[SOLVED] Saw in out match design

[cyberjok]

Hello, I have a question about EPCOS B3479 in/out match circuit:



the SAW output is 730||0.8pF ??? . just check the recommand match circuit on the b3479 datasheet with SMITCH V3.0 TOOLS:
and we got the result is 8-j38 ohm.. NOT 50ohm, what's wrong ?

Thanks;

 

[BigBoss]

730¦¦0.8pF is a parallel equivalent circuit that may be equivalent to 8-j*38 Ohm serial ??Check this ..
SAW filters generally have such impedances, it's normal not to be 50 Ohm.

 

[jiripolivka]

SAW filters have typically high i/o impedance, therefore inductors are used for matching as shown in the specification.

 

[cyberjok]

thanks for your input:
730ohm || 0.8pF @869.3MHZ Z=65-j208
**broken link removed**

- - - Updated - - -

YES, but maybe there has something wrong....

- - - Updated - - -

OK, just got the s-parameter for the SAW, S11 = 1.001+j0.006; so the input zin is 50ohm,,,, so i dont know what does them means (730ohm||0.8pF ),,, and the matching network to 50ohm

 

[BigBoss]

thanks for your input:
730ohm || 0.8pF @869.3MHZ Z=65-j208
**broken link removed**

- - - Updated - - -

YES, but maybe there has something wrong....

- - - Updated - - -

OK, just got the s-parameter for the SAW, S11 = 1.001+j0.006; so the input zin is 50ohm,,,, so i dont know what does them means (730ohm||0.8pF ),,, and the matching network to 50ohm

I got it..
First, the input/output is perfectly matched to SAW filter ( S11=0) with L-C component while the SAW filter is connected then the SAW filter disconnected and the impedance which is seen by the SAW filter measured and the complex conjugate of this impedance ( complex conjugate matching principal ) will give you the essential input impedance of the SAW filter.( Parallel or serial equivalent is possible )
This information has been mentioned in EPCOS App. Note #18

 

[cyberjok]

Ok, thanks. BigBoss

just bulit a circuit in ADS ,and want to measure the Zin @869.55MHZ, the result as follow(B3749_WB.7z,is the S2P file):



i think that is wrong. so can i try this way to get the impedance @869.55MHZ ? i want to use it for matching network design.

sorry i am a newbie on RF design;

 

[tony_lth]

I got the same result. As following: 7.799-j165.93@868MHz, 188.821-j38.55@869MHz, 29.245-j174.97@870MHz.
They are not smooth, may be caused by the extrapolation, you should see the ADS warning message.

 

[cyberjok]

Thanks for your input, just checked the S2P file, that is Y-Parameters, Not S-parameters

------------------------------------
! port | pin | signal
! -----|-----|--------
! 1 | 2 | in
! 2 | 4 | out
! --------------------
!
# MHZ Y MA R 1
! freq magY11 angY11 magY21 angY21 magY12 angY12 magY22 angY22

8.690500E+02 2.048057E-03 9.978873E+00 4.167418E-03 1.144720E+02 4.167418E-03 1.144720E+02 2.221259E-03 3.171905E+01



///

--



so the impedance at 869.05MHZ is 1.000533 + j0.012234 ohm; right ???

 

[tony_lth]

No. Just click the S/Y/Z box of the S_Param in ADS, then you can directly get Z params.
Still doubt your S_param file is wrong, not smooth, from Z param can get Z11=52.97-j82.3, similarly to Zin1 from S11 188.8-j38. Anyway from 868M to 871M your original file is not smooth. So different methods get different results.

- - - Updated - - -

I know the root cause. Your original Y param file using 1 ohm as character resistance, not 50 ohms. So change your both terminals to 1 ohm, then get Z(1,1)=70.1-j358.2,
Just use the default matching as datasheet, and using NB(Narrow band) not WB for simulation.