Another technical problem with HD ---

[rbrucecarter5]

There is a little known station spacing rule that prevents close spacing on stations 10.6 and 10.8 MHz apart, because the IF images from the lower station would cause problems with reception on the upper station. 10.4 and 11 MHz apart are OK - right?! WRONG!!!!

I've had problems with HD dropouts in Houston on KRBE 104.1 - and fairly close to the towers. I now know one reason. I had a dropout the other day on KRBE HD-2, and it was when I was driving by a car that was blasting the local 93.7. With the stations spaced 10.4 MHz apart, there should be no interference. But - both stations are HD, so when you factor in the first adjacent interference potential with HD sidebands, 10.4 MHz (and by implication 11 MHz) spaced stations are no longer usable without jamming from nearby radios with local oscillators.

Just another "gotcha" caused by the engineering "geniuses" at iBiquity.

 

[tomservo]

That's wild. I never considered it. But then, I'm not an engineer so it's forgiveable. The iBiquity people, on the other hand…

At some point when I'm not three minutes from falling asleep I'll have to see if any of my local stations are spaced that way. I often lose the HD close to the towers, but I'm surrounded by so much RF here that I just figured it was overload from the analog FM and digital TV stations all co-located, pumping out millions of watts in my general direction.

 

[tomservo]

So much for going to bed.

I put my Insignia HD radio on top of another tabletop portable and started tuning around. Sure enough, if I tune just right, I can kill the HD, but only on one station with a notoriously weak HD signal. Here's what I found: if I tune so the analog — 101.5 MHz — is wiped out (which it isn't, it just gets fuzzy) the HD remains. If I tune to the lower sideband, so that the birdie is on 101.3 MHz, nothing happens. If I tune to the upper sideband so the birdie is on 101.7 MHz, the HD loses lock.

This required incredibly precise tuning to work, with the portable right up against the HD unit and the portable tuned to exactly 90.95 MHz. even 0.05 MHz either side and the HD came right back. I find it hard to believe any radio could do this from car to car, but that will require some more testing later today.

When I tried with the lowest powered HD signal in town, I couldn't get it to break down at all, which was strange. When I tried it with the most robust signals, I couldn't even detect co-channel interference on the analog OR when tuned to the digital side bands.

Even if I could, none of the IF's correspond to any other local channel as far as I can tell, except for one non-HD instance, where the class C country station on 94.9 would wipe out the other radio tuned to near local class A on 105.7. And even then, the class A won out after less than 10 inches from the radio.

Unless car radios allow for much greater IF emissions, I just don't see this panning out as a legit complaint.

 

[Kmagrill]

Local FM oscillators operate at 10.7mHz which is why there is a spacing requirement for 10.6 and 10.8mHz channels. Since 10.6 and 10.8 are already 100kHz removed from the actual IF of 10.7, there should be no interference at 10.4 or 11mHz which are 300kHz away from the center frequency.

The strength of the interfering IF is sort of proportional to the strength of the local IF channel station, so interference should only occur when you're very close to the IF transmitter. The number of affected stations, either HD or otherwise is, therefore, pretty small. Also, the shielding of the radio's local oscillator will have an effect. For example, a cheap clock radio is likely to radiate worse than a typical car radio, though most cars do radiate enough to be detected. About 15 years ago, Arbitron, or one of the ratings companies, actually experimented with audience measurements made by listening to the local oscillators of passing cars to determine which stations were being heard in real time. However, now that software defined radios are primed to become the next wave in tuners, the local oscillator issue may well fade away, from a minor concern to none at all, over time.

 

[rbrucecarter5]

Local FM oscillators operate at 10.7mHz which is why there is a spacing requirement for 10.6 and 10.8mHz channels. Since 10.6 and 10.8 are already 100kHz removed from the actual IF of 10.7, there should be no interference at 10.4 or 11mHz which are 300kHz away from the center frequency.

The strength of the interfering IF is sort of proportional to the strength of the local IF channel station, so interference should only occur when you're very close to the IF transmitter. The number of affected stations, either HD or otherwise is, therefore, pretty small. Also, the shielding of the radio's local oscillator will have an effect. For example, a cheap clock radio is likely to radiate worse than a typical car radio, though most cars do radiate enough to be detected. About 15 years ago, Arbitron, or one of the ratings companies, actually experimented with audience measurements made by listening to the local oscillators of passing cars to determine which stations were being heard in real time. However, now that software defined radios are primed to become the next wave in tuners, the local oscillator issue may well fade away, from a minor concern to none at all, over time.

