The hobby of treasure hunting using a metal locator started in America about ten years ago and has been growing in popularity every since; in Britain the hobby has grown to enormous proportions. Commercial metal locators are not cheap starting with kits at the £15 mark but with a big gap before most of the built models appear. The average price is in the £50 region (there are notable exceptions of course) yet the circuitry in these is by no means complex. The important part about an induction balance metal locator is the search head and no one should underestimate this - this accounts for a significant part of the total cost and, if you tackle this project, expect to devote a lot of time to lining up and experimenting with this.
The reason for the popularity of treasure hunting is that it works using a reasonable metal locator you can hardly fail to find coins and other items lost or thrown away. Our fields and pathways are littered with metal which has been there for hundreds, even thousands of years. The art of knowing where to look is almost more important than the technical performance of the machine: a good detector helps of course but it's how it is used that's important.

Because of the enormous popularity of the Mark 1 we couldn't resist the temptation of having a good look at the circuit and design to see if it couldn't be improved upon. Readers who are interested in this field are strongly recommended to see the February 77 issue (not unfortunately available as a backnumber) or the reprint in Top Projects No.5 (available).

Our first step was to look at the original design in the light of experience could we improve it? We came up with a dozen variations to try but to our surprise we were unable to make any real improvement on the first circuit using the general principles. We could have reduced the package count by using an LM389 (which in~cludes three independent transistors plus an audio output amplifier) but that would have cost more with no real change.
In the original design the trans­mitter was modulated and the peaks of the detected signal were gated and enormously amplified (See How It Works and Fig la). Although we refer to the signal being modulated, it was actually switched on and off and this resulted in ringing in the tuned circuit.
After literally three weeks solid experimenting we decided to take another approach. We decided to dispense with a modulated transmitter and work with DC until the final stages. In the original design the audio frequency was fixed, being dependent upon the modulator and metal was sensed by an increase in audio level. However, our ears are highly insen­sitive to changes in level but they are however, very sensitive to a change in audio frequency. Once we had decided to tackle it from this side everything fell into place. For a long while our voltage controlled oscillator was a unijunction transistor and although we achieved excellent results we were not satisfied with the unit in practice and eventually adopted the circuit shown in Fig.3. FIp. la (below) shows the block diagram of the Mark 1. In thus the peaks of the modulated signal wore gated and enor­mously amplified. Diagram below shows the new arrangement. the RF signal, which is unmodulated, is converted to a DC signal which delves a voltage controlled oscillator (VCO).

Continued onNEXT PAGE