Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Quality factor of capacitor and inductor in filter design

Status
Not open for further replies.
A

An_RF_Newbie

Member level 4
Member level 4
Joined
Sep 1, 2020
Messages
70
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
572
I designed my desired Chebyshev filter with these parameters (F0=100 MHz, BW=10 MHz, attenuation=40 dB at 20 MHz). The results are acceptable.
uE3R8.png


13.JPG

12.JPG


After changing the quality factors of inductors and capacitors to default values (which are available in markets) from Q=30 to Q=50, the results are something weird (s21 becomes less than -20dB at F0.) How can I solve this problem?

15.JPG

14.JPG


I've seen and as I understand these conventional filters have these kinds of problems. Any idea to improve and correct this circuit?
 

Sort by date Sort by votes
First, the Capacitor and Inductor values are absurd and not applicable/practical.
Therefore "the Sensitivity" becomes high. Play filter characteristics and find more reasonable values.
 
Upvote 0 Downvote
Thank you, BigBoss, for your help.
I try to find the right and approximate capacitors and inductors after finding the right solution for suppressing the side effect of loss because of Qualify factors. (is there any way to standardize the capacitors and inductors in the first step of filter designing?)
Could you let me know more about "Play filter characteristics and find more reasonable values."?
Thanks
 

Upvote 0 Downvote
What BigBoss means is change the topology of the filter
All you have done is the textbook LPF prototype to to BPF conversion at 50 ohms. This almost always gives impractical values.
For the performance you seem to be looking for, look at helical resonator filters perhaps. Component Q for cheap parts has a very determinate effect on performance.
 
Upvote 0 Downvote
John.Jin1990 said:
Thank you, BigBoss, for your help.
I try to find the right and approximate capacitors and inductors after finding the right solution for suppressing the side effect of loss because of Qualify factors. (is there any way to standardize the capacitors and inductors in the first step of filter designing?)
Could you let me know more about "Play filter characteristics and find more reasonable values."?
Thanks
Click to expand...
I think you requested very tight attenuation level at very close frequencies. Loose them a little bit first and try other topologies. Can you post your requested specifications here ? Let me try for you.
 

Upvote 0 Downvote
Thank you very much BigBoss and G4BCH, for your cooperation.

As the same the other, I follow the RF Filter books to design my desired circuit, and I don't know much about different topologies But as I use Filter Solutions software, I was successful in checking other topologies. None of them works fine for Low Q capacitors and inductors.

I need to design a filter with these specifications:

F0=100 MHz, BW=10 MHz, attenuation=40 dB at 20 MH

I really appreciate your help and kindness.
 

Upvote 0 Downvote
If you look at the pole pair Q of your filter (6th order Chebyshev, 0.05 dB ripple, 10 % bandwidth), you see Q numbers of 18 up to 72. Respectively it can't be implemented as lumped filter with the provided low Q components, besides all practical problems of meeting the exact component values. As far as I'm aware of, the Filter Solutions tool has no feature to recalculate filter parameters to compensate for parasitics, so it won't be able to implement a basically feasible optimized lumped filter with lower Q.
 
Upvote 0 Downvote
Thank you, FvM, for your explanation. What do you suggest improving or implementing this filter with low Q elements?
In addition, could you let me know how you estimated Q between 18 to 72?

I appreciate any help you can provide.
 

Upvote 0 Downvote
the capac
John.Jin1990 said:
I designed my desired Chebyshev filter with these parameters (F0=100 MHz, BW=10 MHz, attenuation=40 dB at 20 MHz). The results are acceptable.
View attachment 175476

View attachment 175477
View attachment 175478

After changing the quality factors of inductors and capacitors to default values (which are available in markets) from Q=30 to Q=50, the results are something weird (s21 becomes less than -20dB at F0.) How can I solve this problem?

View attachment 175479
View attachment 175480

I've seen and as I understand these conventional filters have these kinds of problems. Any idea to improve and correct this circuit?
Click to expand...
the capacitor, inductor, and resistor values are ideal and cannot be applied in practice you need to first simulate the circuit to ensure that the components that you will use give the best results.
 
Upvote 0 Downvote
You cannot realize this filter with those specifications by using well known topologies because your specifications are very tight. I think your aim is to filter FM Band.
In order to realize a such filter, you should use "Coupled Helical Resonators" in a metal cavity. Because bandwidth is pretty narrow and this will bring a necessity of very high Q valued inductors.
 
Upvote 0 Downvote
Thanks all for your help, Could you provide me with some references for sharp filter design for F<1Gighz?
 

Upvote 0 Downvote
Try a google search for helical resonator filter design.
this gives amongst many others:
https://www.rfcafe.com/references/electrical/helical-resonator.htm
http://tonnesoftware.com/helical.html I have not tried the software

Helical filters work fine in the 100MHz region, they can get a bit large if you need very high Q
A good book on filter design is Handbook of Filter Synthesis by Zverev, it is pretty much the standard work.
You may get it to work using other topologies if you use high quality low ESR porcelain ceramic capacitors like kyocera 600 &700 series and physically large high Q air cored inductors. It will almost certainly need tuning to get a decent response.
 
Upvote 0 Downvote
I have designed a affordable filter using with standard Murata Components. You can start from this point.
1650320447257.png

1650320478932.png


1650320548851.png
 
Upvote 0 Downvote
John.Jin1990 said:
Thanks all for your help, Could you provide me with some references for sharp filter design for F<1Gighz?
Click to expand...
For me, I think I would refer you to how to do simulations using Proteus EDA. This could help you be able to simulate any type of filter you want to design. Check here to understand how to do the simulations using Proteus EDA. You can also check on the Texas instrumentation filter calculator. The calculator will do the calculations for you and then generate the circuit of the filter that you need.
Thank you!
 

Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top