Magnetic metal detector for sawmill

[hunter63]

Magnet-based metal detector for sawmills

[Detectorist#1]

See information about metal detectors for passengers at airports - your case is very similar. Also for detecting of metals in industrial conveyor belts.

[ivconic]

It is not as easy a task as it seems to some.
Very small filings of special steel that do not have magnetic properties or are very small often break off and end up in the product (for example, flour).
It is a terribly difficult task to make such a "flow" detector.
Finally; that's why they are so expensive.

[Detectorist#1]

Yes, you are right, it is very difficult task.... Small nails in the wood's parts is very hard to find as in contemporary "mine" detectors intended to find very small metal parts. But the problem needs some solution. Happy, in this task, the searching not need sophisticated "ground balance" solution as in mine detectors! This task have to be solved.

[ivconic]

I'm going to write something that isn't exactly complimentary and even less promotional for myself.
X years ago, the owner of a very large and powerful pharmaceutical company came to visit me and asked me to design a similar "detector" for him that would find small steel tool filings in the product.
A large sum was on the table.
Aware of the problem and even more aware of my modest (read as "none") capabilities; I flatly rejected that offer in a few words.
That was the period when I was in unprecedented financial problems and I desperately needed money.
The man left totally disappointed.
The point; a person should stand firmly with both feet on the ground.
And above all, he should be very aware of the problem that is presented to him.
Don't get involved in a game where you have no chance.
My pessimism is correct this time. Forget any DIY attempts on those subjects.

[Detectorist#1]

Hi hunter63,
The environment in sawmill is too noisy - strong electromagnetic fields from power electric motors. The metal detector have to be designed with capabilities to suppress the external fields. As ivconic says - the task is very hard.
Nevertheless, an idea for building I give in attached file. The implementation depends on your level of knowledge.

[pito]

. As ivconic says - the task is very hard.
No is not, food industry is using that.

This arrangement is effectively reducing interferences.

[Carl-NC]

This arrangement is effectively reducing interferences.

Correct, metal detectors are used in factories on conveyor belts driven by motors. A security walk-thru detector often has an X-ray machine (with conveyor belt) right next to it. There are ways to deal with the EMI.

[ivconic]

However, in the food industry when it comes to products like various types of teas, flour... things like that, especially with modern mills that grind;
the vital parts of the machine are mostly made of some special type of steel that does not rust, lasts a long time and has virtually no magnetic properties.
From my story above, it was specifically about the production of teas, that is, the mill that ground plants for tea.
The owner brought me a dozen of the usual sawdust that often breaks off and gets mixed up with the raw material.
At that moment, I had a dozen detectors at home, but the closest to that story was the XP Gold Max Power (18kHz, 22.5cm coil).
XP GMP couldn't even detect those chunks.
They were pieces of metal from 2 mm to 10 mm, of various irregular shapes.
Simple; something happens and the machine crushes some of its moving parts and those metal filings end up in the material.
Conveyor belts system is represented here as well, "food" (material) passes through a small "tunnel" and such sawdust is detected there.
The setup with 3 coils as shown in the sketch above is insufficiently efficient.
When I saw that such a request was far beyond my realistic capabilities; I gave up.
It was very unfortunate to willingly give up the rather large sum of money that was offered!

[Carl-NC]

Sawmills are typically looking for nails and spikes, things large enough to damage a saw blade.

[ivconic]

Ah!
This is a typical example of my ignorance of the English language!
Instead of slowly looking up in the dictionary what exactly the term means; I immediately had a picture of my case in my head!
I apologize!
That's a completely different case and it's already somewhat easier to do!
...
What is actually required here is a detector that first looks like a very good but not too complicated PI detector!
With certain specific conditions.
My humble opinion is that VLF I/B is not a good solution in this case.
PI seems like a logical choice to me.
...
This is not my condescension or anything like that; but the Bipolar PI from ITMD3 seems like an ideal starting platform to me.
I made it, I have some test experiences, I think it is the closest to what is asked here.

[ivconic]

What could satisfy the condition from the sketch Pito gave is the following (Carl is here to judge),
the Bipolar PI already has the option of separate TX and RX coils.
TX remains the same, conventional, ok.
RX can be done as an "8" coil, and "sandwiched" with TX in the middle, ok?

[pito]

This is my first detector, built a loooong time ago. It could detect a nail from 60 cm away, so placing a conveyor belt between the coils will solve the problem.

[ivconic]

When I suggested the "8" coil; I had in mind the "physiological" feature of such a coil to easily cancel the disturbances caused by other parts of the machine.
Of course, the mechanical production of such a "sandwich" will be challenging.
The PI detector itself has already been checked and there are no making problems.
The author is among us, let the "client" agree with the author about details and "royalties"...

[Detectorist#1]

The problem most probably will be solved with my proposal in post#6. An other problem is the fact that we not know what is the level of knowledge of electronics of the "client". No information for that.