Your reference resistor should be about the same size as the measured impedance, so that the voltages across each component are in the same range. If not, then the smaller impedance will have a very small voltage across it, making for inaccurate measurement. Often, if you have no idea of the size of the impedance, you make a couple of measurements: first with an arbitrary reference (like 1kΩ) to get a rough estimate of the impedance, then again with a more appropriate size. I often use different reference resistors for different frequency ranges, if the impedance is largely inductive or capacitive, and so changes a lot with frequency.
@@mohamedahmedkamal8686 If your load is a capacitor, you decide what frequencies you want to test at and pick an impedance of about the same magnitude. If you want to test over a very wide frequency range, you might need to use a few different reference resistors (a large one a low frequencies and a small one at high frequencies). Theoretically, one could use a reference impedance that is not a resistor, but it is difficult to get precision capacitors or inductors, while precision resistors are cheap.
@@mohamedahmedkamal8686 the readings will be roughly right until your resistor is off by a factor of 1000 or more, because then the voltage being measured will be so small that it cannot be accurately measured. The measurement is most accurate when the two impedances are about the same, and get gradually less accurate as the impedances differ.
I don't understand your comment-the impedance-spectroscopy measurements in this video are not applied to diodes, but to either combinations of standard passive components or to things like loudspeakers and electrodes.
@@gasstationwithoutpumps, I was just trying to imply it would be a dirty trick to put a schottky diode (or other non-linear device) in one of these black-box circuit boards and ask a student to characterize its impedance.
@@jessstuart7495 Yes, it certainly would be, as the impedance would depend on the DC bias and the measurement would have to be made with very small AC signals to avoid changes to the DC bias. The Analog Discovery 2 is not really good for very small signals, so measurements would be rather approximate-either because of the low resolution of the measurements or because large signals would make for a fluctuating bias and hence fluctuating impedance. A reverse-biased diode is basically a capacitor (with capacitance decreasing as the reverse bias increases and the depletion region gets thicker). A forward-biased diode is basically a resistor (at low currents anyway-saturating the diode would make for yet another set of complications). The impedance tokens I made used just a single resistor and a single capacitor, in series or in parallel, with a time constant in a range (5µs to 300µs) that made the corner frequency easily found. I considered making some 3-component tokens for more advanced students (two resistors and a capacitors or two capacitors and a resistor), but I never got around to doing it-the loudspeaker measurement and modeling was challenging enough.
@@gasstationwithoutpumps, The diode would also generate harmonics of the driving signal's frequency, which may or may not alias onto the measured voltages and currents, depending on whether the measurement system implements a tunable bandpass filter (analog or digital) centered about the swept driving frequency.
i have a question please , how can i calculate the suitable value of reference resistance ?
Your reference resistor should be about the same size as the measured impedance, so that the voltages across each component are in the same range. If not, then the smaller impedance will have a very small voltage across it, making for inaccurate measurement. Often, if you have no idea of the size of the impedance, you make a couple of measurements: first with an arbitrary reference (like 1kΩ) to get a rough estimate of the impedance, then again with a more appropriate size. I often use different reference resistors for different frequency ranges, if the impedance is largely inductive or capacitive, and so changes a lot with frequency.
what do you mean with ur last sentence? what if it is largely capacitive ,what should i do ??@@gasstationwithoutpumps
@@mohamedahmedkamal8686 If your load is a capacitor, you decide what frequencies you want to test at and pick an impedance of about the same magnitude. If you want to test over a very wide frequency range, you might need to use a few different reference resistors (a large one a low frequencies and a small one at high frequencies). Theoretically, one could use a reference impedance that is not a resistor, but it is difficult to get precision capacitors or inductors, while precision resistors are cheap.
so readings will be deviated if i chose improper resistor ? or it will be totally wrong ?
@@gasstationwithoutpumps
@@mohamedahmedkamal8686 the readings will be roughly right until your resistor is off by a factor of 1000 or more, because then the voltage being measured will be so small that it cannot be accurately measured. The measurement is most accurate when the two impedances are about the same, and get gradually less accurate as the impedances differ.
I would feel sorry for the student who gets a schottky diode.
I don't understand your comment-the impedance-spectroscopy measurements in this video are not applied to diodes, but to either combinations of standard passive components or to things like loudspeakers and electrodes.
@@gasstationwithoutpumps,
I was just trying to imply it would be a dirty trick to put a schottky diode (or other non-linear device) in one of these black-box circuit boards and ask a student to characterize its impedance.
@@jessstuart7495 Yes, it certainly would be, as the impedance would depend on the DC bias and the measurement would have to be made with very small AC signals to avoid changes to the DC bias. The Analog Discovery 2 is not really good for very small signals, so measurements would be rather approximate-either because of the low resolution of the measurements or because large signals would make for a fluctuating bias and hence fluctuating impedance.
A reverse-biased diode is basically a capacitor (with capacitance decreasing as the reverse bias increases and the depletion region gets thicker). A forward-biased diode is basically a resistor (at low currents anyway-saturating the diode would make for yet another set of complications).
The impedance tokens I made used just a single resistor and a single capacitor, in series or in parallel, with a time constant in a range (5µs to 300µs) that made the corner frequency easily found. I considered making some 3-component tokens for more advanced students (two resistors and a capacitors or two capacitors and a resistor), but I never got around to doing it-the loudspeaker measurement and modeling was challenging enough.
@@gasstationwithoutpumps,
The diode would also generate harmonics of the driving signal's frequency, which may or may not alias onto the measured voltages and currents, depending on whether the measurement system implements a tunable bandpass filter (analog or digital) centered about the swept driving frequency.