium.. I INVENTED IT!
30% of my lungs are Aluminium!
It is my middle name, boy!

[ivconic]

Here's some more info to chew on ...

The attached simulation includes the Felezjoo transmitter circuit, plus the preamp. If you run the simulation you can readily see that (even with a 380uH coil having zero parasitic capacitance) the preamp doesn't start to come out of saturation until about 15us, resulting in a possible main sample delay of around 20us. This is exactly as would be expected by just looking at the component values in the schematic.

In that case you would have to ask the question as to why it would even require a ground balance feature?
From Ivconic's tests it is clear that there is a movable hole in the response, which is capable of eliminating a Roman brick. In which case it's conceivable that the so-called iron discrimination is a combination of conductivity discrimination (mentioned in an earlier post) and the movable hole. Personally I can envisage that this is a recipe for missing lots of valuable targets, and I don't expect we will eventually reveal some magic secret in this circuit. If I had the circuit here for analysis, it would be possible to figure out the truth, but I would only be interested to do so in the interests of solving a mini-mystery

Ok, now seriously.
I don't expect no mystery nor miracles also.
It is the way it is; a man did excellent job with just few parts. And that's about all.
Everything what could be done with one lower range MCU and one opamp; author did.
I ventured into this whole story and debate not to convince people here that this is extraordinary ultimate project which will beat all the competitors in all possible aspects and under all possible criteria.
I joined this topic only to support some facts about this nice hobby projects, a fact which are (so it seems) deliberately neglected, suppressed and overlooked!
From the very start of the topic it feels some distaste and excessive criticism towards this project!? Why?
And at the same time we see constant glorifying of few already pretty washed up simpler and poorer projects like MPP, Hamerhead and Crosbow PI???
I am not saying here those are bad, i am only saying; give this nice FelezJoo PI a chance! Let other hobbyists to be interested for it. Don't kill any enthusiasm by constantly spitting on it.
I don't recall that you Qiaozhi payed so much attention and excessive criticism on MPP or HH or Crosbow PI ?? How come?

Just look at the "chronology" of your remarks here;
1) "Not discriminating".
Than i proved you that it is discriminating.
2) "Not having GEB".
Than i proved you that is is having excellent GEB.
Now the last bijou from you is: "...In that case you would have to ask the question as to why it would even require a ground balance feature?..."
Ha,ha,ha! What kind of question and attitude is that?
Why it would require GEB? Simple answer; TO BE PROPERLY ADAPTED ON GROUND CONDITIONS! Without GEB it would give response to mentioned Roman brick.
...
Mannn!
What kind of one sided mind you must have!??
With such frequent remarks which otherwise have no slightest basis in reality you knowingly and perfidiously influencing in a reputation of ideas that you do not like because they are not your personal, so it seems, sorry if by any strange coincidence i might be possibly wrong in this.
Which i certainly am not.
Please come again here and exhaustively specify all the better features that (for example) HH, CrosbowPI and MPP are having against FelezJooo PI??
Please name ONLY ONE!
ONLY ONE!
NONE!
But... you can't earn even one cent on FelezJoo PI and that's what makes it undesirable to be promoted and built here among forum members.
While on the other side; CrosbowPI and MPP are yours "small and shy" projects (actually not your intellectual products but just plain copies of somebody's else ideas) on which occasionally you earn a buck or few....
And that's what is in root of the whole of this story that is going on here.
Spit on any idea "other than mine", discredit it in the very beginning and focus attention of forum members on something which will bring you a buck or two instead.
This is a trend that is clearly visible also on other similar topics too.
...
That's why i came here and am trying to put a new, more honest and objective light on this otherwise indeed splendid project called FelezJoo PI.
FelezJoo PI is just the very BEST small PI diy project on this forum, with no competition at all.
All other already mentioned diy projects here are just funny jokes comparing to FelezJoo PI.
Be a fair and live with it.

[ppodg]

Qiaozhi, check your 1n4148 with a magnet - kovar is ferromagnetic

[jladre]

Her is a short video on aluminum detection. A 5x6 mm aluminum foil folded into two is detectable at 3". A 1x1" aluminum sheet 50% thicker than soda can is detectable at 14-15". Note: it is just an air test. And it is all metal mode. Critics is accepted anyway thats not my problem
NOTE: i dont and never said that 1n4148 diode is a low conductive target. What i say is tiny target.https://m.youtube.com/watch?v=N3EeKpyZIEo