Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now


HD Radio Capture Issue Is Not Always Predictable

O’Neal and Kean examine the vagaries of this phenomenon

Text has been updated to correct a typo in the call sign of WDCX-FM

In a special report late last year we explored the fairly recent discovery of inadvertent “capture” of out-of-market HD digital radio signals by receivers tuned to local stations on the same frequency but operating in analog. 

Two cases were documented: a Class B FM in Buffalo, N.Y. and an LPFM licensed to Asheville, N.C. In both cases, HD-equipped radios tuned to the respective frequencies of these two stations (not transmitting a digital carrier) were forced into their HD reception mode, with programming being delivered from stations in the neighborhood of 100 miles away.

The Buffalo incident was transient, lasting only while the station licensed to that market, WDCX-FM, had to switch off its HD MP1 transmissions due to an equipment problem. In the case of Asheville station WPVM(LP), the capture by Class C station WSOC-FM in Charlotte, N.C., occurs frequently, as the Asheville station does not offer HD broadcasts.

I traveled to Asheville to experience and investigate the HD capture occurring there. Broadcast consulting engineer John Kean became interested in the situation and offered to do computer modeling of stations in other markets where listeners with HD-capable receivers tuned to a non-HD local FM station might encounter the same sort of unintended HD signal capture or “hijacking” experienced in Buffalo and Asheville. The results of these selected test cases were not as expected.

Predicted “capture zones” for HD Radio receivers tuned to LPFM station WLMP(LP) in Fredericksburg, Va., which is co-channel at 102.7 MHz with Baltimore’s WQSR. The latter transmitter is north and east of Fredericksburg. (On all maps in this article, the darker-shaded areas — orange and red — indicate areas where the strongest potential for unintended HD capture exists.)

All clear in two Virginia locations

One of the models indicated a case for capture of HD receivers tuned to 102.7 MHz in the Fredericksburg, Va., area. WLMP(LP) is licensed to Fredericksburg with an ERP of 100 Watts and an antenna height of 27 meters above average terrain. The projected potential unintended capture source was WQSR in Baltimore, Md. It is a Class B station, operating with an ERP of 50 kW and an HAAT of 133 meters. The cities are about 82 airline miles apart.

I equipped my 2016 Chrysler Town and Country minivan with an outboard HD receiver and a separate FM “magmount” antenna to facilitate switching between analog FM reception from the vehicle’s receiver and the audio output of the HD unit. The HD antenna’s base was about 70 inches (1.8 meter) above ground, as compared to the fender-mounted factory antenna’s 45 inches (1.1 meters). 

(Increase in height above ground of the HD receiver’s antenna was expected to provide a slight gain in reception capability. This was borne out by switching back and forth between the “in-board” receiver and the HD receiver operating in analog mode while tuned to a weak station. The “outboard” antenna and radio provided consistently better reception of weak signals. This should have made it easier for out-of-market HD signal capture.)

The Chrysler minivan used in the testing around Fredericksburg and Charlottesville, Va. The mag-mount antenna used with HD reception equipment is at the top of the vehicle and the factory-installed antenna feeding the vehicle’s analog receiver is visible at the left of the outside rearview mirror.

To verify the presence of an HD signal from WQSR, I tuned both receivers to 102.7 MHz at a distance of about 50 miles north of Fredericksburg. WQSR delivered very usable analog and HD signals at that point. As I drove south, I noticed spotty reception of the Fredericksburg LPFM about eight to 10 miles from Fredericksburg, with the low-power station fully capturing the receiver about five miles from the city’s downtown. The WQSR HD signal disappeared slightly before the LPFM’s reception was first noted. 

After maneuvering as close as possible to the predicted HD capture points on the computer model, I observed no capture. I spent a couple of hours driving around the Fredericksburg area while monitoring the HD receiver output. All that I heard was the analog signal of WLMP, which came in very loud and clear at all times.

Later, Kean and I traveled to Charlottesville, Va., some 60 miles southwest of Fredericksburg, to investigate potential unintended HD capture cases there from Washington, D.C. stations. The Charlottesville and D.C. stations are separated by about 100 miles and computer modeling suggested HD capture was likely on several co-channeled stations due to power levels being transmitted from the D.C. FMs. 

Models indicated the best probability of capture existed among the following co-channels stations (Charlottesville stations are listed first followed by D.C. stations):

  • 88.3 MHz: WVTU (0.195 kW ERP) and WPFW (34 kW ERP)
  • 88.5 MHz: WVTW (1 kw ERP) and WAMU (47 kW ERP)
  • 96.3 MHz: W242CL (0.01 kW ERP) and WHUR (16.5 kW ERP)
  • 97.1 MHZ: W246DD (0.099 kW ERP) and WASH (17.5 kW ERP)
Possible capture zones in the coverage area of Charlottesville, Va., translator W242CL on 96.3 MHz by co-channel WHUR in Washington, D.C.