But - when both stations are broadcasting HD, as is often the case, the channels are considerably wider. The math doesn't show overlap. but IF filters are not brick walls. So it is conceivable a strong HD sideband could interfere with the HD sideband of another station.

As for software defined radios - i just don't seen current semiconductor technology being fast enough to directly sample 100 MHz. The clock frequency for the DSP would have to be much higher - perhaps two to three orders of magnitude. If that can be made to work, it won't be cheap and it will drink power. IF PC's are any indication, clock frequencies in the low GHz range are plausible - but they stalled there years ago and the computers were speeded up by having multiple cores. Perhaps that technique would work for software defined radios, but a 29 cent radio on a chip IC and a cheap microcontroller to generate tuning voltage and digital display are hard to beat in consumer gear. Unless you can get something down to less than $5 bill of materials, I don't think the Chinese manufacturing houses that make most of our gear today will touch it.

 

[tomservo]

I was under the impression that both my car's OEM radio (2005 m/y) and all my cell phones had software defined FM radios already. I know in my ancient mp3 player from ~2003 it is also supposed to have a software defined radio, as third party software can be made to access different channel bandplans for different parts of the world.

 

[Kmagrill]

Software defined radios are rapidly gaining ground in the consumer market because they are cheap. Decoding 100mHz is trivial using today's high speed DSP. Recently, I've seen complete software defined tuners that include the software to decode radio or TV for under $20 retail. Ham radio operators have been able to buy them that are tunable from "DC to light" for at least 10 years. I have one on a USB stick in my office that cost $19.95 with free shipping. Many of the HD radios use SWD technology to make very sensitive and selective tuners cheap. I believe I read somewhere that Ibiquity will supply sample HD DSP code to licensed manufacturers for free. The days of traditional rf reception seem to be drawing to a close. Of course, there will be many millions of existing traditional radios out there for the next 50 years.

 

[K6JHU]

"I have one on a USB stick in my office that cost $19.95 with free shipping."

What USB stick are you using?

 

[rbrucecarter5]

Software defined radios are rapidly gaining ground in the consumer market because they are cheap. Decoding 100mHz is trivial using today's high speed DSP. Recently, I've seen complete software defined tuners that include the software to decode radio or TV for under $20 retail. Ham radio operators have been able to buy them that are tunable from "DC to light" for at least 10 years. I have one on a USB stick in my office that cost $19.95 with free shipping. Many of the HD radios use SWD technology to make very sensitive and selective tuners cheap. I believe I read somewhere that Ibiquity will supply sample HD DSP code to licensed manufacturers for free. The days of traditional rf reception seem to be drawing to a close. Of course, there will be many millions of existing traditional radios out there for the next 50 years.

If it is that Silicon Labs USB stick, I have one and inside it is good old superhet, same technology that is decades old, just completely in silicon. Hardly a software defined radio. Direct sampling of 100 MHz, given the Nyquist frequency involved, is just not possible. Cell phones running at 2.4 GHz use triple conversion to get the IF down to where it can be digitally sampled. Digital sampling of the IF is very different than direct sampling of the RF stream. Superhet will reign supreme for many years to come.

 

[JohnnyElectron]

Tell me more about this $19.95 Software Defined Radio USB stick?

 

[rbrucecarter5]

Tell me more about this $19.95 Software Defined Radio USB stick?

http://www.silabs.com/products/mcu/Pages/USBFMRadioRD.aspx

 

[JohnnyElectron]

USBFMRADIO : All gone :-\

--------------------------------------------------------------------------------
Note: This reference design is no longer supported, and is not available for purchase.
--------------------------------------------------------------------------------

 

[spunker88]

Tell me more about this $19.95 Software Defined Radio USB stick?

http://www.silabs.com/products/mcu/Pages/USBFMRadioRD.aspx

Thats actually just a regular radio IC (Si4701) with DSP controlled by a computer. There is a thread over at WTFDA about the Ads RDX-155 which uses the same IC and microcontroller. Its a decent tuner, but not a true SDR where bandwidth and modulation type would be configurable by the user and you'd be able to tune outside the FM band. A true and affordable SDR tuner would be one of those RTL2832 based TV tuner cards on ebay, etc.