 |
 |
How it works
The working principle is simple: an AC voltage, with a known frequency, is applied over a known resistor (R) in series with an unknown impedance (X). Based on the voltage ratio (Ux/Ur), the phase shift between Ux and Ur the unknown impedance (X) can be determined.
However, implementing this principle using a soundcard is not so easy. At high impedances (X) Ur is close to 0 while at low impedances Ux is close to 0. In both case it is difficult to get a sufficient accuracy. Mainly because of the inaccurate phase measurement.
In addition the soundcard inputs have a rather low input resistance and a significant input capacitance, appearing in parallel to X.
In fact a soundcard is not really suited to build a RCL-meter, in particular due to the low input impedance and relatively large input capacitance. In addition there is a lot of variation between soundcards in regard with the input impedance, input capacitance, line-in sensitivity and speaker out level.
But it is probably the nature of a radio amateur to try to use things for purposes they are not designed for ...
These problems were solved by taking the Fourier Transforms of Ur and Ux and use these to calculate the voltage ratio and phase shift. After using some tricks and a lot of calculations it seems possible to build a relative accurate (and very cheap) RCL-meter.
System requirements
- As a lot of math is involved (3 FFT's per measurement) at least a Pentium 200MHz with 8MB RAM (16MB or more preferred) is needed.
- Of course a soundcard that can handle 16 bit ADC / 44kHz sample rate. With older / cheaper soundcards the measurement accuracy can be limited, due to a strong internal coupling between speaker out and line in (see help file for details). In addition the soundcard MUST have a line-in input, as 2 signals must be measured at the same time the microphone input (as it is mono) cannot be used.
- At least Win98.
The working principle is simple: an AC voltage, with a known frequency, is applied over a known resistor (R) in series with an unknown impedance (X). Based on the voltage ratio (Ux/Ur), the phase shift between Ux and Ur the unknown impedance (X) can be determined.
However, implementing this principle using a soundcard is not so easy. At high impedances (X) Ur is close to 0 while at low impedances Ux is close to 0. In both case it is difficult to get a sufficient accuracy. Mainly because of the inaccurate phase measurement.
In addition the soundcard inputs have a rather low input resistance and a significant input capacitance, appearing in parallel to X.
In fact a soundcard is not really suited to build a RCL-meter, in particular due to the low input impedance and relatively large input capacitance. In addition there is a lot of variation between soundcards in regard with the input impedance, input capacitance, line-in sensitivity and speaker out level.
But it is probably the nature of a radio amateur to try to use things for purposes they are not designed for ...
These problems were solved by taking the Fourier Transforms of Ur and Ux and use these to calculate the voltage ratio and phase shift. After using some tricks and a lot of calculations it seems possible to build a relative accurate (and very cheap) RCL-meter.
System requirements
- As a lot of math is involved (3 FFT's per measurement) at least a Pentium 200MHz with 8MB RAM (16MB or more preferred) is needed.
- Of course a soundcard that can handle 16 bit ADC / 44kHz sample rate. With older / cheaper soundcards the measurement accuracy can be limited, due to a strong internal coupling between speaker out and line in (see help file for details). In addition the soundcard MUST have a line-in input, as 2 signals must be measured at the same time the microphone input (as it is mono) cannot be used.
- At least Win98.