Charlotteville stations operating on 88.3 and 88.5 MHz were ruled out in terms of distant signal HD capture, as these co-channeled lower-powered FMs were also operating in HD mode. 

Again, my Chrysler Town and Country minivan was used as the mobile test vehicle, but this time it was also equipped with both a Sony XDR-F1HD FM tuner, which is recognized as a sensitive FM and HD Radio receiver, and an Audemat FM Navigator 10/100 for analysis of a station’s signal and for establishing a measurement point via an internal GPS receiver. Either of these devices could be connected to the rooftop mag-mount FM antenna and used for HD reception. Analog reception was accomplished via the van’s factory-installed radio

The Sony XDR-F1HD tuner and Audemate used in the Charlottesville observations are seen in this picture. The minivan’s factory-installed analog-only receiver, also used in the driving tests, is visible in the center of the dash.

Our listening tests began approximately 10 miles northwest of Charlottesville, with the vehicle being driven towards that city on U.S. Highway 29 and continuing at locations in and around Charlottesville.

The Washington-area stations were received (albeit not fully quieted in analog mode) as we neared Charlottesville, with the co-channeled lower-powered stations there eventually replacing them with solid (analog) reception. At no point in the routes traveled did the anticipated HD capture occur.

Predicted areas of HD signal capture of Charlottesville translator W246DD on 97.1 MHz by co-channel Washington station WASH.

No worries in east Texas

The next bit of field testing took place in a region very different from the rolling hills of northern Virginia and the mountainous real estate dominating western North Carolina. 

In east Texas, FM station owner Chuck Conrad, mentioned in our earlier article, agreed to participate in the testing and was sent the HD Radio receiver and magmount antenna used in the Virginia trial to ensure there were no variables in receiving capabilities. (Conrad used a Chrysler Town and Country minivan identical to the one in the Fredericksburg testing.)

Kean’s computer model predicted an even better case for HD capture of Conrad’s KZQX(FM), which is licensed to Tatum and operates on a frequency of 100.3 MHz with an ERP of 2.35 kW and an antenna HAAT of 158 meters.

KZQX, a Class A station operating on 100.3 MHz and licensed to Tatum, Texas, was expected to be captured by co-channeled Dallas market Class C HD Radio station KJKK as indicated by this map. Tatum (KZQX) is located in the lower right-hand corner. However, extensive field testing revealed no unintended HD receiver capture.

The co-channeled Class C HD station, KJKK(FM), is licensed to Dallas and operates with a 100 kW ERP and an antenna HAAT of 574.2 meters. The cities are about 135 miles apart.

Conrad was expecting the worst, as he had received a listener report of an apparent HD capture last December. He notes that due to the open terrain and proximity to the Gulf of Mexico, VHF “skip” from “sporadic E,” ducting and other natural phenomena are common beginning in late spring and continuing through the summer months. 

However, after driving to the projected areas of HD capture from the Dallas station, he observed no such behavior, with his station’s analog signal dominating.

“At no time did the HD display light up with some other station’s data, nor did I get anyone else’s audio,” said Conrad. He said he also listened for out-of-market co-channel HD signals on the frequencies on several of the FM translators that he operates, with no indication of any distant station capture.

“My conclusion is, at this point, there is no problem with current HD levels. What will happen if the HD sidebands are increased is anybody’s guess. I can see a potential problem, but we won’t really know until someone actually fires up a new higher-power digital signal,” said Conrad.

“Perhaps the good news is not every station broadcasting in HD has the ability [to increase digital power] with their current transmitter. Whether management thinks favorably about investing in more powerful transmitting equipment is a good question. From what I’ve been able to tell, the major reason for HD is to pair it with an analog translator, thus giving the owner another signal without violating any existing ownership rules. If that is indeed the case, there is little to be gained by being an early adopter of a higher-power HD signal as long as your signal chain is working. I know If I were in that situation, I would have a ‘wait and see’ attitude.”

Another projected HD capture “hot spot” is not far from Conrad’s base of operations. This is FM translator K231DK attached to AM station KERO in Longview, Texas. It operates on 94.1 MHz with an ERP of 250 Watts and an antenna very close to the ground (0 meters HAAT, according to FCC records). The potentially interfering HD signal is that of Fort Worth’s KLNO, a Class C running 100 kW ERP from an antenna HAAT of 485 meters. The cities are about 150 miles apart.

