low cost metal detector

GENERALINTEREST

60 Elektor Electronics 10/2003

Metal detectors are available commercially buttheir price tag is often prohibitive to young-sters and those just starting out with the trea-sure hunting hobby. This article proposes ametal detector that is fun to build at a very lowinvestment. Carefully built and aligned, it willclearly pick up a tiny 15 mm dia. brass coin at70 mm in air, or a 25 mm coin at 120 mm. It will

not fail to pick these up vaguely at upto one quarter greater distance. Theability to locate coins buried in thesoil are, of course, dependent on soilconditions, dry sand being the mostfavourable, and clay, the worst,‘medium’ between the treasure andthe search head.

The electronicsThe electronics you’ll need to buildfor this project is relatively simpleand based on common or gardencomponents only — some of whichyou may already have available inthe junkbox or your regular compo-nent supply. The electronics (Fig-ure 1) consists of a pulse transmitterand an associated receiver, withone–way traffic using two coils as amedium.

The transmitter is built aroundIC1. The CMOS (low-power) 555 gen-erates a square wave output signalwith a duty cycle of nearly 50% and afrequency of about 700 Hz. With the555 in astable mode the output fre-quency is determined by compo-nents R4, R3 and C3. The outputpulse is applied to TX (transmitter)coil L1 via series network R8-C4,with the electrolytic capacitor block-ing the passage of DC through thecoil, and the resistor protecting theoutput stage inside the 555. Thepulse edges generated by the 555will excite the coil and causedamped oscillation bursts at the self-resonance frequency of about 10 kHz.

The receiver section (IC2) is pre-ceded by a simple yet effective pre-

Minimalist Induction-Balance Metal DetectorA low-cost treasure hunter

Design by Thomas Scarborough

This design is arguably the simplest IB (induction balance) metal detectorthat could be built from off the shelf components. The IB method of metaldetection has a good depth of penetration and discriminates well betweenferrous and non-ferrous metals.

detector considerably. It also enables a freshpack to be substituted quickly.

The on-board voltage regulator, IC4, isconfigured with R1 and R2 to supply an out-put voltage of 8.6 V. Current consumptionfrom the battery pack will be of the order of20 mA depending mostly on the buzzer activ-ity, of course.

Construction — the PCBPopulating the printed circuit board shown inFigure 2 should not present problemsbecause the board is spacious and onlyleaded components are used. Beginnersshould take strict guidance from the parts listand the component overlay printed on theboard. As you can see from the photographs inFigures 3 and 4, the potentiometer shafts runthrough the board. This is done to allowknobs to be fitted once the shafts have beencut to a length which is determined by theenclosure. As we are dealing a with a prettysensitive piece of equipment, we recommendusing a metal enclosure. This also allows thepotentiometer cases to be grounded via theretaining nuts and washers.

Construction — search coil assemblyThe constructing of the coils takes you intothe much dreaded realms of ‘mechanical con-

amplifier stage based on opamp IC3,which amplifies the signal receivedfrom the RX (receiver) coil, L2 viaC11-R9. The gain of the opamp isadjustable using P1 (coarse) and P2(fine). The second CMOS 555 in thecircuit, IC2, is configured to act as atrue threshold detector, its output(pin 3) going High as soon as the sig-nal level at the input (pin 2) dropsbelow 1/3 of the supply voltage (orabout 2.9 volts). Likewise the outputdrops low again as soon as the volt-age at the THR input (pin 6) exceeds2/3 of the supply voltage or about7.4 volts). So, a 700-Hz signal willjust start to sound from the piezobuzzer if the carefully adjustedthreshold is exceeded. This adjust-ment is extremely critical and the‘crux’ of the circuit.

