R1 = 1K D1,D2 = 1N4148 R2 = 2K2 Z1 = 8V2, 1/4 watt R3,R4 = 22K T1 = 2N3905 (PNP) R5 = 2K7 T2,3,4,5 = 2N3904 (NPN) R6,R7 = 56K 9volt Alkaline battery R8 = *See text suitable loudspeaker C1,C2 = 22nF housing & probes Descriptive Notes: The output from the speaker is not loud but is more than adequate for the purpose. I used a small transistor radio loudspeaker with an impedance of 25 - 80 Ohms. The resistance should be brought up to 300 ohms by adding series resistor R8. Example, if your speaker is 58 ohms, then R8 = 242 ohms. An experiment worth doing is to select the value of either C1 or C2 to produce a frequency oscillation that coinsides with the mechanical resonant frequency of the particular loudspeaker in use. Having choosen the right value, which probably lies in the range of 10n - 100n, the tone will be louder and more earpiercing. A "freewheel" diode D2 is connected across the transducer since fast switching sction of the oscillator circuit can produce a surprisingly high back e.m.f. across the coil and these high voltages might other wise lead to transistor damage of breakdown. Zener diodes do not provide an absolutely constandt volt-drop regardless of current; at the 0.8 mA design current an 8.2 volt diode will quite possibly give only about 8.0 volt drop since test current for zener selection and marking is typically 5 mA or more. A further possible source of error is the battery; the one suggested nominally provides 9V but a brandnew one may be as much as 9.5 to 9.8V until slightly run-down and this "surplus" voltage, combined with an "under-voltage" znere volt-drop will leave considerably more than the forecast voltage available at the probes. A silicon diode D1 is therefore connected in series with the zener to decrease the probe voltage by a further 0.6 volt or so. During your final testing and before boxing your circuit, the most suitable connection, A or B, is selected for the positive probe wire. The aim is to have the circuit oscillating with short circuited probes but to stop oscillation with the least amount of resistance or the inclusion of a diode (try both ways) between the probes. No sensitivity control is fitted because I don't think it is worthwhile nor necessary and would spoil the simplicity of the circuit. There is no easy way to proof the unit against connection to the supply. Be careful if checking AC line wiring and switch off first. In a similar way, if checking electronic apparatus for unwanted bridging between Veroboard tracks, for instance or a suspected crack in a PCB (Printed Circuit Board) track switch off power first also. Good luck! |
|