A projection of potential areas of HD Radio capture of 94.1 MHz FM translator K231DK, which serves Longview, Texas, AM station KERO. The out-of-market station is Class C KLNO, which serves the Ft. Worth-Dallas area. Despite the strong potential for capture of the 150-watt translator’s signal, no such activity was observed.

Translator owner Scottie Rice was provided with a map of the projected HD capture areas and agreed to do a “field test” to see if the out-of-market HD signal “took control” of his vehicle’s HD receiver. Just as in Conrad’s trial, Rice reported no problems. Sometime later, Conrad performed independent listening tests in and around Longview and verified Rice’s findings.

Back in North Carolina

Just to complete the picture, the operator of the Asheville, N.C. LPFM that we identified as having a signal unintended HD capture issue, Davyne Dial, was asked to monitor the frequency of another LPFM, WRES, licensed to Asheville on 100.7 MHz, for potential issues from co-channel HD station WUSK, licensed to Cleveland, Tenn., and serving the Chattanooga market.

WRES(LP) operates with an ERP and an antenna HAAT of 30.7 meters; WUSK transmits at 100 kW from an antenna HAAT of 363 meters. The stations are separated by about 130 miles, slightly greater than the distance from the Charlotte, N.C., HD station that is capturing her WPVM(LP).

Dial was provided with the Asheville regional map with projected potential areas for HD 100.7 MHz captures. While she observed none on this frequency, she did report that digital radio is alive and well in that region of the country, with other distant HD signals easily receivable on her Mini Cooper’s radio.

Other areas

Following the appearance of the first part of this story, reports of other HD Radio captures have been received, including in or around Denver; Reno, Nev.; Lafayette, La.; and Los Angeles. 

Denver-market station KKPL, licensed to Laporte, Colo., operates on 99.9 MHz with an ERP of 13 kW. Co-channel HD station KVUU, licensed to Pueblo some 185 miles distant, operates with a 79 kW ERP).

Projected areas of capture of receivers tuned to Laporte, Colo., station KKPL, near the top of the map, by co-channeled Colorado Springs-area HD Radio station KVUU, in Pueblo at bottom. Field observations did not correlate with predictions.

Conflicting reports have been received about this case, with an individual who asked not to be identified stating that the HD capture was primarily in the Greely area east and south of the station but that it was dependent on the particular vehicle and radio, with the effect being observed in a BMW primarily south of the Laporte station. However, when monitoring via a Toyota truck radio, the captured continued a significant distance to the north of the stations, as far as Cheyenne, Wyo.

However, another observer, Fred Baumgartner, reported minimal capture while driving south from the Laporte station in the direction of the Pueblo area.

“The area in which the unintended HD capture is occurring is somewhat questionable,” said Baumgartner. “The only place where I observed it was in the ‘interference zone’ between the two stations where their signals were about equal.”

Baumgartner stated that he experienced the normal (analog) FM capture ratio effect, with the slightly stronger signal being received in preference to the weaker one.

“I would have thought that the HD station would have the advantage,” he said. “However, the HD doesn’t lock up immediately. The FM capture kicks in first, and then the HD Radio capture occurs. I don’t think there would be any real impact on a listener, but to a broadcast engineer trying to resolve an interference issue, it could be a problem.”

South of the border

Another case of receiver capture by an out-of-market HD Radio signal was reported by Stephen Lockwood, president of the Seattle Wash.-based Hatfield and Dawson consulting firm.

He said the incident happened several years ago and involved a Mexican broadcaster client, XETRA, which broadcasts from Tijuana and serves the Tijuana and southern California area, including San Diego. 

“We got a call about an interfering signal on XETRA’s 91.1 MHz frequency,” said Lockwood. “I made the trip down there and drove all over the market with their chief engineer. We spent the whole day looking for the interfering signal with a spectrum analyzer but didn’t find anything. I told them that the next time this happens, take a picture of the radio display and that will help us figure out what’s going on.

“A month or so later, the XETRA manager contacted me to report the same problem. I looked at the picture of the display and realized that it was an HD Radio signal. The receiver was doing a really nice job of receiving it.” 

Lockwood recalled that it took a while to figure out what was happening, as the display seemed to indicate the signal was from a station licensed to Los Angeles — KUSC — but which was not co-channel with XETRA, operating instead on 91.5 MHz. The mystery eventually was solved when Lockwood found that KUSC programming is carried on a number of outlets in Southern California, including KDSC, licensed to Thousand Oaks and operating on 91.1 MHz. 

