On Mon, Sep 19, 2022 at 02:28 AM, Gerard wrote:
Hello, Good morning,
Very interesting article. I did a translation in French.
Now tell me if I¡¯m wrong about the measurement method.
1) We calibrate the NanoVna and then we measure the filter, we obtain a raw
state of the filter (2 peaks).
2) The right measure would therefore be to calibrate the NanoVna with the 2
resistances in series (In our example 940 ohm --> 2*470 ohm) and not 50 ohms
3) Measure next.
4) Observe the curve
Similarly, I did not find the NDK 21F15B filter features to check if the
manufacturer impedance is around 470 ohm for our example.
I understood that in our case, it would take 2 adjustable resistors, and
repeat the operation to find the ideal curve.
Yes you have the correct procedure. Below is the test setup used by a Monolithic Crystal Filter (MCF) vendor using series resistors and a shunt capacitor to get the required termination impedance..
Note that if the MCF termination resistance is 520 ohms the series resistor would be 470 ohms (if the test equipment is 50 ohms). When you calibrate with the series resistors connected together the S21 measurement range of the NanoVNA will be reduced. Assuming a 70 dB range for S21 with a NanoVNA and using two 470 ohm resistors in series the S21 range will be reduced by 20 dB so you can only see MCF attenuation to 50 dB. The larger the resistors the worse it gets. With a termination resistance of 2000 ohms the series resistors would add up to 3900 ohms and the S21 measurement range will be reduced by 32 dB.
Roger
Roger Need
