SoylentNews is people
SoylentNews
First, I debated whether to put this on stack exchange or here, but it seems like it is a tech question that suits this site fine.
Background
I have a room with a 115 V, 6000 BTU window AC unit plugged into one outlet. Then a bunch of electronics (~800 W measured) plugged into a 1500 VA, 900 W UPS plugged in to a second outlet across the room. Finally, I have two 50 W strands of Christmas lights in series (100 W total) I tried to plug into various outlets around the room.
Problem
The first problem is that whenever the room gets too hot, the compressor for the AC unit turns on and the Christmas lights will all flicker. This is not just an annoyance, because the first strand of lights I had in the room actually got burned out one by one, starting at the light closest to the wall outlet.
So I got another strand and was surprised to see the flickering happens even if they are plugged into the UPS (which does have an internal automatic voltage regulator). This made me concerned for the electronics plugged into the UPS, which includes a PC and monitors. However, I do not notice any flicker on the monitor when the compressor turns on. On the other hand, I have been getting some strange pc crashes lately (which would make some sense because only recently did it cool enough for the AC to not be running constantly) that may be related. This could also be due to installing a second gpu recently, etc though.
Questions
I have two main questions:
1) What is the best way to stop the flickering?
2) If the lights are flickering even when plugged into the UPS, should I also be concerned about the other electronics that are obviously also experiencing a momentary power reduction?
Some secondary questions:
3) Does it make sense to put another AVR between the UPS and the wall, eg something like this?
4) Is there something I can put between the AC unit and the wall to help?
5) This is a rental so I would prefer not to do any maintenance on the AC unit, but is this an issue you would report to the landlord?
Any ideas?
(Score: 4, Interesting) by RS3 on Sunday November 10 2019, @09:58PM (25 children)
Are you able to measure the voltage? Like with A/C on and off? If it drops more than a few volts, maybe the A/C has a problem and draws too much current?
Or maybe there's something loose- generally a wire terminal screw.
Among many things I occasionally do wiring both new and retrofit. I often find loose screws. Outlets are wired in a jumper or "daisy-chain" method, so the electricity getting to the outlet you're referring to may go through many outlets, and there may be a loose screw somewhere in the chain.
Also, there are "push-ins" in the back of outlets- internal spring tabs that supposedly make contact with the wires. Someone might have used them. They're often very unreliable, and I pretty much never use them.
Also check at the breaker (fuse?) panel, if you know what you're doing- both the "hot" and the neutral (white return, assuming you're in USA or somewhere with similar wiring.
(Score: 0) by Anonymous Coward on Sunday November 10 2019, @10:11PM (16 children)
It depends on how long it has been since the AC was last on. But yes, I have a killawat that reports voltage and the UPS also reports voltage to me. What exact measurements do you think would help?
(Score: 1) by NPC-131072 on Sunday November 10 2019, @11:10PM (15 children)
Watts to confirm problems are caused when the AC switches on.
Yes, confirm it's a problem and then get the AC unit and board checked over by an electrician via your landlord.
(Score: 0) by Anonymous Coward on Sunday November 10 2019, @11:15PM (14 children)
You mean put the KIllawatt between the AC and the outlet? I've done this but not noticed anything crazy. Possibly because it reports only the average of the last x seconds (where I am not sure how long x refers to). The flickering lasts for ~1/10th of a second.
(Score: 1) by NPC-131072 on Monday November 11 2019, @12:17AM (11 children)
If it's happening when the AC is switching, there's an abnormal power draw. You could check with a decent (safe) multi-meter if you have one?
Fixed something like this last month on an air compressor, contactor was sticking and needed cleaning and re-greasing but the compressor was also on an old wire fuse at the board and the bakelite holder fell apart when I pulled it. Shot foam into cable holes at the board and used insulation tape around the fuse holder. My fren is now saving for a rewire and new board with RCD's because the arcing at the board had scorched the insulation on one phase of 2 adjacent circuits. Also found a single phase run in metal conduit where some psychomonkey left a live protruding from a screw terminal, replaced that mess with wago connectors housed in a junction box.
