FMI: More than Just a Problem on the Fringe

Last year, this column explained how strong FM Band signals can overload ATSC receiver front-ends generating second-order intermodulation (IM) products. All second order IM products generated by signals in the 88–108 MHz band fall in channels 7–13.

Channel 10 is subject to much more FMI than channels 7 or 13 because so many IM products fall in channel 10. FMI shows up where the DTV signal on a high VHF band channel is weak, so the RF amplifier is at or near maximum gain. Strong FM band signals can then overload the mixer thereby generating second-order IM products.


This year, the FCC may take up to 120 MHz of broadcast TV spectrum for wireless broadband. There would be a repacking of broadcast TV spectrum, that much is certain.

The FCC expects some broadcasters to voluntarily surrender their license in order to share in spectrum auction proceeds; others may elect to share one of the remaining TV channels, splitting the available bit rate per channel.

Fig. 1: The relationship between the received power of a DTV signal and the total received FM Band signal power for NTIA-approved "converter boxes 1-13." The diagonal dashed black line has a slope of 2, the slope which would exist if the interference is due to second order distortion. An alternative to sharing a 19.39 Mbps bit stream has been proposed by the CTIA (The Wireless Association) to convert terrestrial TV transmission to a network of lower power transmitters distributed throughout the community. Tower height would have to be restricted to protect other stations allocated the same channel from co-channel interference—in short, a single frequency network (SFN).

Based on the lower effective radiated power limits, which would be imposed after repacking, and lower tower height above average terrain, the field strength at many sites well within the community would be, I believe, significantly reduced.

However, the field strength of FM band radio signals will not be reduced, so I expect the conditions under which FMI occurs to be more frequently encountered. It won't be just a fringe area problem. Field experience with FMI to date shows that a well-designed FM trap installed at the antenna "F" connector of the DTV receiver usually solves the problem if one knows this fact and acts accordingly.

Fig. 2 depicts the performance of the other units tested: #14-26. The spread between these converter boxes is almost 20 dB. However, traps don't solve the problem of receiving channel 6 DTV signals. At present, there are only a few channel 6 DTV channel allotments, but when plans for repacking are announced, I believe many broadcasters will become interested in a low VHF band allotment once again.

So perhaps some further study of FMI is justified. My colleagues and I have recently measured FMI into channel 11 with 26 NTIA-approved DTV converter boxes using a Rohde & Schwarz DTV Test Signal Generator. The results are shown in Figs. 1 and 2. Each shows 13 plots of the maximum FMI power for a range of desired DTV received signal power. That this interference is due to second-order nonlinearity is proven by the fact that the slope of these plots is 2 (shown as a dashed black line). This means a 10 dB reduction in the minimum DTV power for a 5 dB decrease in total received FM signal power.

In the Portland, Ore., market, we have 22 FM stations. The number of second-order IM distortion products of 22 signals is 225, distributed unevenly across channels 7–13.

Charles Rhodes is a consultant in the field of television broadcast technologies and planning. He can be reached via e-mail at