RF Field Notes
Both raise the uplink noise floor. Telling them apart early saves weeks, because the two problems live in completely different places.
Same symptom, different world
Elevated RTWP is a symptom, not a diagnosis. It says the uplink noise floor has risen. It does not say whether the energy is arriving from outside the site or being generated by the site itself.
Those are two entirely different jobs. External interference is a search problem that happens off site. Passive intermodulation is a hardware fault that happens on the antenna line. Teams lose weeks by starting the wrong one.
The tells that point to PIM
- It tracks with downlink traffic. PIM is generated by the site’s own transmit power mixing at a non-linear junction. More downlink load, more intermodulation product, higher noise floor. If the RTWP curve follows the traffic curve, look at the line.
- It is sector-specific and stays that way. PIM belongs to a path. It does not wander between sectors.
- Weather changes it. A joint that is marginal when dry and worse when wet is a classic failing weatherproofing or corroded connection.
- It appeared after site work. A connector remade without correct torque, a jumper crushed during install, or hardware disturbed during a swap.
- The affected band is predictable. The intermodulation products of the site’s own downlink carriers land in specific places. If the rise sits exactly where the arithmetic says it should, that is not a coincidence.
The tells that point to external interference
- It is independent of traffic. Noise floor elevated at 3am with almost no downlink load is not PIM.
- Neighbouring sites see it too. Particularly on the sectors facing a common area. PIM does not propagate to the site next door.
- It has a schedule. Weekday business hours, or evenings, or bursts — something being switched on and off by a person or a timer.
- The site tests clean. Sweeps pass, PIM tests pass, the radio has been swapped and nothing changed.
- The signature does not match the site’s own products. A narrow carrier sitting somewhere unrelated to your downlink arithmetic came from somewhere else.
The test that settles it
When the indicators are ambiguous, the definitive check is straightforward: reduce or disable the site’s own downlink transmission during a maintenance window and watch the uplink noise floor.
If the noise floor drops with the site’s transmit power, the site is generating it — PIM, or a hardware fault. If the noise floor stays where it is, the energy is arriving from outside, and the investigation moves off site.
That single test, run properly, is worth more than another week of speculation. It needs an outage window and coordination with operations, which is exactly why teams avoid it — and exactly why the problem stays open.
Where each path goes next
The expensive mistake
The expensive mistake is not picking the wrong path. It is picking a path and never checking back. A crew sent to climb and remake connectors on a site whose problem is a repeater three blocks away will remake the connectors, find nothing wrong, and report the site as clean — and the ticket will still be open, with a climb and a truck roll spent.
Spend the first hour on triage. It is the cheapest hour in the whole investigation.