Archives

April 2001 Issue
Broadband: Two-Way Success Secrets Revealed
By

Pssst!

I’ve got something to tell you. It’s not widely known, so we need to be discreet about this. Wanna know the secret to making your reverse path work well?

Gotcha. There is no secret weapon to successful two-way operation. For the most part, it’s nothing more than back-to-basics stuff you’ve known for years, such as proper network alignment, signal leakage management, compliance with relevant technical specifications and effective maintenance and troubleshooting practices.

I’ve seen too many attempts to "plug-n-play" when it comes to two-way operation. You know—plug in the reverse amp modules, equalizers and attenuators, and assume the job is done. In some cases, this task is accompanied by turning the reverse amp gain controls to maximum. That’s not the way to make two-way work.

Network alignment

Assuming your network was designed with unity gain in mind (see "Seek Balance in All Things: A Look at Unity Gain in the Upstream Coax Plant," June 2000 Communications Technology), and you’ve plugged in the reverse amp modules, you should start with downstream and upstream network alignment.

You must properly align the downstream and upstream—including fiber links—if you want the reverse path to work well. It’s not possible to characterize the extent of upstream ingress, common path distortion (CPD) and other potential impairments without first aligning the network. From a big picture perspective, you can’t tell how good or bad the reverse is if it’s not correctly aligned. For instance, if overall levels are too low, a spectrum analyzer display in the headend might indicate very low levels of ingress or other problems. Conversely, if overall levels are too high, the reverse spectrum might look like it’s absolutely unusable. Of course, if the reverse is running wide open, it probably will be unusable.

Several years ago, I assisted a cable operator with its corporate headquarters engineering lab. The lab included a simulated cable network, complete with a two-way fiber link, an eight-amplifier cascade and a section of tapped feeder. I was teaching the operator how to align the network for two-way operation, and we started with downstream alignment. As we progressed through the amp cascade, we noticed what appeared to be common path distortion in the reverse spectrum. The problem eventually was traced to what we all believed was a defective diplex filter in one of the amplifiers because it appeared that downstream signals from that amp were overdriving the upstream module.

Well, sort of.

The real problem was improper alignment of the downstream amplifier. We discovered that the multi-output amplifier had individual internal level adjustments (plug-in pads) for each output port. The amp’s main output had been adjusted correctly, but the other two were cranked to the max. The downstream was indeed overdriving the reverse amp. When all of the amp’s downstream outputs were correctly adjusted, the upstream worked just fine.

I can’t overemphasize the importance of properly aligning the network. Furthermore, I’ll preach until the cows come home that alignment must be done using broadband sweep equipment. Some argue that modern hybrid fiber/coax (HFC) networks with short amplifier cascades after the node don’t need to be swept. I’ll never buy that argument. While it’s true that amplifier frequency response signature built-up is less of a problem in short cascades, sweep alignment still is necessary to see what’s going on across the entire spectrum.

Broadband sweep will identify frequency response roll-off problems that aren’t evident with simple balance alignment techniques. As well, sweeping will identify defective or missing end-of-line terminations, water damage, response suckouts, over- or under-equalization and other problems that are difficult or impossible to see without sweeping. One word of caution: Make sure you program sweep guard bands around the downstream and upstream digitally modulated carriers. I can’t tell you how many times I’ve seen degraded bit error rate (BER) performance that was caused by the sweep transmitter sweeping right through the digitally modulated carriers.

For more on alignment, see "Mystified by Return Path Alignment? Get Your Upstream Fiber Links Aligned" in the March 2000 issue of CT.

OK, now that the network has been properly aligned, what’s next?

Managing ingress and other gremlins

Look at the upstream spectrum in the headend, and see just how good or bad the ingress is.

If the spectrum is relatively clean, and your network meets or exceeds the minimum recommended downstream and upstream parameters outlined in the Data Over Cable Service Interface Specification (DOCSIS) RF Interface Specification (this is on-line at www.cablemodem.com/SP_RFI_I05_991105.pdf), you deserve a tip of the ol’ hat.

If your reverse spectrum looks like most, you’ll need to tighten up your downstream leakage.

"But the plant meets the Federal Communications Commission’s 20 microvolt per meter (µV/m) leakage limit," you argue, "and our most recent cumulative leakage index was in the low 60s."

That’s probably not good enough.

Most operators who have upstream ingress under control keep downstream leakage well below the FCC’s 20 µV/m limit-often as low as 5 µV/m or less. These same operators have annual flyovers in the 99th percentile, or cumulative leakage index numbers far below 64.

Jim Kuhns, director of field support for Terayon Communication Systems and SCTE chairman of the board and Region 7 director, is a frequent return-path speaker at industry seminars. He told me that he is constantly approached by seminar attendees looking for the upstream ingress silver bullet. "We came to learn about two-way, and half of your presentation was on downstream leakage," they complain.

If your system has leakage, it has ingress. Where cable signals leak out, over-the-air signals will leak in. You’ve got to have a tight plant if you want reliable two-way. "It’s not rocket science," says Kuhns. Getting ingress under control is largely "low-tech stuff," he adds.

The real issue is whether or not you’re willing to devote the necessary resources—time and manpower—to tighten up the plant. Fix your leakage, and the worst of your ingress will go away.

Once you get over the alignment and ingress hurdles, you’ll be able to tackle things like CPD. Common path distortion can be masked by excessive ingress or noise, and may be made worse by improper downstream alignment. If you don’t have CPD or it’s low enough to not be a problem, consider yourself fortunate. Otherwise, get a couple bottles of aspirin—common path distortion can be a real headache.

But it doesn’t stop here. Keeping the network properly aligned, leakage and ingress down to a dull roar and other reverse path gremlins at bay all are on-going battles. "It just takes a lot of TLC," Kuhns says.

Ron Hranac is a consulting systems engineer for Cisco Systems, and senior technology editor for CT. You may reach him at .

 Back to April 2001 Issue

Access Intelligence's CABLE GROUP
Communications Technology | CableFAX Daily | CableFAX's CableWORLD | CT's Pipeline
CableFAX Magazine | CableFAX databriefs | Broadband Leaders Retreat | CableFAX Leaders Retreat