The Case for Preamplifiers
It seems to me that the case for using a preamp is
any case where there is a fringe signal.
However, I acknowledge that a set of problems do often arise when there are also very strong signals present. I had one visitor to my
Antenna Hacks website who lived just a few miles South East of Detroit and was trying also to pickup a UHF station from Cleavland about 50 miles away. He lived right on the line of sight from Detroit to Cleavland, and was using a CM4228HD for both VHF and UHF signals, aimed at Cleavland. Given his proximity to Detroit, the backside response was sufficient for the Detroit stations of interest. But he could not insert a preamp to strengthen the Cleavland signals because it overloaded from the strong Detroit signals. E.g., all of his channels degraded. This happens when intermodulation distortion sets in as a result of over driving an amplifier.
When I looked at his TVFool report, the differences in signal levels from his Detroit and Cleavland signals
exceeded 50dB in a number of cases! Even with the reduced backside gain the strongest signals were overloading his preamp.
I have to wonder if the observations cited by
haykarenhardy might be the result of some combination of strong and weak signals, and degradation as a result of an over driven amplifier? Mainly:
- One station gets worse when amplifier installed,
- Other stations disappear when not installed.
Now, its not completely clear to me that keykarenhardy was using the amps cited independently or in series, but none are particularly impressive from a Noise Figure standpoint (2.8dB for the preamps, and 4.5dB for the line amp (I've never seen a "line" amp with a respectable NF. I would be happy to learn of one). Such Noise Figures are acceptable whenever you have sufficient margin to begin with, and maybe your're just trying to push the signal through the loss of the coaxial distribution system. But where I've resolved really fringe area ("cliff") issues in the past I've had to resort to one of the Kitz Technologies 1dB NF amps, or the newer 0.4dB design, or one of the (very expensive) Research Communications models, in order to fully optimize.
But, back to the issue of strong and weak signals; there are two ways to mitigate a large difference in signal levels:
- Attenuate only the strongest, or
- Amplify only the weakest.
Both approaches require filtering and depend on serendipidy. E.g., how close the strongest and weakest signals are to each other, and/or to other signals of interest.
Attenuating the strongest signal(s) requires a notch filter, and is the easiest to engineer. Whether it is practical or not depends on whether or not weaker adjacent signal would also get unacceptably attenuated. You can construct a parallel resonant LC filter inside a die cast aluminum box with inexpensive parts. You can use only the inductor and capacitor and get a pretty narrow band filter, or you can dampen the Q with a resistor to get a broader band and hit several channels potentially (at the expense of less attenuation), and you can also cascade them to hit channels that aren't so adjacent.
You then just run the main signal through the filter and then to the amplifier, thus decreasing the level of the strongest signal before amplification.
Amplifying only the weakest is requires a band pass filter and is more complex. Its really only practical when using a separate high gain antenna (such as the CM4228HD or 91XG) for the weak signal. We can then filter out only the desired frequency, amplify it, and then just use a common splitter/combiner to insert the 20dB or so hotter single channel with the main antenna.
If you try to do this with a single antenna its gets very tricky and I don't normally recommend it. In concept, you need to isolate the single channel of interest, amplify it, and then recombine it with the rest of the signals without creating a feedback loop from amplifier output back to its input, and without degrading the signal to noise ratio of the weak signal. It is not possible to accomplish either of these with common splitter/combiner devices.
Now, there was a marvelous device made by Channel Master called a "JoinTenna" that can be successfully employed here, as well as for the attenuating scenario above. The JoinTenna is a 3-port device with two inputs labeled "All Channels" and "Channel N" (where N is a single channel of interest), and one output port labeled "To TV". On the "All Channels" port is a notch filter attenuating Channel N, and on the "Channel N" port is a band pass filter only passing Channel N. Both ports have losses around 1dB, significantly less than splitter/combiners. You can google for JoinTenna and maybe get lucky and find one tuned for your channel of interest. Also, quite a few of them can be retuned to nearby channels with the use of an oscillator, and signal level meter, or better yet a spectrum analyzer with tracking generator. The only off-the-shelf alternative to these are the filters from Tin Lee, which are much more expensive, but very good.
The magic of the JoinTenna is that it is symetrical. I.e., the passive filter characteristics allow it to be used either as a comibner, or reversed and used as a splitter.
That means you start by connecting the antenna to the "TV Set" port of one device to be used as a splitter, and connect its "All Channels" port to the second device's "All Channels" port; the second device to be used as the combiner. Now take the "Channel N" port of the splitter and connect it to amplifier input, and the amplifer output to the "Channel N" port of the Combiner device. There is sufficient attenuation on the "All Channels" ports to avoid a feedback path around the amplifier, and the approximately 1dB of insertion loss ahead of the amplifier can be overcome with a very low noise preamp (like the KitzTech).
You can also use just the "All Channels" port for the "Attenuate the strongest" approach, inserting it between antenna and amplifier, and mistuning it somewhat off the offending channel. That is, attenuating the offending signal only partially.
Alternatively, I have successfully built and deployed DIY filters using little die cast aluminum boxes, a little brass sheeting for a solid ground (internal to the box), a little 18AWG enamel coated copper wire to serve as inductor, a 2-10pf variable cap (for lower UHF fequencies, less for higher freq, more for lower freqs), and a couple of panel mount F-Connectors. You can make notch filters (a parallel resonant configuration), band pass (series resonant), and insert them in the signal path to reject or select specific frequencies. In some cases you can cascade multiple notch filters in series to reject multiple channels, and cascade multiple bandpass filters in parallel to pass multiple channels. But there are interactions when you cascade and things can get tricky; you'll need good test equipment.
But What Happened To Mr. Detoit?
The serendipidy here was that the offending strong stations were VHF. The back side of the CM4228HD is almost as hot as the front side (poor F/B ratio at Hi-VHF) so facing it away from Detroit towards Cleavlend didn't reduce the VHF very much.
But: the main thing here was that they were VHF and the desired station was UHF. For this scenario there are common $3 VHF/UHF splitter/combiners (like the "UVSJ" model) that can be employed to implement the "Amplify the Weakest" approach.
Using one UVSJ as a splitter, he connected the antenna to the common port, connected the two VHF ports of the UVSJs directly together, took the UHF signal from the splitter device into his ampifier, and the output of the amp to the UHF port of the combiner device.
Viola! With the VHF stations removed from the amplifier input he had is Cleavland station rock steady!
Hope the info useful...
t
