| Products > Test Equipment |
| XJW01 Auto LCR meter review ($120 bench top LCR meter) |
| << < (11/26) > >> |
| Yansi:
No no... The V and I measured accross the DUT is not 90 degree apart, the phaseshift of those is defined by the impedance of the DUT at the tested frequency. Up to the synchronous detector, you have two analog signals, theoreticaly sinewaves. You will then need to measure how those two are phase shifted each apart (and what is their amplitude). Because also the voltage may be (and is) shifted a bit against the source (DDS output or any other generator of the sinewave test signal), you have to measure both V and I and take the phase difference between them. Lowpass? That is exactly right. The result of the synchronous detection is a DC voltage (look at the dannyf's diagram). If you feed the synchrodetector with an in-phase modulation, you will then integrate the magnitude of the real value Re{V} or Re{I} on the DUT. If you feed the reference as a quadrature, then the detected value will be the magnitude of the imaginary part, Im{V} and Im{I} respectively. As a result, you have measured both V and I accross the DUT, both as a complex values. It is then simple matter of the ohms law to calculate Z = V / I (in complex math of course). |
| dannyf:
--- Quote ---I speculate that he is switching between PA0-high/PA1-low and PA0-low/PA1-high at 1/4 the signal rate, thus producing 0 and 90deg sampling of both voltage and current for a total of 4 readings. --- End quote --- No. He is measuring the relative phase angle of V vs. the reference (switching) signal (PA0/PA1), and then I vs. the same reference signal. From that, it is easy to get the relative phase between V and I. |
| dannyf:
A big part of the complexity, in both models, is rooted in the desire of using and reusing cheap parts, as in thee form of fixed gain amplifiers, mcu control of gains, single sync detector forr both channels, etc. I think if I were to build one today, I would use two sync detectors, four adc modules, pgas, touch screen, etc. This would have greatly simplified the design. |
| happydad:
In the synchronous detector, are we alternately sampling half of the sine wave? And the integrators (low-pass filters) filter the output to give us DC values? I'm still trying to grasp the concept of synchronous detectors. Thanks. |
| Yansi:
The basic principle of the synchrodetector is to multiply the incoming signal with the reference one. It's a basic math with goniometric functions: Multiplying two sinewaves with the same frequency leads to a dc component in the result (also dependent on their respective phase), however multiplying two with different frequencies leads to zero DC component. The DC component is what you get on the output after filtering. Therefore it is telling you, how much amplitude has the synchronous frequency present in the input signal. you can try have some fun in a math tool (at least wolframalpha.com) to see, what the dc result is, if you multiply your reference sin(t) or cos(t) with some harmonic input signal. For example, if your input is sin(t-pi/2), you will get zero in-phase (sin(t) reference multiplied) result, but maximum quadrature (cos(t) multiplied) result. From the two resulting DC components (in-phase and quadrature), you can easily tell, what amplitude does the input synchronous wave has (module of the vector, sqrt(I^2+Q^2)) and whats the phase difference between the reference and the input: tan(phi)=Q/I. The chinese meter uses somewhat simplified technique, multiplying the input only with a +1 or -1, like only the sgn(sin(t)) or sgn(cos(t)) references. This technique is cheap, easy, but somewhat shitty and deprecated these days, where you can get cheap enough MCU with enough grunt to sample both inputs (V and I) at 1Msps or so and do the multiplications and integration in real time. |
| Navigation |
| Message Index |
| Next page |
| Previous page |