RF notch filter for suppressing intense interference

You probably remember my post about the impossibility of reception in the C range due to powerful out-of-band interference in the "Small talk" section.
I thought a lot about how to suppress this interference or the interference that I have at 10730 MHz and interferes with reception at the beginning and throughout the Ku range.
I thought for a long time, and recently calculated and made a rejection filter in five minutes, which is shown in the figure

{ Rejection 10700 MHz -24 dB }

I was counting on a rejection frequency of 10700 MHz, so as not to completely cut off useful frequencies at the beginning of the Ku range. In fact, it turned out that the interference on the original spectrum, which is drawn by the blue curve at a frequency of 10730 MHz, after the inclusion of a rejection filter in the gap of the antenna cable is suppressed by 24 dB (see the black curve on the spectrum)

{ Rejection 10730 MHz -24 dB }

At the same time, the local minimum of the noise level is at frequencies of 10640 - 10740 MHz.
And C/N at a frequency of 10390 MHz increased by 5 dB!

Please compare the results of the blind search during registration after the inclusion of a rejection filter in the antenna cable and before (without a filter) in the figure:

{ Before_&_after_Rejection }

{ Without_Rejection }

{ Before_&_after_Rejection }

Pay attention to the 11263 MHz transponder, whose "legs" are exactly at the noise level along the black curve, and the transponder lock reserve has increased by 2.9 dB!
And the SNR of the 11881 MHz transponder has increased from 16.9 on the original (blue) spectrum to 18.4 dB with a rejection filter, i.e. by 1.5 dB.
The 11895 MHz transponder, which was not locked without a filter, after the inclusion of a rejection filter received a confident lock with SNR = 17.4 dB!

This post is based on my article Notch Filter on my blog

Quote from Notch Filter

https://filatov-yuri.livejournal.com/2025/02/15/

. . . Notch filter of frequency 10724 MHz . . .
======================================
filatov_yuri
March 14th, 14:59

14 March 2025 - Филатов Юрий ака СТРАННИК

14 March 2025. Спутниковое телевидение

filatov-yuri.livejournal.com

Vertical polarization spectrum with intense (-35 dBm) interference at frequency 10724 MHz

Intense interference at frequency 10724 MHz (intermediate frequency 974 MHz) creates a strong background of intermodulation interference, which degrades the reception of satellite signals in the entire Ku range.

To combat this interference, a filter was designed and tested from two sections of coaxial cable, an RF tee and three F-type connectors.

A rejection filter is connected between the satellite antenna drop cable and the PCI-exp card input. The spectrum obtained with this setup with the results of a blind search is shown below.

Vertical polarization spectrum with a rejection filter, which suppressed intense interference by 17 (-24) dB

Interference suppression at a frequency of 10724 MHz was 17 dB, which had a very favorable effect on the SNR of useful signals. It increased by 0.2-1.9 dB. In addition, a signal from two more transponders was additionally received, one of which with SNR = 9.7 dB!

Below you can compare the spectra WITHOUT a notch filter (on the right) and the same frequency ranges recorded WITH a notch filter (on the left):

Beginning of the Ku band — spectrum with a notch filter on the left, spectrum without filtering on the right

SNR gains (from left to right) were 0.2, 0.3, 0, 9.7 and 0.9 dB

Middle part of the Ku band — spectrum with a notch filter on the left, spectrum without filtering on the right

Here the gains are 0.4, 1.0, 1.9, 0.7, 0, 0.5, 0.3, 0.7, 0.1, 1.0, 1.5, 0.5, 0.3, 0.5, 0.6, 0.9, 0.9, 1.0, 0.9, 1.1 dB

End of the Ku band — spectrum with notch filter, on the right is the spectrum without filtering

At the end of the range, the gains are 0.5, 0.3, 0.5, 0.6, 0.9, 0.9, 1.0, 0.9. 1.1, 1.2, 0.8, 0.6 and 0.7 dB.

My costs for making the notch filter were only $0.57

Quote from satesco

What do you say?

What can I say?
Taking money from old friends and colleagues is not in my rules...
As one wise fisherman said when he was asked for a fish:
"I will give you a fish, but you will be full for one day.
I will give you a fishing rod, and you will be full for life"
I will describe the design of the rejection filter and help calculate and adjust it, and you yourself will cut two pieces of coaxial cable with a known coefficient of electrical shortening, screw on three F connectors and close the appendix with a metal cap from a ballpoint pen.
First, please make a spectrum of the working frequency range with the spectrum of the intense interference that you want to suppress in IQmonitor. And publish it here, in the topic.

Quote from SaTom

A few years ago I read a Hungarian Satellite DX-r book or magazine published in the 1990s, and which this very solution was proposed and described, for KU band applications.

Quote from satesco

I was wondering if it would be possible to make one for me, for a fee, including shipment.
What do you say?

