Questions about some continuity tester schematics/circuits I tried

[neo2001]

Hi!

In the last few weeks I tried to build some continuity tester schematics I found on the web (using breadboards). Most of them worked as expected - but some did not - or at least not completely. I really rechecked everything at least twice, even replaced most of the components - but still can't find the problem.

(1) This one doesn't work at all (LM1458-based):
http://kripton2035.free.fr/Resources/continuity%20checker%20adjustable.gif

It doesn't matter what I do with the test leads. The only way to make the buzzer sound (= Q1 switch) is to turn R3 until PIN5 reaches about 8V. Then it sounds permanently. Obviously that's not what was originally intended. AFAIU the voltage on PIN6 should change - which means, the voltage on PIN3 needs to change first. But it doesn't.
I even checked all the jumper cables and measured the resistor values twice - no success.

(2) This second one (LM358-based) does work fine so far, but the red LED (D1) constantly glows very slightly, as soon as the LM358 isn't powered anymore (= test leads open, Q1 off).
http://www.electroschematics.com/9232/continuity-tester-differentiates-resistance/

If I connect VCC (PIN8) directly to 9V+, D1 turns completely off (like it's supposed to). But I guess the whole idea of Q1 is to turn the OpAmp off so it doesn't draw any more power from the battery. The only way to make the glowing D1 LED (about 0.4mA draw) stop is to remove PIN1, PIN3 and PIN4 (GND) all at the same time (so there is basically no possible connection to ground anymore). But as soon as one of these pins gets reconnected, the D1 LED comes on again. This seems odd, because this circuit was actually tested by other people and worked.

Maybe some of you guys have any ideas what could be the problem here? I really would appreciated any kind of help.

[neo2001]

Any ideas or hints?

[Andy Watson]

The inputs of the LM1458 are not specified to include the negative rail (or in this case, the ground). Typically, they are specified to work (i.e. its "common mode range") to within 3V of the power supply rails. This continuity tester has the first amplifier operating with a common mode voltage of 90mV. Either, supply a negative supply to the opamp, raise the common mode voltage, or use an opamp type (LM358 ?) that includes the negative supply as part of its input common mode range.

[alsetalokin4017]

I've breadboarded the second circuit and it works as stated on the webpage. I don't get any dim glow of the Red LED unless I touch the leads of the 10K resistor, and this makes the Red LED glow dimly. Scope shows that this induces an oscillation in the circuit. So if your Red LED is glowing it is probably because your circuit is oscillating. To cure this you can try adding another 0.1 uF capacitor between Pin 2 and Ground. This works for me to get rid of the oscillation when I touch the 10k resistor.
(I'm using a BC556B transistor instead of the 3906 but I doubt if this makes any difference.)

I'm trying to find a 1458 in my parts stash so I can check the first circuit for you.

[alsetalokin4017]

OK, I found some 1458s. The first circuit does not work with this op-amp, but it _does_ work with LM358 (same pinout) ! 

I also tried my favorite TL082 and it does not work with that one either. But it works fine with LM358.

(I did not try the variable setpoint voltage divider using the 100k pot, I simulated it with two 47k resistors, with the Pin 5 from the midpoint. With these values it triggers with 10R but does not trigger with 22R between the test leads. )

I'm using a BC337-25 transistor instead of the 3904,  and an LED+resistor instead of the buzzer, in a low-side switch configuration. (Led Anode at +9V, 1k resistor from LED Cathode to transistor Collector, Emitter to Ground, Base to Pin 7 as usual.