Half the people here could help you fix but on a rental with potential fire / electrocution risk, just get your landlord involved.
(Score: 4, Informative) by Runaway1956 on Monday November 11 2019, @01:34AM (10 children)
We don't grease contactors. Grease attracts dust and dirt, making the contactor stick worse. Be aware that the contactor points generate the dust that you find in there. Most greases are also flammable, which will start a fire. Unless you meant dielectric grease? Contactors should be bare-assed copper-to-copper contact. Even if I were to read a maintenance manual on an exotic piece of equipment that specified lubricating the contacts, I would be reluctant to do so.
If contactors are worn to the point that they need to be maintenanced, you should just replace them. Cleaning and filing them means that you will return in the not-distant future to replace them anyway.
(Score: 2) by RS3 on Monday November 11 2019, @03:14AM (5 children)
Oh gosh, agreed- NEVER grease contacts! That said, I bet someone has some kind of compound for contacts, but normal contact arcing would turn anything into quarks, muons, and neutrinos.
You might argue in favor of using an anti-corrosion paste on the wires in the terminals / lugs, but never on the contacts.
You shouldn't need lube on the moving armature, but you could use a light silicone grease or oil, but again, I can't imagine where you'd need it- maybe a moisture compromised location? It might get on the contacts eventually, so either way you'd be replacing the thing.
(Score: 2, Informative) by NPC-131072 on Monday November 11 2019, @03:28AM (4 children)
Petroleum jelly melts off well before it catches fire. Read any manufacturer advice on isolator switches and it'll state PJ is preventative against arcing since it is non-conductive and gets smudged out of the way at the mechanical point of direct electrical contact.
(Score: 0) by Anonymous Coward on Monday November 11 2019, @03:50AM
https://www.youtube.com/watch?v=BF4WDVVDTp0 [youtube.com]
(Score: 2) by RS3 on Monday November 11 2019, @01:41PM
Thanks, that's very interesting. I tend to avoid risk, so I'd rather not put anything on the contacts which might leave a residue. But if it's helpful, I'll try it. But I'm curious- if it melts off, how is it helpful after the first contact?
(Score: 1) by Sally_G on Tuesday November 12 2019, @07:53AM (1 child)
Many of us prefer to read the NEC.
https://www.browntechnical.org/content/PDF/nec%202014.pdf [browntechnical.org]
https://www.nfpa.org/Assets/files/AboutTheCodes/70/70-A2013-ROPDraft.pdf [nfpa.org]
http://www.eng.usf.edu/~fehr/files/2008NEC.pdf [usf.edu]
One day, I'll locate a current, online, searchable, version of the NFPA 70. Suffice to say that no searches yet have mentioned oiling, greasing, or otherwise lubricating an electrical contactor. Don't take that as an authoritative "NO" yet. The advice already given, to never lubricate electrical contact surfaces is sound advice. Petroleum jelly will supply fuel for a fire, and should not be found inside of an electrical cabinet.
(Score: 1) by Sally_G on Tuesday November 12 2019, @09:05AM
310-9. Corrosive Conditions.
Conductors exposed to oils,
greases, vapors, gases, fumes, liquids, or other substances
having a deleterious effect on the conductor or insulation
shall be of a type suitable for the application.
(Score: 2, Interesting) by NPC-131072 on Monday November 11 2019, @03:18AM
Cleaned the mechanism and greased electrical contacts using petroleum jelly to prevent arcing. Tested out of circuit using a bench supply and variac - problem was a buildup of crud (oxidization and contaminants) slowing the mechanism from engaging. The rubber seal on the control cabinet had perished and the compressor was in a small room by a grinding area. Didn't trace but basically the contactor was coming from one phase off the control to engage the motor on a big old hydro-vane supplying tool-holders for a bunch of CNC mills at ~100 psi each. Clearly while I wasn't the first in the control cabinet I should be the last, a previous visitor had bypassed a fuse. The thing is a beast and a beat-up, resprayed and franken-fucked pressure vessel like that is a disaster waiting to happen. Told the guy to budget for a modern screw compressor when he's had his electrics redone.
