A genius solution !

[Riss]

The underwater SECON UWM20 was sent to me for repair - https://ebingergroup.de/wp-content/u...ffekt_engl.pdf ; I could not find a manual, there is very little information about this model. https://buergerrechtler-micha.blogsp...gerat-mit.html. Extremely simple to use and superior sensitivity - opinion of the owner. https://www.djlaan.nl/secon-uwm-20-w...inter-detector - ,,The sensitivity of the detector can be adjusted by turning the adjustment ring on the front of the handle'' - this got me very interested...- what is the principle ? There was no radiation - I disassembled it - white paste on the front of the board, near two electrolytic capacitors ... I replaced the key IRFD9220 because it was covered with a white paste (cuts the current through the coil) and it worked immediately. The adjusting ring is very close to the coil ! - it has an average diameter of 13 cm. . A plate of magnetic material is placed on the inner diameter of the adjusting ring - it strongly sticks to a magnet ! , it was not covered with rust. The plate is 12 mm wide, its diameter is 44 mm and has the shape of an arc -180 degrees - it covers half the diameter of the adjusting ring. I measured the detection delay – about 120 microseconds !!! How does this miracle register coins and rings at about 20cm and small gold crosses at 3-5cm with this huge delay?!? .Brilliant solution by Ebinger ! The whole secret is in the arc of magnetic material, but what is the principle?

[Altra]

Hi, did you take any photos?

Maybe there are Hall effect sensors inside which adjust the circuit when the external ring is rotated?

I searched for a second hand unit for a long time. But they seem to be rare in north America and hard to find.

[Riss]

I feel uncomfortable with my too hasty opinion. The other morning I "downloaded" its circuit. Transmitter - the clock is from an 8-pin microprocessor Atmel 1245 / ATTINY 13 / 20 PU. Pulse length through the coil - 32 microseconds ‚ period -1.5 milliseconds. The current through the coil increases linearly up to 0.8 amperes. The coil is 420 microhenry / 2 ohms. Almost all chips have erased inscriptions. The input operational amplifier is double - NE5532 or similar. Standard scheme - 1 kilohm, two diodes - to the inverting input. The gain is 500 times. The other half is an integrator that compensates the input offset (feedback from the integrator output -pin7 -to the non-inverting input –pin3 through 1megohm ) The integrator does not integrate during the TX pulse and 120 microseconds after it .It fooled me and it was the end of the day and I was tired. The other " erased " chip is a MAX323 - two analog switches. At its pin 3 for 150 microseconds the voltage is 0 . On pin 7 is 60 microseconds sample, delay relative to falling edge of TX imp -18 microseconds. Follows a DC amplifier on a single operational amplifier, amplification 220 times - to the pin 1 of the processor. Has a 78L05, LMC7660 (also deleted). DC-DC converter with 5 terminals - smd code S52B. Since it is non-motion, the sheet metal ( arc ) in the ring is used to set the threshold (sensitivity) of the vibration motor. I apologize for the blank shot ...

[Riss]

The owner said that the bane of this detector is the fishing hooks, but so far there are no detectors on the market with good time constant separation. In 2015, I assembled a working prototype with the monocoil, which I tested on many terrains, including the wet strip of the seashore and terrains with strong mineralization. Maximum error - about 25-27 percent (out of 100 dug signals about 25 are iron - error). But unfortunately fishing hooks would in most cases be perceived as a digging signals - they have a abruptly falling front, they are not like the larger iron ones targets - 5 cm nail for example)