HD Radio Subchannel Signal Strength

[Deleted member 76036]

I was just curious if it’s possible to limit the signal strength of a single HD subchannel, while keeping the primary and other subchannels at a higher power?

The reason that I’m asking is because Audacy moved Channel Q to 103.7 KVIL-HD2 here in DFW on Thursday. When it was on the HD3, there was a lot of dropouts where I live and across the metroplex. My neighborhood is less than 8 miles from the tower site. I could even have a clear view of the towers and I would still constantly experience dropouts. The dropouts happen on the radios in my home, as well. They aren’t as frequent, as I’m not moving, but they do happen.

These dropouts don’t seem to be an issue for other HD subchannels from stations of the same ERP and similar HAAT from the same tower farm. I was hoping the move from the HD3 to HD2 would eliminate this. The previous smooth jazz format on the HD2 didn’t have the dropouts, but now that It’s been replaced with Channel Q, Channel Q’s dropout problem has followed the format. It’s also weird, because the dropouts will sometimes cause the entire station to lose the HD lock. However, this only happens when I listen to Channel Q. If I listen to the HD1, the signal is solid and rarely ever drops out.

I was driving on I-20 in Dallas last night and had a clear view of the tower. Meanwhile, 103.7 HD2 was breaking up so badly that it was barely listenable.

Any ideas on why this may be happening?

 

[Kelly A]

I was just curious if it’s possible to limit the signal strength of a single HD subchannel, while keeping the primary and other subchannels at a higher power?


Any ideas on why this may be happening?

All HD1 and ancillary channels are packed into the same bitstream and sidebands. By your description, I'm guessing they're having some kind of malfunction. Maybe an importer or exporter problem.

 

[CentralFL]

Channel Q has the same problems on WOMX in Orlando. Including the HD dropping from the primary hd 1. It must be something the Audacy is doing wrong.

 

[Deleted member 76036]

All HD1 and ancillary channels are packed into the same bitstream and sidebands. By your description, I'm guessing they're having some kind of malfunction. Maybe an importer or exporter problem.

Thanks! I appreciate your response. I communicate with Channel Q’s Brand Manager from time-to-time, as he’s a great guy! I’ll pass this info on to him, in hopes he can pass it on to engineering. He’s passionate about making sure Channel Q runs properly with Artist/Title info, logos and album art and sounds good!

@CentralFL I’ll also include info about WOMX in Orlando.

 

[semoochie]

I have noticed that some HD3s and 4s have dropout problems that don't affect the main channel or HD2. There are a few ways of transmitting the HD signal and my guess is that one or more of those ways is somewhat incompatible with the additional channels.

 

[Kelly A]

I have noticed that some HD3s and 4s have dropout problems that don't affect the main channel or HD2. There are a few ways of transmitting the HD signal and my guess is that one or more of those ways is somewhat incompatible with the additional channels.

It may have to do with the way the bit allocations are configured. Earlier brands and models of HD-FM gear used to require the bitrate allocation to be set up by the end user. With one company it needed to be configured as a hex decimal formula. Of course, HD 2, 3, etc. are always lower rates than HD-1. Modern all in one HD exciters or exporters set the allocated bitrates automatically, depending on how many additional HD channels are activated. Nautel, for example, it's just clicking on a box for which channels you want active.
Although, I did get caught one time by not knowing that a new Nautel exporter defaults to "Extended HD", where the sidebands are something like 12% larger to accomodate additional bandwidth. I returned from the station and looked at the spectrum analyzer and couldn't figure out why the sidebands appeared to be so much wider than usual. Prior to that, I'd never heard ot 'extended HD'. Just another box checked on a different page in the GUI that I'd missed.

 

[Dr. Bob]

I may be able to provide some additional information. Many stations use the extended hybrid mode to transmit their HD3 signal known as the MP3 mode (no relation to the MP3 audio codec). The MP3 mode contains additional digital carriers closer in frequency to the analog carrier, both above and below it. The data rate for the primary hybrid mode carriers (usually HD1 & HD2) is 96 kbps and the data rate for the extended hybrid carriers (usually HD3) is 24 kbps.

The information in the primary digital carriers is duplicated above and below the analog carrier. This effectively means a receiver can lose part or all of the digital carriers above or below the analog and still receive the 96 kbps data transmitted the the primary partition (usually HD1 and HD2). The data in the extended partition is not duplicated above and below the analog carrier so the digital carriers both above and below the analog must be received without interference for the HD3 to work. Since the extended hybrid digital carriers are closer to the analog, they can also receive interference from the analog sidebands during analog modulation. All of this makes the HD3 more “fragile“ than the HD1 or HD2 when the HD3 is transmitted using the MP3 mode.

 

[pclover]

The information in the primary digital carriers is duplicated above and below the analog carrier. This effectively means a receiver can lose part or all of the digital carriers above or below the analog and still receive the 96 kbps data transmitted the the primary partition (usually HD1 and HD2).

This is a common misconception. It's not not duplicated. Both sidebands make up the entire channel.

Good receivers are able to decode from only a single sideband with a strong signal due to resilient FEC. The error rate is still very high but you have enough data to correct errors and get a usable stream.

 

[Dr. Bob]

The sidebands are duplicated through redundancy in the data, I.e. there is enough original data in one digital P1 sideband for the system to decode with the loss of the other P1 sideband. I was employed in the 90’s by one of the three companies that owned USADR. One of the primary design goals of the FM system was to decode with one of the digital sidebands missing since multi-path looks like a notch filter inserted into the transmission path.

 

[Deleted member 76036]

Hey there, I just wanted to say thank you! I shared this thread with Channel Q’s Brand Manager on 1/14. He ended up emailing me back and then sent another update on 1/31 to let me know that he was bringing in the big dogs to tackle the issue. Anyways, a couple days after that I noticed there were no longer any dropouts or the loss of an HD lock and it hasn’t happened since!

@CentralFL - I did mention to the brand manager that the same issues were happening in Orlando on WOMX. If this is still an issue, let me know and I can pass the info along.

Thanks again for everyone’s help in solving this longtime problem!

 

[CentralFL]

Hey there, I just wanted to say thank you! I shared this thread with Channel Q’s Brand Manager on 1/14. He ended up emailing me back and then sent another update on 1/31 to let me know that he was bringing in the big dogs to tackle the issue. Anyways, a couple days after that I noticed there were no longer any dropouts or the loss of an HD lock and it hasn’t happened since!

@CentralFL - I did mention to the brand manager that the same issues were happening in Orlando on WOMX. If this is still an issue, let me know and I can pass the info along.

Thanks again for everyone’s help in solving this longtime problem!

It seems to be fixed on WOMX as well! Good work contacting them.

 

[Deleted member 76036]

It seems to be fixed on WOMX as well! Good work contacting them.

Thanks! 🙂

 

[pclover]

The sidebands are duplicated through redundancy in the data, I.e. there is enough original data in one digital P1 sideband for the system to decode with the loss of the other P1 sideband. I was employed in the 90’s by one of the three companies that owned USADR. One of the primary design goals of the FM system was to decode with one of the digital sidebands missing since multi-path looks like a notch filter inserted into the transmission path.

I guess that's one way to look at it. Perhaps it's easier way to explain it like this instead of the complicated information theory.

However, I still look at it that the redundancy is in the from of FEC(Forward Error Correction) each packet that is sent out extra data is sent along with it to correct errors.

If you take a commercial decoder or something with the rtl-sdr and a take out a sideband you'll notice the MER(Modulation Error Rate) and BER(Bit Error Rate) is very high but it still works due to extra data from FEC. There is enough redundant data from the FEC to put it back together. Sometimes you'll still get a CRC error.

HD radio was designed with multipath in mind. You can get some pretty nasty multipath and it still works fine due to DSP magic. Even if you get selective frequency fading that you can call a notch filter you can still recover data from it.

If you look at the actual NRSC-5 spec documentation it explains it all in detail.