Component Selection in a 4Ghz System
I'm wondering about component selection regarding a 4-4.5GHz system. I need to select some inductors for RF-chokes in my circuit. There are two inductors of same inductance value but different tolerance and SRF.
I have this duality problem in different cases with similar numbers.
1nH with SRF@6GHz and tolerance with +- 0.1nH
1nH with SRF@15GHz and tolerance with +-0.3nH
I can make my circuit work with high tolerance by selecting other component values accordingly;but I'm not sure how working with a part with SRF close(?!) to my carrier frequency would affect my circuit or better yet is it viable to take that risk?
Any opinions about that? I'm not sure how I can simulate effect of SRF in simulations and I don't want to order some hundreds of wrong inductor.
Best Regards,
ktr
Different inductors may have quite different impedance, eventhough they have the same nominal value. The tolerance is not critical for RF-choke purpose. The SRF is more important.
Simsurfing from Murata is a good tool to know the actually impedance at different frequency points.
http://ds.murata.co.jp/simsurfing/index.html?lcid=en-us
As I know RF inductor manufacturers will at least provide you the S-parameter. You need to convert S-parameter to impedance. Murata simfurfing allows you to view the impedance directly, that is why I recommend it to you.
You should select an inductor who has highest SRF, there is no constraint for capacitors @ 4.5GHz.
But also, you'd better use Microstiplines as much as possible to match or other purposes.( If you have sufficient place )
You have to work on the circuit and tweak the components to tolerate components' tolerances.
In my opinion, operating the choke near SRF is a high risk because the data sheet values are somewhat idealized, and in reality you add capacitance by mounting it onto the PCB. Substrate thickness will make a difference for these parasitics, too.
Matching with Tlines is quite hard in terms of space, it is too crowded in terms of PCB space. I tried using some 0 ohms to give me additional shunts; but their parasitic effect is not something I am comfortable with, they do work(I will write some stuff about their effect later on when I have time)
I found some inductors with SRF as high as 20GHz albeit at higher price; but I took the warning from you and @volker seriously.