Categories
Statistics
Flag Counter
Since 08.08.2014
Counts only, if "DNT = disabled".

Your IP is 54.161.178.52
ec2-54-161-178-52.compute-1.
Info
Valid HTML 4.01 Transitional Creative Commons Lizenzvertrag
rss
เราจะทำแบบวิศวกรผู้ยิ่งใหญ่
We love the King
Your valuable opinion :
4 stars

Avg. 4.45 from 11 votes.



Quartz_Crystal_Filter_Designer_3.php   7259 Bytes    16-07-2014 00:20:01


Quartz Crystal Bandpass Filter Design

Verifying the results

measuring the quartz crystal filter




In a first step, the results are verified by simulation. A lot of tools are out there to do this.

Our filter is made of 10.7 MHz quartz crystals. We measured / calculated the following values :

QUARTZ CRYSTAL BANDPASS FILTER DESIGNER
http://www.changpuak.ch/electronics/Quartz_Crystal_Filter_Designer_2.php
Javascript Version : 30. June 2014
-------------------------------------------------------------------------
QUARTZ CRYSTAL USED
Frequency : 10.70000000 MHz
L : 8.6920 mH
Cs : 25.470 fF
Cp : 6.300 pF
R : 10.660 Ohm
Qu : 54801
-------------------------------------------------------------------------
DESIGN DATA FOR A 4 ELEMENT QUARTZ CRYSTAL FILTER.
Center : 10.7000 MHz, Bandwidth : 10.00 kHz
Source / Load Impedance : 735 Ω
-------------------------------------------------------------------------
C[0] = 62.85 pF (series)
C[1] = 39.76 pF (shunt)
C[2] = 50.24 pF (shunt)
C[3] = 39.76 pF (shunt)
C[4] = 62.85 pF (series)
-------------------------------------------------------------------------
POSSIBLE E12 MATCHING NETWORK TO 50 Ω (HIGHPASS) :
C = 82.00 pF (series)
L = 2.70 µH (shunt)
POSSIBLE E24 MATCHING NETWORK TO 50 Ω (HIGHPASS) :
C = 82.00 pF (series)
L = 3.00 µH (shunt)
-------------------------------------------------------------------------
APPENDIX
k[1,2] = 0.6850
k[2,3] = 0.5421
k[3,4] = 0.6850


The filter (for the simulation) looked like this : (some values changed, after)

Simulating the Quartz crystal filter


On the left, there is an input matching network (C1, L4), followed by the filter (C2 ... C6) and again a matching network (L5,C7). The matching networks have been changed, as the treasure chest offered only 1 µH, 2.2 µH and 4.7 µH. However, the following amplitude response could be achieved, BEFORE tuning, just using the calculated values :

Simulating the Quartz crystal filter


An increase of C2 and C6 from 68 pF to 470 pF slightly reduced the insertion loss. The increase of C4 to 44 pF made the passband more symmetric. However, the amplitude response looked like this :

Simulating the Quartz crystal filter


Now it was time to design a pcb. The files below give detailed information about it. We use four quartz crystals, because they nicely fit in our standard case No.: 3.






picture of the Quartz crystal filter





S21, measured of the Quartz crystal filter





S21, measured of the Quartz crystal filter




✈ Share your thoughts


✈ Nothing more to be seen below this point ? Maybe your browser blocks the facebook iframe.
The webmaster does not read the comments regularly. Urgent questions should be send via email.








 
t1 = 3876 d

t2 = 589 ms

★ ★ ★  Copyright © 2006 - 2017 by changpuak.ch  ★ ★ ★

PRchecker.info Impressum