Butterworth Filter design question

Thread Starter

JohnnyB

Joined Mar 24, 2011
3
Is there any additional calculation required to determine the -3dB point for a passive Butterworth filter? (I'm not strong on the heavy formula stuff). I've been using the online calculators to do a 6 pole 400MHz filter (and modelled it in PSpice as well), but when I get the values and proto it up, it's coming up at 336MHz at -3dB. I thought it might be a change in the inductances as the parts were listed as being at that inductance at 100MHz, so I designed and built a 100MHz version, but that came out at 94MHz at -3dB. So am I missing something? Or is there something specific I'm unaware of (I'm new to filters) in regard to needing to use extremely low impedance inductors or something that would cause this issue? Been using 5% parts for both L and C
 

bertus

Joined Apr 5, 2008
22,270
Hello,

How did you build the filters?
Did you also take a look at parasitic capacitance?
Can you post a schematic?

Bertus
 

Thread Starter

JohnnyB

Joined Mar 24, 2011
3
We have a small pcb with the RF plug (with SMA connectors) on one side, and on the other, running between the two SMA signal points, a standard 6 pole butterworth, laid out so we can put up to two SMT components in each DiPole position so we can easily make pluggable filters of various cutoff frequencies. 2oz copper, and nice wide traces. Straight and tight.Parasitics came in at about .1pF equiv. per Inductor and grinding that through Spice only brought the frequency down to 398Mhz. Here's the calculator I was using (along with Spice) entering 400MHz, 50 Ohms, and 6 components:

http://www.calculatoredge.com/electronics/bw pi low pass.htm

All I've got for a schematic is a marked up printed note sheet. I talked to the guy that was really involved with this before and he said he started out using the calculator's values, and then "tweaked" the design by varying the C values, until he got the Freq. he wanted. (shotgunned for results, without solving why the calculated values didn't work--and we've verified the needed values through a couple of different models so I don't think we're off there. I guess I was wondering if there was a basic fact like "only use air core/multilayer L's" or "make sure resonance freq. is 2X the working area freq", or some other thing I'm just missing.
 

bertus

Joined Apr 5, 2008
22,270
Last edited:

Thread Starter

JohnnyB

Joined Mar 24, 2011
3
Thanks, I'll research those. Late in the day we got our Component Analyzer back, and some quicky tests are making it look like the resonant frequencies of the inductors are nowhere near where the Manufacturer has them rated at in their specs. May have to give them a call. I've run into this (inductor specs issues) before where we cannot duplicate their numbers, and then come to find out they have this elaborate cicuitry tied into it all (that they neglect to mention) when they are doing their spec listings.
 
Top