Arduino based LC meter - advice needed

[boylesg]

I am following this design to create my own LC meter that can measure small values of capacitance and inductance (pF and nH).

https://electronoobs.com/eng_arduino_tut10_3.php

For starters the tank circuit part of the schematic has got to be wrong - one end of the tank circuit cannot be connected to ground.

I tried a tank circuit on my bread board (100nF and 100uH) and connected to 5V through 150R and 1N4007 and with one end of the tank connected to GND.

However upon removing the connection to 5V, I cannot detect any frequency with my multimeter.

If I break the connection of the tank circuit to GND them my multimeter detects a frequency just fine.

Also I simulated in multisim and in here the oscillation only occurs if I leave the 150R + 1N4007 connection to 5V floating. Connecting it to GND, equivalent to Arduino digitalWrite(nPin, LOW), kills the oscillation.

So in Arduino code I would need to do pinMode(nPin, INPUT) in order to initiate the oscillation.

Am I correct or am I missing something else?

Anyway soldered as specified by this author my device cannot measure capacitiance - pulseIn(....) simply times out.



Also I want to be able to measure very small values of capacitance - down to a few pF.

According to the formula F=1/2PiSRT(RC), I need a very large inductor in order to get the oscillation period up to 1us or more that an Arduino can measure.

So I have wound my own at 130mH, which is going to create quite a kick back.

Is this likely to be a problem for the LM393 or the Arduino? Are there any other suggested modifications to the schematic I should make as a precaution?

 

[betwixt]

From what I can see, it works by pulsing the LC circuit and measuring the period of the 'ringing' waveform. For example, if you momentarily connect a 1.5V battery across any LC circuit, it will oscillate but only for a few cycles. The losses in the L & C and radiated magnetic field quickly reduce the oscillation so your multimeter wouldn't have time to measure the frequency.

There are better methods using similar circuits but it should work, at least at low resonant frequencies. The LC circuit SHOULD be grounded at one end, remember it isn't in a feedback loop and it isn't a continuously running oscillator, it is just being 'kicked' by a pulse through the diode to get it excited for a few cycles.

Brian.

 

[boylesg]

From what I can see, it works by pulsing the LC circuit and measuring the period of the 'ringing' waveform. For example, if you momentarily connect a 1.5V battery across any LC circuit, it will oscillate but only for a few cycles. The losses in the L & C and radiated magnetic field quickly reduce the oscillation so your multimeter wouldn't have time to measure the frequency.

There are better methods using similar circuits but it should work, at least at low resonant frequencies. The LC circuit SHOULD be grounded at one end, remember it isn't in a feedback loop and it isn't a continuously running oscillator, it is just being 'kicked' by a pulse through the diode to get it excited for a few cycles.

Brian.

Apparently arduino has the same problem as my multimeter then.

After some more research Colpitts and Hartley's oscillators seem far more promising.

Set up a Hartley on my bread board with KIA359AN, gain of 10, 2 x 10mH inductors (center grounded) and 1x 100nF cap (parallel with 2 x inductors) gives me a frequency of about 2.7kHz.

Formula gives me 3.5kHz with these values - I suppose either or both of the inductors or capacitor are not quite the value specified.

With 2 x 10mH inductors I should be able to measure down to 10pF with Arduino. Below that and it might get hard to reliably distinguish different values.

Will make the Arduino programming easier because I don't need a charge pin - the oscillation will start autonomously as soon as I insert the cap.

Opinion Brian?