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

[Qiaozhi]

C4, C5, C7, C8, C9, C14, C18, C20, C21, C23, C26 - Those should be ceramic caps
C11 2,2µF - Tantalum
C6, C12 100µF - Tantalum

According the 3D image of the PCB, and the silkscreen, C11, C6 and C12 are shown as electrolytics.

[eclipse]

In the iranian document from the designer it's mentioned these caps are تانتالیوم which translates to Tantalum



FelezJoo_PI_v3_00.rar

[Qiaozhi]

In the iranian document from the designer it's mentioned these caps are تانتالیوم which translates to Tantalum

OK - in that case the 3D image of the PCB, and the silkscreen are incorrect. Perhaps the PCB layout program that was used cannot handle tantalum capacitors, and hence they were shown as electrolytics.

[koohyar]

I am not criticizing the detector per se, but merely pointing out that it will never be good on low conductive targets in its current form.

When I quickly calculated the gain as 10,000, the input was taken from the limiting diodes. If you include the 390R, then I agree that the gain is closer to 2,000. However, this is still too high for low conductive targets.
You also mentioned C11 (2u2), which incidentally is an electrolytic capacitor. This type of capacitor is well known to be a noise generator, and should never be used in a filter circuit. It would be better to completely replace this with a two-stage pre-amp (similar to the MPP). With the design as it is, you're caught between a rock and a hard place. In order to use direct sampling, the preamp gain needs to be quite high (since the Atmega ADC has 10-bits resolution), but this then limits the detector to searching for large relics. You could try using oversampling and decimation to increase the ADC resolution beyond 10-bits, but the response will become slow. Perhaps the software is already doing that ... I don't know ... which might account for the audio delay you can see in the videos.

Also note that the potential divider formed by R8 and R7 attenuates the preamp output to less than half its amplitude before being input to the ADC. This is yet another reason for looking at a redesign of the input stage.

One final point ... what is the range of the TX pulse rate for this detector, if it's not fixed to 200pps? I can then change the notes section on the schematic accordingly.

My dear friend
I wonder how you ever did even once the files that I had sent.
To find the answer to these objections unreal?
I put to you some of the parts list, which was sent as pdf
And is about the type and amount of capacitors.
Capacitors
--------------------------------------------------------------------------------------------
C1………………………. 10uF-16v electrolytic
C2, C3 ……………………….18pF ceramic
C4, C5, C7, C8, C9, C14 ……………………..….100nF (or Code 104) ceramic
C18, C20, C21, C23, C26 ……………………….100nF (or Code 104) ceramic
C6, C12 ……………………….100uF-16v (or Code 107) Tantalum
C10 ……………………….5pF (or 4.7 pF) ceramic-high quality
C11………………………. 2.2uF-16v (or Code 225) Tantalum
C13, C17, C25 ……………………….470uF-16v electrolytic
C15, C16 ……………………….4700uF-16v electrolytic
C19, C22 ……………………….1000uF-16v electrolytic
C24 ……………………….2200uF-25v electrolytic
* Ceramic capacitors voltage isn’t important.
*you can Use a capacitor with higher voltage but don’t use capacitors with lower voltage at all.
* If you use more than 24 volt battery, C24 should be 35 volts.

As for the objections that you did, I'm trying to explain one by one:

1. This metal detector is actually very sensitive to small targets and low conductivity.
And the Iranian forum found that some people even complain that it is very sensitive to small metals
However, this also depends on the loop to sense small gold should be flat or Spider
And good loop cable may be used. And all settings effectively in the pulse detectors can be controlled by operator.
Despite the varying features such as sensitivity, frequency(35-999hz), pulse width(100-500), delay, speed, Ground and Integ.W ,Virtually all factors affecting the pulse detector can be controlled by the operator.
So with the right settings, it can achieve the best results for Sense any kind of metals.
2. How do you see only the resistance of 100 ohms, but do not see that there is a resistance of 390 ohm in the same field?
If you suspect that you can simulate the circuit and see the amplification factor to the maximum of his case, according to the frequency is 2000, not 10,000!
This policy exists in many other circuits. I really wonder how you don’t know this?

3. It is true that there are 2.2 micro capacitors, for blocking a DC signal,
But the value of this capacitor is intended to minimize entry of 50 Hz noise.
In this respect and based on the tests 2.2 micro is the best value that is the best answer.
Besides, who said that the electrolytic capacitor should be?
It clearly states that should be at least tantalum and even mentioned that it is better to be a kind, MKT or polyester.
In the end I recommend Friendly Once read files are uploaded And compare the features of the system with other systems that you know To see that it is really worth making And undeniable advantage compared to other similar projects.
Good luck.

[BILLY]

Just my personal view. Please do not be upset. But your detector would be worth about 100 dollars and is not to be better than a cheap detector bought on the market. But as far as a experiment, good for you trying., Keep it up, never know what may happen.

[bernte_one]

i dont know what you want
it is a diy Project
it is free
the part Count is not much
pcb is easy
and it is a sort of high powered pi

so first build it and then make a comment

[eclipse]

Just my personal view. Please do not be upset. But your detector would be worth about 100 dollars and is not to be better than a cheap detector bought on the market. But as far as a experiment, good for you trying., Keep it up, never know what may happen.

