AC mains monitor

Something like this perhaps?

Hamonics maybe. Kill the fundamental with a notch or highpass filter.

I don't know. What good will "knowing" this do me? Why does the UPS get so upset? Is there some risk -- other than not being able to deliver sufficient power to the load? If I had a 'scope on the line, could I *tell* what was problematic??

Is it complaining about an unimportant problem?

I don't know. Ask the UPS :)

My crystal ball is in for a service so I don't know. But if the UPS is complaining about distortion then a harmonic analysis of the line when it complains might be revealing.

Probably. If you're a UPS then you'll want to make sure the customer knows that you're doing your job.

Your unspecified model UPS probably has adjustable limits for how much distortion (i.e. harmonic) in can tolerate on the AC input. For example:

"Adjusting Input Sensitivity on APC Smart-UPS SMT & SMX Series | Schneider Electric"

"Understanding the full implications of harmonic distortion"

I run into distortion problems when utility power is switched to generator backup power during power outages. The inverter generators, which produce low distortion sine waves work the best. The modified staircase approximations of a sine wave, produced by cheap commodity generators, are horrible.

Another source of distorted power line waveforms are non-linear loads.

"Understanding Total Harmonic Distortion" <https://powerquality.blog/2021/02/11/understanding-total-harmonic-distortion/><https://powerquality.blog/wp-content/uploads/2021/02/understanding-total-harmonic-distortion.pdf>

"Analysis of Total Harmonic Distortion on the U.S. Electric Grid"

There should be something here. I'm not familiar with todays power line test equipment.

"Power quality meter THD"

"Clamp meter THD"

Incidentally, IEEE 519-2022 sets "voltage distortion limits" at 8% for power lines under 1000v: "Power quality - IEEE 519-2022"

Depends on the UPS's topology. "Line interactive" tend to let the mains through UNLESS there is a disturbance (low line, high line, outage, "distorted input", etc.)

An "online" UPS rectifies the mains and uses that to charge the battery as well as drive the inverter. ALL THE TIME. There is no "switching" involved. E.g., this is how POTS worked.

An "offline" UPS lets the mains flow to the output UNLESS there is an outage -- in which case the inverter powers the load. This necessitates a switch to select between power sources (mains or inverter out).

A line interactive UPS usually dynamically controls an (auto) transformer to adjust for variations in sensed mains voltage. When things go beyond its ability to compensate, the inverter is switched in to carry the load. (switch == relay click) But, most of the time, the load sees the mains directly, even thought the inverter is ready and waiting. There is less control over the line but less effort to do so.

Online is the gold standard. But, it also does the most work as the inverter is ALWAYS powering the load -- so, throws off lots of heat, etc.

Thes are in bogunits -- low, medium, high. No real way to look at hard data and determine how that should affect your choice of setting.

Given that the UPS *rarely* engages the inverter -- and, that the lights in the room were "shimmering" -- I imagine there was *something* going on with the mains.

We have some switching transients that happen, daily, at reasonably predictable times. Not much you can do except wait for the click on, click off, as the transient passes. I.e., a UPS that can carry the load for "many seconds" is often all that is needed.

Knowing your line quality is X doesn't really tell you much (except that it is not *Y*!)

E.g., many devices rectify the line so the only significant issue is an extreme voltage spike (that exceeds the rating of some front-end component) *or* a significant dropout that the bulk filter can't carry (at the present load).

Maybe use a resistor attenuator and record the mains with a PC sound card? There exists lots of software that then can replay and show the waveform, and do all sorts of math on it.

Here one part of mains lead is ground, other live, PCs are all grounded, use the live one :-)

The one that upsets mine is when one of the three hot phases goes down and instead of 240vac we get 87% of that. Everything being on switched mode PSUs these days you don't notice until you try to boil a kettle!

In the good old days of CRTs the height and width would shrink (if it worked at all). Incandescent lamps would be redder and dimmer too.

Easiest way is put a scope on it to see the actual mains waveform, or perhaps on an old ac transformer output at a safer voltage.

Here, the worry is an open neutral as few homes have truly balanced loads (especially when users can switch individual loads on and off)

But that still doesn't tell me what to look for or if I should care. If the UPS gave better clues as to *why* it was "intervening on my behalf" (instead of just complaining that the line was abnormal), then I could decide if the condition was likely a transient or might persist. In the latter case, I could elect to just power off the powered device until things settled down a bit.

E.g., when I've lived places with overhead services, storms would often cause momentary disruptions and fluctuations in power. Without a UPS, it was wisest to just shut everything down and wait for the storm to pass -- at which time, power would restabilize.

Here, services are below grade so disturbances tend to be more network-wide. This time of year, likely related to excessive load on the grid. Much harder to predict if that will come back under control or fall apart, entirely.