I offered to make a filter together and for free. To begin with, I asked to make a spectrum of the range of interest to you, on which the interference would be visible:

Quote from strannik

First, please make a spectrum of the working frequency range with the spectrum of the intense interference that you want to suppress in IQmonitor. And publish it here, in the topic.

In IQmonitor, unlike other similar programs, there is a gray scale at the bottom of the spectrum, which is not striking, but which displays RF frequencies regardless of the range of received waves, the frequency of the first heterodyne and the method of conversion (up or down) of useful signals.
These frequencies (the beginning and end of the operating range), as well as the frequency of intense interference at the Z receiver input, are the basis for calculating the length of coaxial lines in the rejection filter.
If for some reason you change your mind - let me know - we will wait until someone is really interested in this (no longer mine) problem.
femi
Many thanks for the link to the pdf file. Unfortunately, it is copy-protected and cannot be translated into Russian... And I do not know German, although I had two graduate students from Germany on an internship (Frank Volkersdorfer and Catherine Schöne, with whom we communicated in English).
I was pleased to look through the articles that are my peers and which I see for the first time. I looked through the drawings, graphs and formulas - they fully correspond to the classical theory of wave propagation in lines.

Quote from satesco

You overwhelm me, dear strannik, with your generosity, in the not-at-all peaceful conditions in which you live.

Я ценю Вашу иронию по поводу небывалой "щедрости" в условиях военного времени, ведь Вы наверняка читали

Quote from strannik

My costs for making the notch filter were only $0.57

57 центов - это копейки в нашем понимании.

Гораздо хуже, что Вы не понимаете, что я от Вас хочу - СПЕКТРЫ полезных волн и волны-помехи как функцию мощности сигнала от радиочастоты на входе Вашего приёмника. Знаю, что у Вас есть программа IQmonitor, с помощью которой это делается за пять минут, и Вы нею умеете пользоваться. Не хотите пользоваться - насильно мил не будешь - сделайте спектры на любом другом программном продукте или спектро анализаторе. Вам для расчёта заградительного фильтра нужны именно RF частоты. И выберите, пожалуйста, один объект для наших показательных расчётов режекторного фильтра

Today I reverse engineered the "silver" notch filter to show its design and, at the same time, to increase the notch frequency. First, I unscrewed the quarter-wave section of the filter (1) from the tee:

Then I removed the protective cap from the ballpoint pen as shown in photo (2) :

The open end of the coaxial cable (I used a Finmark 690 cable, in which the insulator is foamed polyethylene, with a shortening factor of 0.86) is wrapped with black electrical tape to prevent the central core from shorting to the cable braid, and it to the outer screen.
In the figure (3) the insulation is unwound, and two marks with a black felt-tip pen are visible on the cable:
- the first is at the calculated distance from the bare end of the central core and along it the cable was cut after winding the F connector;
- the second is applied exactly according to the size of the cable, which is hidden under the blue cap of the first copy of the rejection filter, the rejection frequency of which is higher and is at the bottom of the Ku range.
Finally, you can see in the figure (4) how the cable was cut with metal scissors. The cable deformation was eliminated by finger-crimping, and before insulating with electrical tape, the absence of a short circuit of the foil or wire braid hairs on the central core of the cable, which should be strictly on its axis, was checked.
Having insulated the end of the cable and inserted it into the handle body, I screwed the quarter-wave section of the rejection filter onto the tee.
Having registered the spectrum, I made sure that the rejection maximum had shifted up in frequency, remaining below the beginning of the Ku range.

In general, the spectrum has changed for the better and has become noticeably cleaner, and the SNR of many transponders has increased, especially in the area of the 3rd and 5th subharmonics:

The strongest effect was observed for horizontal polarization:

A pleasant surprise for me was the 11069 H transponder, which I had not seen before in the shadow of its powerful neighbors...

Today I reconnected the antennas:
- the 90°E antenna, which was connected via the "blue" notch filter to the A input of the TBS 6983, was connected via the "silver" notch filter to the B input of the TBS 6983;
- the 42°E antenna, which was connected via the "silver" notch filter to the B input of the TBS 6983, was connected via the "blue" notch filter to the A input of the TBS 6983
I loaded the spectrum of 42°E registered yesterday from the database and displayed it on the tablet in gray, as a basis for comparing the spectra. I registered the spectrum of 90°E in blue on it, and then registered the spectrum of the vertical polarization of the 42°E satellite in purple:

From the comparison of the spectra it clearly follows that the interference waves in the RF frequency range of 970 - 1010 MHz are interference on the cable from the antenna and they are completely suppressed by the "blue" rejection filter. "Blind" scanning of the satellite position 42°E was carried out, which confirmed the high efficiency of the filter, especially when receiving on the 3rd and 5th subharmonics of the internal heterodyne, as well as on horizontal polarization: