T1 interfaces have found common use in STL applications
The field of broadcast engineering keeps getting more complex. And though I sometimes pine for the days before computer automation, one thing I never miss is an old analog radio STL. While I still make use of one or two of the classic analog radios here and there, always in the back of my mind I remember that parts just aren't available any more. I have to keep hoping that they never break. The little parts drawers are nearly empty. Fortunately most of them are back-ups anyway.
With the trend toward digital audio transport, and lately data transport, a whole new generation of radio STL systems has been developed. Their inherent advantages completely outweigh the problems that they happen to share with their ancestors. For example, with an older style analog radio, co-channel interference could easily be noticed (at best) as some low-level audio chatter and (at worst) not-so-low level audio and a beat note to go along with it. My experience is that digital radios can operate more easily in a co-channel fashion, as long the desired-to-undesired ratio is great enough for the local demodulator to ignore the “noise” signal from the undesired transmitter. (If you opt to rely on this, by the way, be certain that your desired signal doesn't go through fades making it weaker, and that your undesired signal isn't in a fade when you measure your desired to undesired ratio. Changes in the desired-to-undesired ratio while you aren't looking may lead to great frustration later.)
The QAM modulation scheme of many of the digital radio systems seems to make them particularly susceptible to de-sense. Other transmitters keyed on and off (like paging transmitters) can make some of the digital demodulators lose lock, even with the receiver being far removed in frequency from the offending transmitter. That isn't typically a problem with analog radios because they use FM (unless you are unlucky enough to have an intermod show up on your STL frequency). The solution in a case like this is generally a sharp-skirted filter to keep other RF out of the front-end of your receiver. Getting a better antenna and using horizontal polarity (if possible) are other ways to reject signals from many other communications systems transmitting from mountaintops.
The current generation of digital radio STLs sound great, but like many other systems in radio, they aren't necessarily plug and play.
The primary advantage of any digital system vs. an analog system that operates within a limited dynamic range is the consistent quality of the sound. The dynamic range of audio encoded in a digital system is essentially fixed and is defined by the word length.
The dynamic range of an old analog system is defined by two things: the noise floor (in the case of an analog radio system or analog tape) and the limit of headroom (100 percent modulation in the case of an analog radio, the saturation point of tape).
Wireline STL systems have come a long way in the last 20 years. I never had the opportunity to use copper loops for an STL — although it was common everywhere in the old days. (You were indeed lucky if your studio and transmitter site shared the same central office) Getting a 15kHz pair was, in many instances, a crap shoot; the performance depended on the skill of the telco techs that aligned the system. If you sent audio between central offices, then the audio had to be converted to digital (and truncated to 11 bits), sent over a T1, and then reconverted back to analog on the far end. If you could get 65dB of dynamic range you considered yourself lucky.
Trying to explain the frequency response vs. headroom trade-off to some telco techs was indeed something that could try your patience — and it used to be that if we got a good field tech, then we knew the lines would work. We always tried to get the same tech for the next time. (Now they're all retired.) Once again, advances in other communications technologies (like the cellular telephone) have had a positive benefit for broadcasters: with more users at typical transmitter sites, many telcos installed fiber, dramatically increasing the reliability of wireline STL systems.
The copper loop has entered the digital age with the Pulsecom PCAU and HD PCAU.
Perhaps more significant though was when we were allowed to generate the data ourselves for subsequent transport over telco T1s. After that happened, the floodgates opened. Intraplex began selling its channel banks directly to end users. It included the A/D, and at the far end, the D/A, that gve broadcasters more control over the sound of the radio station and many other handy features. Other manufacturers jumped at the same chance not long after.
We have actually come full circle with wireless connections now that Pulsecom and APT have partnered on the PCAU and HD PCAU. Both work with dedicated lines to provide digital links via copper.
The nature of data transport via digital data circuits (such as a T1) is that it is bi-directional — in fact full-duplex — and so signals can easily be brought back in the TSL direction. A common example would be the backhauling of air monitor or audio from an RPU receiver. Those are not functions that you are typically going to carry out with a radio system.
The ability to extend the station's LAN to a transmitter site has nearly become a necessity in the last couple of years. First, it was only because it was handy to have a computer at the transmitter site to retrieve e-mail or look at manuals online. Now there are many other reasons for it, not the least of which is HD Radio (although you only need the data going in one direction for that). Radio systems that can do that also are fairly easy to come by, but nearly all of them work in unlicensed ISM bands and as such, use low power, limiting their operational distance, and are ultimately subject to interference from other unlicensed users. That being said, I recently installed a 5.8GHz ISM band radio making use of a four-foot dish pointing right at downtown Seattle. Using a spectrum analyzer, I saw absolutely nothing else but the transmitter on the other end of the link.
Any digital STL acquired in this day and age should have the following capability, at minimum: the ability to send one AES data stream, embedded with left and right channels, sampled at 32kHz (44.1kHz if you want to be able to generate the HD MPS at the transmitter site) with 16-bit resolution; and full-duplex data connectivity for TCP/IP applications. Alternatively, if you intend to make use of the exgine architecture for an HD Radio application (and you don't need a LAN extension) then a simplex data connection will be adequate. The recommended minimum data rate is 300kb/s.