[hung]

Hi. Thanks for the link. Today I had time to read that thread.

No wonder you blame the chip. This project employs a PIC. The old 16F877A MCU from Microchip. This line of PICs are a nightmare to progammers because of known instabilities and fragility due to transients leading to malfunction or loss of data to happen out of the blue.
Why he did not employ an AVR/Atmega a much better choice, is a mistery to me.
So you have all the reasons to complain. And I tell you, I think you will not remain alone too long. If Nicholas sold this PIC to others, sooner or later there will be people facing the same problem and he will need to do something about it. What will he do? Tell these customers that they will need to buy another one just like he did with you? I refuse to believe that he chose to use that PIC on purpose just to earn more money. This would be disgusting. I tend to believe that he did this only because he is familiar with that particular PIC. But anything is possible...

Anyway, it's been more than 2 days now I think, and he did not step forward to comment on this or provide any help to remedy the situation. This means that he apparently does not care about a customer after the sale is done. He is out of my list of decent people now.

Pablo, I checked the project and I can see it's a simple device. In sum, the PIC is sending a PWM signal through the CCP1 pin which is amplified and sent to the antenna. Then the incoming response signal is received by another antenna, amplified and sent to the RC0T1clock pin, captured/compared and it activates RC1/CCP2 pin which happens to be the buzzer. It features 2 amplifiers and filter. It also features a standard 16X2 LCD display with the usual text programming according to what I saw in his videos.
However, I see two issues with this project from the start.
First, even if the output signal sent by PWM to 'energize' the target element is a PCM or CW type and of different frequencies than the ones expected to be 'sent' by the elements and received by the device, the filters are inadequate to support a precise response. Gold for instance is set at 5KHz but the device might detect 10, or 15Khz also. Even if the filters were strictly designed only for this band, the device would detect anything ranging in this frequency, from TV interference to maybe even bats flying.
Also the antennas appear to be telescoping type and there is no mention of how they are built. I see no inductors or apparatus to match the low frequencies intended. So they are completely underutilized in terms of energy sent and received.

Now the good news, Pablo. I was touched by your case and I decided to help you.
This is what I am gonna do. I will redesign this circuit, featuring better filters, integrator and detector stages, and will write a code that will take care of all of this.
If you send me your email by PM if you wish, I will send you the new schematic and also the hex code. You will have in your hands a much better device that will trully detect gold or silver. And I will not charge anything to you . All of this for free.
With one condition tough.
You will not disclose this schematic and code to anyone here in this forum. This will be given only to you and I hope you respect this pact.
Allow me 3 to 4 weeks to send this to you as I am quite busy these last days of the year.

Cheer up. It's Xmas time.