Deciding I wanted to built another kit project, and with my interest in metal detecting rekindled, I purchased a Surf PI 1.2 kit from Silverdog,
The Surf PI is a Pulse Induction type metal detector. This makes it ideal for beach detecting. The documentation on the project is somewhat sparse, but the PCB is very well made, if a little tight. The sequence of photos below shows the board being populated
I wont go into the electronics of this thing. Suffice to say that the addition of a couple of potentiometers, a power supply and a search coil are all that remains of the electronics.
The search coil of course is the hard bit! A pulse induction machine works by pulsing a current into the coil, then rapidly switching back to its receive circuits to look for disturbances in the decay of the magnetic field. This requires a coil of a specific inductance, and low resistance to give a good strong current. Low capacitance helps make the recovery time faster.
I have now built a test coil. Its somewhat bigger than normal, at 50cm diameter (20cm being more usual), but thats the diameter I needed to make the right inductance from the multicore cable I decided to use. 16 core shielded computer cable, hence 16 turns, the calculations saying that would give the specified 350uH.
The following photo sequence shows the steps needed to make the coil
The shield of the cable has a messenger wire, which conveniently provides a way to connect to it. No attempt has been made to keep the pairing, but then again none has been made to obscure the pairing either, I have no idea how this will affect the capacitance. The shield only connects to one side, in this case the green coil wire. The final binding of insulation tape serves to keep the connections bundled and stop the sleaving from moving.
The final thing to do was to test the coil and measure its resistance and inductance. This was done on a very cheap but surprisingly versatile little tester from China
So, the resistance is nice and low. The inductance is a little higher than the theory said it would be, at 410uH, but I think this is within the tolerances allowable.
The reason I chose to build such a wide test coil, is that I wish to see how this sort of cable performs for search coils. This cable is available in a range of numbers of pairs, this piece being 8 pairs (16 wires). A smaller coil, such as a standard 10 inch coil, would need around 22 wires, or 11 pairs. This cable can be had in that form.
The next step will be to rig the coil and Surf PI module up in a metal free test environment, and see if it actually works.