Coupled coilsThe TX and RX coils are criticallycoupled so that the presence ofmetal will disturb their magneticcoupling and with it the carefullyadjusted ‘quieting’ of the thresholddetector. Both coils have the samesize and have a partial overlap. Thisenables the RX coil to pick up a pos-itive as well as a negative (inverted)

portion of the magnetic field gener-ated by the TX coil. Because at care-ful balance of the coils the positiveand negative signals cancel out eachother, in theory the RX coil will sup-ply zero output signal. This we willcall a ‘null’. However, because of (for-tunate) practical restrictions, a verysmall residual signal will always begenerated. Once the delicate bal-ance between the fields is disturbedby the presence of a metal object(which will absorb magnetic fieldenergy) the RX coil will start to sup-ply a higher output signal, causingthe electrical threshold set up on IC2to be exceeded, and the buzzer tostart sounding. In practice, thedetector adjustment is optimal withthe passive piezo buzzer producinga soft crackling sound in the absenceof metal. At this setting, the soundlevel increases considerably when ametal object is detected.

The adjustment of the ‘null’ posi-tion of the coils is critical and will bedescribed further on.

The circuit is powered by a 12-Vbattery or a battery pack consisting of8 penlight (AA) batteries or recharge-able cells. The use of an external bat-tery pack carried on the shoulder willreduce the overall weight of the metal

GENERALINTEREST

6110/2003 Elektor Electronics

IC1DIS

THR

OUT555

TR

CV

2

7

6

4

R

3

5

8

1

IC2DIS

THR

OUT555

TR

CV

2

7

6

4

R

3

5

8

1

C7

100n

C9

100n

C3

10n

C8

100n

C10

100n

BZ1

R3

10

0k

R4

4k7

R1

2k7

R8

10

0Ω

R5

10

0k

R6

10

0k

R9

10

0Ω

R7

1k

R2

470Ω

C2

47µ10V

C1

100µ25V

TL081

IC3

2

3 6

7

4

P2

2k5

P1

50k

C11

100n

C4

10µ10V

317T

IC4

D1

1N4001

K2K1 020290 - 11

L1

TX

L2

RX

+12V

20mA

11V3 8V6

7V36

4V9

700Hz

4V3

4V3

4V3

5V7 5V7

C6

10µ10V

Figure 1. Circuit diagram of the metal detector.

struction’ although some would argue for ‘tin-kering away on a Sunday afternoon’. What-ever, it will present a welcome change fromhandling your soldering iron and those com-ponents you know so well.

Below is a search head description (assuggested by the author) which we have noreason to question for veracity! Because a pic-ture tells more than a thousand words, refer-ence is made to Figure 5.Both coils are identical. If you can get it, use33 SWG (0.26 mm) enamelled copper wire,winding 100 turns in a clockwise direction ona circular 15 cm diameter former. The wirediameter is not critical — anything between0.2 mm and 0.3 mm will do just fine. The coilis then temporarily held together with stubs ofinsulating tape passed under it and pressedtogether over the top. It may be jumble-wound. A second coil is wound in the sameway. Each coil is then tightly bound by wind-ing insulating tape around its entire circum-ference.

Next, add a Faraday shield to eachcoil. This is accomplished with somelong, thin strips of aluminium- or tin-foil. First scrape the enamel off thebase of each coil’s end wire. Solder a100 mm length of bare wire to thebase, and twist this around the coil,over the insulating tape. This pro-vides electrical contact for the Fara-day shield. Beginning at the base ofthis lead, the foil is wound aroundthe circumference of the coil, so thatno insulating tape is still visibleunder the foil — but the foil shouldnot complete a full 360 degrees.Leave a small gap – say 10 mm – sothat the foil does not meet after hav-ing done most of the round (if you goall the way, you’ll create short-cir-cuited winding that will introduce aformidable amount of unwantedattenuation). Do this with both coils.Attach each of the coils to qualitybalanced (or ‘symmetrical’) screenedmicrophone cable, with the Faradayshields being soldered to the screensof the cables. Do not use “stereo”

screened (microphone) cable — thismay cause interference between thecoils. The cable you need has twotwisted signal wires inside a com-mon screening braid. Each coil isnow again tightly bound with insu-lating tape around its entire circum-ference. Last of all, wind strips ofabsorbent cloth around each coil(dishwashing cloth should suit),using a little all-purpose glue to keepthem in place. Later, when resin ispoured over the coils, this clothmeshes the coils into the resin.Gently bend the completed coils untileach one is reasonably flat and circu-lar, with each end wire facing awayfrom you, and to the right of the begin-ning wire. Now bend them further,until they form lopsided ovals — likecapital D’s, as shown in the diagram.The backs of the D’s are positioned sothat they overlap each other slightlyon the search head — this is the criti-cal part of the operation, which will beperformed after the printed circuitboard has been completed.

