RF matching question with BLF571
I have the datasheet for a BLF571 in front of me , I'll try and attach it at the end.
The impedance information is quoted as ZL = 31.7+j29.3 with the arrow pointing OUT of the drain terminal.
Does that mean it's the conjugate of the output impedance of the transistor , and hence looking into the
transistor drain one would "see" 31.7-j29.3 ohms. If so , why not just quote the transistor impedances as they are?
Why quote the load impedance?
I now want to match this to 50 ohms using some equations I've gotten from an RF book.
Qs = Qp = (sqrt)(Rp/Rs-1) , Qs = Xs/Rs and Qp = Rp/Xp . Frequency of interest is 225MHz
I can get "rid" of the output capacitace( - j29.3) with a series inductor of +j29.3 ohms or 20.7nH
I can now match 31.7 ohms to 50 ohms using the above equations. Giving an additional series inductance
of about 17n and parallel capacitance of 10.7p
Is this correct so far. I realise the Q is less than one , so there will be very little attenuation of harmonics ,
but I assume I could then design a low pass or band pass filter to go between the antenna and the output of
the matching network.
Is my methodology correct , or am I missing the boat altogether here. I realise more experienced designers
will probably have better ways of doing things , but you have to start somewhere :0)
Cheers
Ned
Search the net for WinSmith or any other Smith Chart software, and this would make your job easier.
You are heading in the right direction. To your question about "why give ZL and ZS", that is to make it easy for the designer to know what his/her load should present to the transistor. If you give them Zout, then they have to think about "what is the conjugate load". It's really simple, but this just removes some of the confuction... i.e. "make our load look like this, and it'll work".
Without running the numbers, your approach to matching looks sound. As for adding a filter after the amplifier, beware that it's impedance outside the passband will be all sorts of crazy impedances, so your transistor will be poorly matched at it's harmonics. This can cause strange things to happen to the bias point of the transistor, so work up incrementally. Design and implement your matching network, test and verify, then add your filter, test and verify.
Good luck!
Also, +1 on vfone's recommendation for some freeware Smith Chart apps, they make tweaking L's and C's very easy to see the impact on your match. FYI, keep your trajectory on the Smith Chart as close to the middle of the chart as you can. The further away you get from Z0, the more VSWR you have, and the narrower your match's bandwidth will be.
Thanks for the help :0)