“The signal was ducting down the coast,” said Lockwood. “One minute you’re listening to 70s rock [on XETRA] and then all of a sudden you’re getting classical music. It was an interesting problem, as the co-channeled Thousand Oaks station was operating legally. The only thing we could advise was for XETRA to put an HD signal on the air.”

Lockwood noted that the ducted analog signal from KDSC was not usually strong enough to achieve conventional FM capture in receivers tuned to 91.1 MHz in the San Diego area. 

What’s going on?

Consultant John Kean provided the computer modeling for these listening tests. (He also created a bench setup for ascertaining the desired analog FM to co-channel HD Radio signal ratio necessary to trigger digital decoding circuitry in an average HD receiver. This setup is described in the accompanying sidebar below this story.)

I asked him for his thoughts as to why this unintended HD capture phenomenon is so difficult to pin down. 

The Buffalo incident and the 103.7 MHz LPFM case in Asheville are real. Yet these don’t seem to occur in other markets, even though the computer models make a good case for such behavior.

“Since the capture effect may be less frequent than [the Buffalo and 103.7 Asheville cases] show, real-world conditions may be different,” said Kean. 

“For example, I used 10 percent time variability for both the diagram and the ITM field strength predictions. These are, of course, for 10 percent of the time, averaged over a long period, such as a year. So, the real interfering signal could be different from day-to-day and even hour-to-hour.”

He added that while the median or 50th percentile of the time variable could be used in the calculation, it’s always been assumed that 10 percent is “plenty” for interference conditions.

“Since the capture effect is always done in motion, it could be that the Rayleigh fading reduces the HD Radio receiver’s ability to capture a signal by a number of dB,” Kean said. 

“However, this was considered in the link budget calculations to derive a predicted threshold. My bench test was not done with an RF channel simulator. I have one, but not enough signal generators to set it up, so my 25 dBµ measurement might have been a few dB higher. Nevertheless, it should occur below 30 dBµ.”

(Kean noted that he had calculated contour distances for three classes of FM stations, “A,” “B,” and “C,” relative to that of a Class A, with all of the pairs situated at minimum allowable separation distance for co-channel operation. He noted that from his calculations, it’s apparent that a field strength of 25 dBµ is the threshold for digital reception with a suitable antenna and height above ground; however, this is something that needs to be verified in actual automobiles.)

“There is definitely a need for more field testing to compare with the calculations and bench testing,” said Kean. “The problem with relying on real car radios and non-technical users is that the variables are too great and lots of tests would be needed to get a reasonable average.”

So, as we wrote in the earlier report, the jury still seems to be out. A lot more testing needs to be done to determine if this issue of HD capture is due to some freak of nature in the case of Asheville’s WPVM(LP), or if it is manifesting itself in other areas of the country among other co-channeled HD “haves” and “have nots.” As noted earlier by Stephen Lockwood, if this is a consistent problem for an individual station, the desired station could resolve it by employing an HD signal on the air.

We would like to hear about your experiences with unwanted reception of HD signals from out-of-market stations. Please document and email your findings to Radio World at [email protected]

Sidebar: Measuring the Threshold for HD Radio Capture 

As described above, consultant John Kean worked with me on the potential for an out-of-market FM station operating with an HD Radio digital signal to capture a receiver tuned to “local” co-channeled station. Kean devised a setup for measuring the exact amount of signal necessary to activate the digital reception circuitry in an FM receiver.

Here is how he describes it:

For the HD Radio station signal I used an active “Mini-Whip” antenna outdoors, having a whip length of about 5 inches and a high-impedance (FET) input. The FET was coupled to an emitter-follower configuration to provide a match to the coax used to feed the off-air signal to the Sony XDR-F1HD tuner on my bench. The outdoor antenna’s height was about 1.5 meters above the ground to simulate the height of a typical car radio antenna. A Tiny SA spectrum analyzer’s built-in oscillator provided the simulated co-channel FM signal.

Using inline attenuators at the XDR-F1HD input I determined that the threshold for HD Radio (hybrid digital) reception from a local station — WAMU(FM) operating on 88.5 MHz with HD Radio — was a consistent −96 dBm for the HD Radio signal only (with no added noise or interference), with no digital decoding occurring at −97 dBm, and solid reception at −95 dBm. (When the receiver stopped decoding and switched back to the analog FM service, the sound was moderately noisy — even with mono reception — at this very low signal level.)

This signal power was measured with WAMU’s hybrid signal (at the time, with upper and lower digital sidebands at −12 dBc and −14 dBc, respectively) and using the power of only the analog host, in dBm. If the antenna system was a dipole (having a gain of 2.15 dBi) the corresponding FM field strength would be approximately 19 dBµV.