by Peter W. Wed Aug 02, 2017 8:03 am
https://www.cdw.com/shop/products/Tripp-Lite-1000VA-800W-Line-Interactive-UPS-Sine-Wave-120V-Rackmount-2U-TAA/1923328.aspx?cm_cat=GoogleBase&cm_ite=1923328&cm_pla=NA-NA-TRI_BA&cm_ven=acquirgy&ef_id=WUgmggAAAKtl5okB:20170802113356:s&gclid=CjwKCAjw8IXMBRB8EiwAg9fgMBoz0yS-m4ApcExYGxNmXe24JTqUyNy1SkFLHpnitwuGwWn68dVXXBoC1skQAvD_BwE&s_kwcid=AL!4223!3!47988697099!!!g!64450928333! Does not seem overly complicated nor particularly expensive to me for the basics - if one's power is as bad as described.
National Standards:
https://www.pge.com/includes/docs/pdfs/mybusiness/customerservice/energystatus/powerquality/voltage_tolerance.pdf A little bit dense to read, but the potential for variation is considerable.
Those of you at some distance from the source, in areas where there has been considerable growth, you are at particular risk, as suppliers have raised the voltage in the distribution lines so that existing conductors can carry additional capacity. Meaning that you may be running at the limit of the local step-down transformers.
I will state, again, for the record, that variable autotransformers, while better than nothing, are not designed to create a steady-state output, but to drop (or raise) input voltage at some proportion to the setting of the transformer. They also can be a source of noise, and if run constantly at anything approaching their load rating will get hot. Note that the largest risk of an VAT is the wiper - a small graphite button (if quality devices, otherwise a simple copper/bronze leaf-spring) that carries the entire load. Hence the need for significant over-sizing if used as a constant control.
We are blessed with "good" power - inasmuch as we are in a very old neighborhood with almost no additional development since 1900 - so the increase in load has been relative to additional consumers - central AC, appliances and so forth, not 300 new houses on the same distribution line. But if I were in an area with such wild fluctuations as described here, I would go with a decent line-conditioner. Were I to be into belts, suspenders and Velcro, I would feed such a conditioner through a heavy-duty VAT at say.... 118 V or so as a midpoint. And let the LC handle the rest. We get about 118 V as our steady-state at the wallplate here. Even mid-summer when the AC load is at its highest.
You pays you money, you makes you choice. What is appropriate for my location is unlikely to be entirely appropriate to yours.
On Kill-A-Watt meters - they are about as precise as a diesel engine, and that is no big deal when one considers their basic purpose, which is a snapshot of actual consumption at any given location at any given time *relative* to a previous future or time. Does anyone here remember the term "speedometer error"? This was the amount of slop expected in the typical automotive speedometer that the "Cops" would typically permit a drive before pulling one over for speeding. In Pennsylvania, for one example, the 'courtesy' was about 15% of actual speed. A study was done in about 1968/9 or so, and it was determined that at a cost of about $0.02 per car, this error could be eliminated, or reduced to a variation based on tire wear (really). Two things happened: The automobile industry howled that this cost would be excessive across total production (about 9,000,000 units in the US alone - $180,000. Really?), and the AAA howled that this would cause all sorts of unjust arrests for speeding. Never happened. Look at the Kill-A-Watt meter as a similar device. It is only as accurate as it must be. A Fluke Meter is designed to be accurate - and costs accordingly.
Last edited by Peter W. on Wed Aug 02, 2017 10:23 am; edited 1 time in total (Reason for editing : Punctuation.)
. What I really did was to add a little bucking transformer to drop the output of a power strip by about five volts. Nice sine wave in, slightly smaller but equally nice sine wave out... just for my ca. 1960 tube gear. The appliances hum along with the full voltage all day and night. 
