PC power drops when UPS switches from grid

[CntrlAltDel]

My wife and I have different make PC's. When we have brownouts hers shuts down/resets but mine doesn't, so it depends on your power supply.

Depends on power supply to an extent. They have a hold-time usually in the 10-25ms range but that is almost irrelevant to a line interactive UPS that has a general switching time of 2-6ms. Mine has a hold time of 22ms. This is more related to the heavy power draw of 600w that the batteries that are 2x12v9ah in series are not able to hold with severe voltage sag that causes the low voltage protection of the UPS to kick in.
This boils down to over optimistic UPS specs by all manufacturers. Only solution would be to get an online UPS (big no-no because I literally only need a UPS to hold the load for 1-2 seconds for lithium battery bank to kick in) or the better alternative, lifepo4 batteries that can handle heavy loads MUUUCH better than SLA.

 

[Gnome]

View attachment 1585202
Devices in the household with a low power draw or good holdup time will survive the first loss of power (eskom) and continue to work through the 2nd loss of power (vdroop)

I'm nitpicking a bit here, your information is generally good, but not all UPS function this way.

APC line interactive UPS have synchronization. They are always following the line and can switch in ~4ms and the switch over would not look the way you illustrate here. It would have a very short interruption followed by a wave that continues on. Before switching back to grid, those UPS would then slowly adjust the sine wave to get it back in sync with the incoming AC line voltage.

Any UPS that actually offers less than 16.7ms (60hz) / 20ms (50hz) change over time would need voltage following. Without voltage following you could end up switching a supply at peak back into another peak which would basically be an enormous over voltage condition.

Most cheap UPS therefore take more than a single AC wave to switch to avoid that complication.
Any reputable UPS brand would attempt to switch at the zero crossing when switching back to grid.
Doing otherwise is beyond stupid for a lot of reasons (these cheap inverters just YOLO it so no doubt it could look the way you illustrate)

This problem is also compounded on a "square wave"/"modified sine wave" inverter because it is a square wave (very commonly actually) or best case stepped approximation (much more rare on "higher quality" UPS, pretty rare these days). Stepped approximation is also only typically applied on most of those UPS at low duty cycles (meaning high duty cycles just revert back to square wave).

Overall you provide a lot of good info about hold-up time, but there are a quite large number of different behaviours in terms of UPS and saying the wave form above is definitive would be fairly far off the mark because there are many implementations out there.

There are two guys (different channels) on YouTube that review these units and inverters all the time that you can check out to see the various approaches and so on.

Only with an actual sine-wave inverter do things get more predicatable.

Alternatively, investing in an 'online' UPS/inverter would shorten the time to operation, allowing a 'to-spec' PSU to achieve its holdup time.

To be clear, a online UPS has no switch-over whatsoever. There is no disruption whatsoever, literally nothing. If it doesn't work like that, it is not an online UPS (more commonly referred to as double conversion)

 

[Mirai]

My wife and I have different make PC's. When we have brownouts hers shuts down/resets but mine doesn't, so it depends on your power supply.

You should swap PCs and let your wife have the more stable computer.