Parts for this detector cost around 40$ And that includes box and batteries and a day of soldering work. If you can point us where people can buy for this price PI metaldetector with same capabilities please enlighten us.

[ossman22]

eclipse,koohyar,Michael,you are Iranian.( perhaps bernete-one)
OK. no problem.
but please give us acceptable tests or evidences not by filling pages to make this subject HOT to bring
credit for yourself to sell metal detectors at your country including LRL which i saw you are selling to people each for 1000$ named FPL35.
it is your country and you can do whatever you want at your home. it does not to me and i don't care.
but i have read other stories on your own forum.

[greylourie]

koohyar mentions, "frequency(35-999hz), pulse width(100-500)". This is a useful range for experimenting.

When this project was first introduced to Geotech, I was under the impression that the device only had a frequency range somewhere around 200-300. But actually it is much wider. Much more interesting now.

[daverave]

it does not matter how cheap a detector cost to make for me if it cannot produce excellent depth on gold rings and some coins !!! if people want a detector for large objects then detectors like this are good value...but for me useless if not good depth on rings...microcontrollers and lcd display's do not give you depth !!!!

[Qiaozhi]

I put to you some of the parts list, which was sent as pdf
And is about the type and amount of capacitors.

Much of the confusion here has been caused by the information being written in Persian.
Also, the supplied 3D PCB image (and the PCB silkscreen) both show C11 and C6 as electrolytics.

This metal detector is actually very sensitive to small targets and low conductivity.
And the Iranian forum found that some people even complain that it is very sensitive to small metals
However, this also depends on the loop to sense small gold should be flat or Spider
And good loop cable may be used. And all settings effectively in the pulse detectors can be controlled by operator.
Despite the varying features such as sensitivity, frequency(35-999hz), pulse width(100-500), delay, speed, Ground and Integ.W ,Virtually all factors affecting the pulse detector can be controlled by the operator.
So with the right settings, it can achieve the best results for Sense any kind of metals.

Of course everyone knows these adjustments are possible. But even so, it is immediately obvious that with a preamp gain of 2000, followed by an attenuation of >50% and then directly sampled by a 10-bit ADC, it will not be sensitive to small gold. A video was posted earlier of a test with a gold ring. Although the target consists of a low conductive metal, the shape will enhance the response. As you should know, eddy currents like to flow in a circle, and therefore a gold ring is not a good test.

It is true that there are 2.2 micro capacitors, for blocking a DC signal,
But the value of this capacitor is intended to minimize entry of 50 Hz noise.
In this respect and based on the tests 2.2 micro is the best value that is the best answer.

Yes, I agree. Calculations indicate that the best [preferred] value for this capacitor is 2u2. That has never been questioned.

Besides, who said that the electrolytic capacitor should be?
It clearly states that should be at least tantalum and even mentioned that it is better to be a kind, MKT or polyester.

The supplied images show C6 and C11 as electrolytic capacitors. May I suggest that you update the information to be consistent with the parts list and the description.

As I have stated before. I am not criticising the detector or the design. However, there are some claims being made here that require some clarification. Firstly, the claim of being sensitive to small non-conductive targets conflicts with the topology of the design, since it uses a single-stage pre-amp with twice the normal gain of other designs. In which case it is impossible to sample the RX signal before the eddy currents have decayed. Secondly, there is the claim of iron rejection. So far the conclusion here seems to be that it is based on target conductivity, which means the rejection ability will depend on size, shape and orientation of the target. This technique is well known in the industry, and the GB feature of White's TDI can be adjusted to provide a similar capability. Hence it is also well known that this method of iron rejection is not reliable in the field.

Finally, I would just like to emphasise that no-one here is "knocking" the design or its capabilities. We're simply trying to understand the limits of this particular design, and to test if some of the claims will stand up to scrutiny. Obviously there is also a language translation problem that doesn't help. If the designer hadn't been so secretive in the first place about supplying a schematic, then this back-engineering would not have been necessary.

Please try not to get upset. We're only investigating the design ...

[koohyar]

i dont know what you want
it is a diy Project
it is free
the part Count is not much
pcb is easy
and it is a sort of high powered pi

so first build it and then make a comment

[eclipse]

eclipse,koohyar,Michael,you are Iranian.( perhaps bernete-one)
OK. no problem.
but please give us acceptable tests or evidences not by filling pages to make this subject HOT to bring
credit for yourself to sell metal detectors at your country including LRL which i saw you are selling to people each for 1000$ named FPL35.
it is your country and you can do whatever you want at your home. it does not to me and i don't care.
but i have read other stories on your own forum.

I think there's a lot confusion in your mind going on. I'm not from Iran (nor bernete-one, I guess), also whats up with the hate??
If you need to display your hate against the designers (owners) of these LRL and PI I don't think you'll get satisfaction here, people in this forum aren't
interested in things like that. As you can see nobody in this topic has made false claims for the capabilities of this design, in fact I mentioned that I can't get more than 1 meter detection depth with my current coil, does this sound like advertising?
I appreciate your comments left in this topic but this last comment is scratching my nerves.

[bernte_one]

bernte_one is not from iran
profile says
Join DateJan 2011
Location Berlin
Posts 1,316

[6666]

Eclipse what software and programmer did you use to burn the hex code ?
I tried Winavr and the bascom programmer but would not recognise the uC.