Two of the older scrambling methods, before suppressed sync methods became popular, are done by using Notch Filters. These notch filters are used to remove unwanted carrier signals from TV channels, such as the annoying "beep-beep-beep". But, they can also be used to remove entire channels. Though Notch Filters are considered an old, archaic technology (and probably aren't used on any new systems), notch filters are still in use at many places around the country as the transition to modern gear hasn't been complete.
Notch filters that remove entire channels are called 'Negative' Filters. Notch filters that remove sideband carriers to make a channel viewable are called 'Positive' Filters. Thus we have two different scrambling methods. The difference in the two types of notch filters is in the exact positioning of the poles - that is, the tuning of the frequency (or frequencies) that the filter attenuates.
If the cable company suspects that more people will subscribe to a particular premium channel on average, they will use a negative filter at each subscriber location to remove the channel if the subscriber doesn't subscribe to that particular channel. To defeat such a filter, you must remove the filter from the line. However, if the cable company figures that less people on average will subscribe to a particular channel, they will add an extra signal half-way between the video and audio carriers that "drown-out" the main audio and video carriers. These are channels that have the annoying "beep-beep-beep" sound. When a subscriber purchases one of these channels, the cable company will add a positive filter at the subscriber's location to remove the unwanted carrier signal and provide a clean signal. To defeat this scrambling, you need only to build your own positive filter.
One of the major suppliers of notch filters for cable companies is Eagle Comtronics. Below are links for some of their documents that describe notch filter operation and selection:
Companies like Eagle Comtronics typically only sell notch filters to cable companies, so don't expect to just call them up and buy one. You might can trick them into thinking you are a cable company, but don't count on it. The best way is to probably just build your own. You can build your own from scratch, if you wish. But it is usually more economical to purchase a kit. I've tried several kits and the best I've found thus far is made by Rainbow Kits and is the DF-222 kit. As per the policy of this site, you can find the complete information about this kit here. Please use this information for informational purposes only. And if you decide to build this kit, I encourage you to purchase it from either Rainbow Kits or one of their resellers such as Jameco Electronics.
As the DF-222 Document mentions, you can use a sweep generator and a spectrum analyzer to tune your filter, but you can also use a Network Analyzer. Internally, a Network Analyzer is effectively a sweep generator and spectrum analyzer combined. But this combination yields a powerful test device. The Analyzer will show the transformation function of the device under test as well as input/output impedence, frequency response, and many other neat things. However, as you'll find out below, both don't yield much more than can be acheived by tuning by-hand from trial and error other than speeding up the time it takes.
I have built 3 positive filters, one for channel 14, one for channel 22, and one for channel 23. Details for building a filter for channel 23 is not included in the DF-222 assembly instructions, but if you extrapolate the capacitor and inductor values in the part selection table, you can easily choose parts that will work for channel 23. Problems come into play as you get higher and higher into the frequency spectrum, as the notch width becomes wider and wider and you end up blocking more of the signal than you want to. Click on the links below to view plots created on the Network Analyzer that I used in tuning each of the three:
Notice that the notch width gets higher as you move up to higher frequencies. Also, notice the extra sideband at 198.43 MHz on the Channel 23 filter created by poor part selection. This falls at the bottom end of Channel 11, but causes no noticable (visible) problems. Both the Channel 14 and Channel 22 filters also have an extra "blip" to the side. But again it isn't noticable.
The good thing to note is that these charts show the results after tuning totally by hand. That is, this tuning was done by trial and error from visually watching the channel for clarity and the center frequencies of all 3 were dead-nuts on! That means that having fancy calibration instruments and tools isn't necessary to create a usable filter!
If you do build your own filter, you'll notice that objects near the filter will affect its tuning, and it doesn't really matter what type of object, but the type of object determines how much it is affected. Therefore, if you mount it in a plastic enclosure, you must position the device in the exact position it will operate in and use a plastic tuning instrument. This is necessary because the tuning will change as you move it near-to or away-from other objects (including yourself).
The best solution is to mount the device inside of an all-metal enclosure with miniature holes to insert plastic tuning tools. This allows you to move the device after you've tuned it without affecting the tuning.
Another problem with any passive type device, such as these notch filters, is that they attenuate the entire spectrum a little bit. With only one or two filters on the same cable system, this doesn't mean anything and isn't noticable, as the losses are negligible. However, if you have 3 or 4 filters, channels that are transmitted weaker at the cable head-end, may become fuzzy when they get attenuated by the series of notch filters. These problems can usually be solved by purchasing a cable amplifier.