GENERALINTEREST


(C) ELEKTOR020290-1

BZ1

C1

C2

C3

C4

C6

C7

C8

C9

C10

C11

D1

H1

H2 H3

H4

IC1

IC2

IC3

IC4

K1 K2

P1

P2

R1

R2R

3

R4

R5

R6

R7

R8 R

9

- +

020290-1

(C) ELEKTOR020290-1

Figure 2. Circuit board design (board available ready-made).

COMPONENTS LIST

Resistors:R1 = 2kΩ7R2 = 470ΩR3,R5,R6 = 100kΩR4 = 4kΩ7R7 = 1kΩR8,R9 = 100ΩP1 = 50kΩ potentiometer, linearP2 = 2kΩ5 potentiometer, linear

Capacitors:C1 = 100µF 25V radialC2 = 47µF 10V radialC3 = 10nFC4, C6 = 10µF 10V radialC7-C11 = 100nF

Semiconductors:D1 = 1N4001IC1,IC2 = 555C, TLC555, 7555 (CMOS)IC3 = TL071 CP, TL081 CPIC4 = (LM)317T (in TO220 case )

Miscellaneous:K1, K2 = 3-way PCB terminal block, lead

pitch 5mmBZ1 = passive (ac) piezo buzzerPC1,PC2 = solder pins12V battery or battery block (8 AA penlight

cells)36-30 SWG (0.2-0.3mm dia.) enamelled

copper wire, 2 x 50mEnclosure: Bimbox 5004-14 or Hammond

1590B (109x58x25mm)5m symmetrical screened microphone cable

(two wires with common screening)PCB, order code 020290-1

A completed and properly working printedcircuit board is needed before we can ‘pot’the coils. They are potted with plastic resinin a hard non-metallic base-plate. Any base-plate of suitable size will do, on condition thatit is rigid. The author’s prototype was pottedon a suitably cut piece of masonite, with‘walls’ made of 5 mm wood dowel gluedaround its edges. These walls should be‘resin-tight’. Do not use any metal in thesearch head. Also, cover a 4 cm segment ofcoil with Blu-Tack (Pres-Stik) so that itremains unpotted, and may be bent for finaladjustment if required.First place the coils on top of one another —ensuring that they are correctly orientated,with each end wire facing away from you,and to the right of the beginning wire. Turnthe gain pots fully counter clockwise (mini-mum gain) to reduce the amplification to min-imum. Attach a 12 V battery pack, and switchon. Now slowly move the coils apart until thebuzzer is quiet. This is where the voltages inthe Rx coil ‘null’. Increase the gain somewhatand tweak the coil positions again. Repeatthis adjustment several times, each timeincreasing the gain. The higher the gain canbe set when still finding a null, the more sen-sitive the detector will be. Always move thecoils from fully meshed to fully apart, notthe other way around. If you adjust at the‘wrong side’, the signal level on the coils willfirst drop when a metal object is detected,pass through the null, and then increaseagain to a level where the buzzer starts tosound. That’s ok in principle, but the detec-tor will be very insensitive.Take your time to positively identify the opti-mum coil positions. If necessary use awooden jig for easier mechanical adjustmentand comparison of results. Once you’ve foundthe precise point where the coils need to beset, take a marker pen, and mark for holesaround both coils. These holes are used topass cable-ties through, to hold the coilstightly to the base-plate. Also use a cable-tieto hold the microphone cables to the base-plate. Use some Blu-tack (or Pres-stik) totightly seal the holes underneath the platebefore pouring the resin — plastic resin canbe very ‘runny’, and sticks faster than manyglues. Carefully bend the coils at the centreof the plate until you reach the exact balanceat which there is neither silence nor scream-ing in the buzzer or headphones, but acrackle. A little drift should not matter at thispoint. You might also wish to attach a swivel-joint to the search head at this point, whichwill be attached later to the metal detector’sshaft. One-inch (or larger) PVC kitchendrainage conduit may be used for the shaft.Now you are ready to mix and pour the resin.

Again, the coils need to be screened(with tin- or aluminium foil) to makesure they only respond to their ownmagnetic field, eliminating the riskof signal breakthrough (caused byeven the smallest degree of capaci-tive coupling). Defective or inade-quate screening will result in fastpulse edges from IC2 arrivingdirectly at the input of IC3. Thesestray pulses will make it impossible

to find a ‘null’ in the search coiladjustment (see below).

AdjustmentAs you may have gathered, con-struction of the search head is notyet finished. At this point, however,some construction aspects start tointeract with the adjustment of thecircuit. Confused? Read on.

GENERALINTEREST


Figure 3. Top side of completed board.

Figure 4. Underside of completed board.

Use the correct amount of catalyst,so that there is not too much heatand shrinkage in the resin. Pour theresin over the cloth which surroundsthe coils, so as to soak it, and keepon pouring at least until the entirebottom of the plate is covered withresin. The circuit may no longer func-tion correctly at this point until theresin has hardened, so make no moreadjustments at this stage, butswitch the circuit off.When the resin has set hard, keepthe search head away from all metal,and away from computer equipment,which can cause serious interference,and switch on. Adjust potentiometerP2 (fine-tune) to their mid-points.Then adjust P1 (tune) until the metaldetector is just at the point where acrackle is heard, between silence anda scream. Use the tune and fine-tuneknobs for any further tuning. Move acoin over the search head, and thepiezo sounder should ‘scream’.

GENERALINTEREST


Beginning

Tx and Rxcoils

End

150 mm

230 mm

45°

200

mm

50 m

m

100 turns33 s.w.g.

Secure with stubs of insulating tape,then bind tightly all round

Beginning

Tx coil

End

From shield

Again bind tightly all roundwith insulating tape

5 mm dia. wood dowel 020290 - 12

Wind a 20 mm widestrip of foil around

all but 10 mm ofcircumference

Wind a 20 mm widestrip of absorbent cloth

aroundTx and Rx coils

Beginning

Rx coil

Tx coil Rx coil

End

From shield

Again bind tightly all roundwith insulating tape

Position and secureTx and Rx coils with

cable ties on rigid base.

Add swivel joint andblu-tack to base before

pouring resin

Wind a 20 mm widestrip of foil around

all but 10 mm ofcircumference

Figure 5. Search head construction guide.

Figure 6. Alternative search head assembly made from Perspex.

time square shaped, is used to hold the semi-circles together, secure them to the PVC pipe,and still allow the semicircles to be adjusted.The latter is achieved with the aid of nylonscrews. This assembly, when finished andmechanically adjusted, should still be cast inresin or securely glued to warrant the neces-sary ruggedness.

In actual useYou’ll soon find that the adjustment of themetal detector will be affected by the miner-alisation of the ground you are searching, aswell as temperature and voltage variations —so readjustments to P1 and P2 are inevitable.One trade-off for the extreme simplicity of thisdesign is some drift. While it is not excessive,re-tuning will be necessary on a fairly regu-lar basis.

At the centre of the search head, the circuit’srejection of iron is very high, so that once youhave a ‘feel’ for the detector, iron can be vir-tually excluded. This is a boon for anyonewho is searching in the first instance for coinsor noble metals.

(020290-1)

Alternative construction method

Figures 6 and 7 show an alternativeconstruction method for the search

head, based on the use of pieces ofPerspex. The TX and RX coils aretucked away in a groove milled inthe side of each semicircle of Per-spex. Another piece of Perspex, this

GENERALINTEREST


Figure 7. Milling the grooves in the Perspex semicircles.

low cost metal detector

Documents