Passing only a narrow band of frequencies -- sounds easy enough. Well Not!
I we fail to design the Band Pass Filter (BPF) properly we risk having out of band signals, spurs and maybe a problem with the FCC. This is where a simulation tool like LT Spice is worth its weight in gold.
The Band Pass Filter (BPF) is such a critical circuit element yet often is done in a haphazard manner. At a bare minimum you only need 7 components consisting of five capacitors and two inductors. Now that sounds pretty easy, but the problem begins because often the values may not be integers or standard parts. For instance you may have a section coupling capacitor of 1.5pF or an inductor like 2.304uH.
So what happens -- not having an extensive junk box, you slap a 5pF cap in there and what you actually wound on the toroid is 3uH. Sure it will limit the signals outside the band pass but the range may be way off, and the response curve resembles a camels back!
I often build my BPF's so that they will favor the SSB portions of the bands as that is mostly where I operate and then a homebrewer builds my radio and slides down to FT-8 where he (or she) sees a drop in output. My answer -- it is a really good BPF for not too far away from the SSB portion of the band the BPF is kicking in to play.
Thus those 7 components in a simulation operate in the range you want, but what may get installed is not the same as in the simulation. That is when the problems start.
The next email to me starts off by saying I used my new $30 Nano VNA and your design sucks. The curve is awful and the range is wrong. Putting in a too large of a section coupling cap causes a double hump (5 versus the 1.5pF) and the higher value of toroid will shift the Center frequency downward.
Still another piece is the tolerance (and quality) of the capacitors outside of the section coupling capacitor. In a fit of not thinking carefully I bought a box of caps from Banggood. (There you go with those curious Chinese names of companies -- back in my Navy days that name meant something entirely else especially in Cebu.).
So having an issue with a BPF I went back to measuring every cap in that assortment -- none were even close to the value printed on the component. I trashed them and chocked that up to experience -- bad experience.
While many abhor surface mount capacitors, as that is a part that is hard to see and hard to use. If you only build ugly style then indeed a problem! SMD capacitors can be found in close tolerances and in the odd ball values called out for in the design. If placed in a lower level stage they will not have an issue such as running 50 watts through the little cap.
I tend to use either high quality trimmer caps (Jameco Electronics) or trimmer caps in conjunction with an SMD part. You can get close just by listening to the band signals and then as a final tune it is OK to now connect your Nano VNA assuming it is terminated properly with 50 Ohms. In the photo you can see the 50 Ohm termination for measurements.
If you really want to understand how to build a filter then put away your Nano VNA (you can use it later as described above) and find a copy of SSDRA by W7ZOI. One of the Appendix topics is scratch designing a BPF. You learn how to set the bandwidth and all about the component dependencies. Then you can take that to LT Spice and create solid BPF designs for your homebrew rigs. Then get out your Nano VNA and measure what you have built.
The are several different forms of the BPF and we have the two section above. W7ZOI in SSDRA covers two and three section BPFs. For those who have dabbled with the Bitx rigs Farhan often uses a form of BPF that looks like these.
A Plot of the 3 Section 40M Filter~ It actually covers the 40M band!
Now something significant about the multi-section filter above and comments made in a recent soldersmoke podcast by Bill N2CQR. Bill was relating that in his 10/15 homebrew rig there was some issue with the BPF not attenuating certain signals higher than the operating frequency.
The answer suggested by Farhan is shown in the plot. Note the steep rise in the BPF response curve for frequencies BELOW the band pass. But the attenuation is not as good for frequencies ABOVE the band pass. It is possible to adjust the circuit constants to reverse that curve so that the steeper response is on the higher side and not the lower side. I think Bill did that and the problem was solved. Again this is where LT Spice is worth its weight in gold.
Complete Blog Post on this from Bill.
https://soldersmoke.blogspot.com/2023/09/crushing-17-and-12-meter-spurs-with.htmlhttps://soldersmoke.blogspot.com/2023/09/crushing-17-and-12-meter-spurs-with.html
So today we covered Band Pass Filters and for only 7 components have caused some to take up drinking or going bald. The way forward is to take the time to really learn how the BPF is designed (SSDRA) and how to simulate the design and the all important build to print. A tool I would buy before getting a Nano VNA (and have one) is a high quality LC Meter such as the AADE (no longer made).
I am often surprised that you wind 10 turns on a toroid in a haphazard, jumble wound, closely spaced array versus 10 turns carefully wound and evenly spaced around the core are different inductance values! Still 10 turns but not the same inductance!
73's
Pete N6QW
PS: For those who love to nit pick, in the photo of the PSSST BPF, the section coupling cap consists of two close tolerance, high quality 3PF Capacitors in Series that measure awful close to 1.5pF.