The Day I Stopped Trusting

For thirty years, my field strength meter was the most important thing in the truck. Before the laptop. Before the smartphone. Before any of the platform tools we have now. The meter told me whether the signal was good. If it was good, I went home. If it wasn’t, the meter usually told me why.

I had a specific meter for fifteen of those thirty years. Replaced the original after I dropped it down a manhole in 1998, different story, involves a thunderstorm, my own poor judgment, and a coworker who was laughing too hard to be useful. The replacement had its quirks. I learned them. I knew the calibration drift. I knew exactly how to hold it for a stable reading on a windy day. Held it the way some men hold a putter, which is to say, with feeling. That kind of relationship with a tool isn’t sentimental. It’s earned. You learn what a tool will and won’t tell you, and you trust it accordingly.

Got a ticket on a customer in a row of duplexes. Slow internet, intermittent. Standard call. The customer had labeled every cable in his house with a Sharpie. I want to go on record as saying I respected that. He wanted me to know that the cable behind the entertainment center was specifically his, the one in the master bedroom was specifically his, and the one in the guest room was where his mother-in-law watched her shows. We spent the first six minutes of the visit on a verbal cable map I did not need.

I rolled up, hung the meter, took readings at the ground block, the splitter, the wall plate, and the modem connection. Every reading was within spec. Forward levels, clean. Return levels, clean. I tightened the connectors, swapped the modem out anyway because of habit, and left.

Customer called back the next day. Same complaint. I went out a second time with the same result. Levels were perfect. By every measurement my meter could take, this customer should have had the best service on the block. He told me again about his mother-in-law’s shows. I assured him I was working on it. I was, in the broad sense of the word. Just not effectively.

A younger tech showed me PNM data on the modem. I will note for the record that he was not smug about it, which I appreciated, and that he could have been, which I would have deserved. FEC was at 78%. MER on certain OFDM sub-carriers was eight to ten dB below where it should have been. There was a clear notch in the sub-carrier MER plot, sitting in a frequency range my meter doesn’t independently measure. Off-air ingress was landing exactly under the OFDM, contaminating the sub-carriers in that specific band.

The meter didn’t see it because the meter wasn’t designed to. The meter measures level. The level was fine. The problem was about quality, error correction load, sub-carrier modulation error, and that’s data the meter doesn’t produce.

I went out with a DRV3 the third visit, drove the plant from the customer’s drop, and found the source in eleven minutes. Bad shield on a span clamp three poles up. The customer asked if the problem was finally fixed. I said yes. He asked if his mother-in-law could go back to watching her shows. I said yes. We did not discuss what specifically had been wrong with the previous two visits. I considered that a kindness on his part.

Here’s what I mean: what the meter measures is still important. You still want signal levels in spec. If the level is wrong, the customer has a problem. But “level in spec” is now necessary and not sufficient. The customer can have perfect levels and degraded service. The thing that matters most about the customer’s signal is no longer something the meter can independently measure.

So the meter has gone from primary diagnostic tool to validation tool. I use it at the end of a repair to confirm the levels are correct. Before that, I use the dashboard. The work changed. The tool’s job changed with it. The meter has been demoted, basically. We don’t talk about it.

A tool you’ve trusted for decades stopping being able to answer the question on the table is a strange experience. There’s a brief temptation to deny that it happened, to insist the level should tell you everything, because it always did before. That instinct gets you nowhere on a DOCSIS 3.1 plant. The carrier modulation changed. The diagnostic implication changed. The tool can’t catch up because it doesn’t measure the new metric.

The hardest thing about the job in my last five years has been letting go of the version of myself that read the meter first. The meter is still in the truck. It still works. It’s just not where I start anymore. It rides along now. Like a relative you bring on errands because you can’t quite imagine the truck without it.