While the FCC DTV channel allotment process was based on the DTV Planning Factors in OET Bulletin No. 69, those Planning Factors were based on the assumption that each DTV station might face one, but not more than one, significant source of interference, and also on the assumption that receivers would not generate third order intermodulation products (IM3) that might fall in a local channel causing co-channel interference. This column has tried to alert the industry, receiver manufacturers and broadcasters to the interference that will result from receivers that generate IM3 in the presence of multiple undesired signals.
Before we get into the details of how combinations of DTV and unlicensed device (UD) signals could cause havoc, it may be worthwhile to review how one and two DTV signals can interfere with reception of another DTV signal on another channel. How do one or two undesired signals jump into the channel to which the receiver is tuned? This review will refresh those who follow this column regularly, and will be vital to new readers.
In January, Triplets of undesired signals will be discussed in grave detail.
(click thumbnail)Fig. 1: One DTV signal and IM3 productsA QUICK REFRESHER
In my last column, I noted that an undesired signal on say Channel 31 will “hop over” another undesired signal on Channel 34 and land in Channel 37 and visa versa. This “hopping over” in the frequency domain is the change in the frequency of a signal by receiver nonlinearity of the IM3. This is a significant problem with DTV receivers marketed in 2005 and 2006, according the FCC/OET Report 07-TR-1003, March 30, 2007.
Fig. 1 shows one DTV signal with third-order distortion: sideband splatter occurs on both first adjacent channels. The signal output power of the amplifier under test is –1.1 dBm per channel, while the third-order distortion is –41.4 dBm in each adjacent channel. No distortion product is above the noise floor in either second adjacent channel.
Fig. 2 shows two DTV signals on Channels 31 and 32, –1.1 dBm power each and their third-order distortion in channels first and second adjacent channels. The IM noise in each first adjacent channel averages about –33 dBm, 8.4 dB more than with a single signal.
Fig. 2 also shows noise in the second adjacent channels, –40 dBm. The total IM noise in one first adjacent channel and one second adjacent channel is –32.2 dBm. This exceeds the noise for the single channel scenario by 9.2 dBm, which is in good agreement with the theoretical value of 9 dB. Doubling the power into the nonlinear device increases its third-order distortion by 9 dB.
(click thumbnail)Fig. 2: Two DTV signals on adjacent channels and IM3 This is a huge increase in the noise under the signal in the desired channel. If the desired channel is either Channel 29 or 34, the noise under that signal is comparable to the noise one undesired signal puts into its adjacent channels as shown in Fig. 1.
Fig. 3 has two DTV signals offset by two channels. Fig. 3 is very important, especially where the desired signal (Channel 32) is surrounded by undesired signals on both first adjacent channels. The noise floor in the desired channel is higher than in any other channel. Fig. 3 also shows what this author has named “Bee-Hives” of third-order IM distortion products. These Bee-Hives are always three channels wide with the power in each side channel 6 dB below that in its center channel (29 or 35 in Fig. 3). If you look again at Figures 1 or 2, you will see that the IM3 there is also shaped like a Bee-Hive with the DTV signal superimposed. Bee-Hives are even more apparent in Fig. 4.
In Fig. 4 the offset is three channels and the noise power in the center of each Bee-Hive (Channels 28 and 37) is about –34 dBm. The noise power in the side channels of the Bee-Hives (27 and 29, along with 36 and 38) is lower by 6 dB. These figures show that IM3generated by a pair of signals is much stronger than the IM generated by one signal at the same power. This is because IM3 varies with the cube of the signal power. IM products act like random noise lowering the SNR when the IM falls in the desired channel.
(click thumbnail)Fig. 3: Two DTV signals and IM3 separated by one channel
This added power may be distributed in different ways between several channels depending on the channel-to-channel offset. In Fig. 4, there is significant noise power in all channels from Channel 27 to and including 38. This spectrum spreading of IM3 is extremely important. In Fig. 4, a desired signal on any channel from 27 to 38 is possible. In Fig. 1, the only IM3 falls in two channels only.
Theoretically the noise power in Channels 27 and 29 is 6 dB lower than the noise power in Channel 28. This is an important concept. As a hypothetical example, if your station is on channel 27, 28, 29, 36, 37 (Channel 37 has never been assigned for television broadcasting purposes), or 38 your signal may have its noise floor raised above what it should be (–99 dBm according to OET Bulletin No. 69). Where DTV reception is interference limited, the noise floor under the desired signal is mostly IM3.
In Fig. 4, the noise power in Channels 28 and 37 would be –33.9 dBm (averaged). To receive a DTV signal on Channel 28 or 37 the desired signal power must exceed –33.9 + 15.2 dB = –18.7 dBm. To receive a signal on Channel 27, 29 or 36 or 38, the minimum desired signal power must exceed –24.7 dBm. The minimum received signal power in a noise-limited channel (no interference) is –99 dBm + 15.2 dB = -84 dBm.
(click thumbnail)Fig. 4: Two DTV signals offset by three channels. Note the IM3 Bee-Hives.DTV-DTV interference is much more of a problem when there are undesired DTV signals on two channels than for one undesired signal. This interference spreads out as shown in Figures 2, 3 and 4, so that a given pair of undesired signals can interfere with many more than just one station. Now you know how to calculate the channel numbers of the centers of Bee-Hives, which may be lurking in your community, and you know that their side channels can also be subject to interference. Interference of this kind can be produced by the effects of corrosion of rooftop antennas and their connection to the downlead.
Most rooftop antennas still in use have been exposed to the elements for decades. In apartment houses there are distribution amplifiers fed by a rooftop antenna. Such amplifiers can overloaded and generate IM3 products. So it isn’t always the DTV receiver or downconverter that is faulty. I suspect that once off-air analog TV broadcasting ends (soon) broadcasters will experience complaints from the public concerning their DTV reception. Interference to DTV off-air pickup by CATV System headends may affect carriage of your DTV signal. Especially in that case, your knowledge of such problems can be extremely helpful.
Now you understand why signals on certain channel pairs can cause interference to DTV reception. If two is worse than one, three will be worse yet. Stay tuned.