So I'm confident the contactor will hold until the compressor can be replaced but an electrician hoping to bring the shitshow of his wiring upto code... I made it safe but didn't have the heart to tell him that his existing wiring and '50s bakelite wire fuse board will need replacing along with GFCI breakers. 3x 3-phase circuits plus the single phases for his workshop and probably(?) the sub-main feeding his house :-o
Shit like this is why I concentrate on my social justice activism!
(Score: 0) by Anonymous Coward on Monday November 11 2019, @05:04AM (2 children)
Well, sometimes one has to [wikipedia.org]
(Score: 2) by Runaway1956 on Monday November 11 2019, @08:02AM (1 child)
There's no grease in there, now is there? If/when a fumble-fingered tech gets his fingers inside of there, the contactor is no longer good for anything other than recycling.
(Score: 2) by RS3 on Monday November 11 2019, @01:46PM
> There's no grease in there, now is there?
That, and the wiki doesn't mention it, but I suspect it's under pressure- certainly could build up pressure. And it's a sealed system. The point is, there's no air around the contacts for oxidation / burning to occur.
(Score: 1) by paul_engr on Monday November 11 2019, @07:51PM (1 child)
The kill-a-watt is probably a few orders of magnitude too slow to measure the dip you are experiencing on line voltage.
Inrush on a motor load will be 6x operating current, but for a few microseconds(? Not sure, don't quote time scale). You will dip the line voltage enough that your lights will see a corresponding rise in current and net effect will blow first one in the line.
What kind of lights are they? Regular series incandescent?
(Score: 0) by Anonymous Coward on Tuesday November 12 2019, @05:41AM
These are the lights: https://www.amazon.com/gp/product/B0773L1THZ/ [amazon.com]
But yea, I think that the killawatt is too slow to observe the extent of the surge.
(Score: 2) by bzipitidoo on Monday November 11 2019, @01:22AM (5 children)
Assuming USA? The article mentioned Christmas lights and A/C. Now, that could mean 1 of 3 things:
1) They are in the southern hemisphere.
2) They like Christmas lights so much they use them at any time of the year.
3) Global Warming has gotten really bad.
I would guess southern hemisphere, and that English is their native language, so that makes Australia a likely location. South Africa is another possibility. I don't know what S. Africa uses, but Australia is 240V AC.
I had a loss of power to several outlets when the shifting foundation pulled a wire out of a push-in outlet. If the home builder had put in an inch or two of slack, it would never have happened, but they were cheap.
Anyway, A/C takes a lot of power. Best to have it on a separate circuit. Move the other items, or run extension cords to distant outlets.
(Score: 1) by jelizondo on Monday November 11 2019, @01:58AM
South Africa uses 240V AC, with European style outlets
(Score: 2) by maxwell demon on Monday November 11 2019, @09:25AM (3 children)
The article also mentioned electronic equipment drawing 800W and Christmas lights of 100W total. That is, the room is constantly heated with almost 900W (note that even LED lamps put more energy into heat than into light; moreover, a large part of that light gets absorbed inside the room, thus also contributing to the room's heating). Not to forget that every human in the room also adds 100W of heating.
Moreover, the article explicitly mentions 115V, which strongly hints at USA and in particular rules out Australia, since
The Tao of math: The numbers you can count are not the real numbers.
(Score: 2) by bzipitidoo on Monday November 11 2019, @01:55PM (2 children)
Ah, yes, I overlooked the explicit mention of 115. Officially, mains in the US are 120V, not 115V. But I often hear 110V thrown around. The US does have places where it does not get cold during the Christmas season, such as the coast of southern California, Florida, the Gulf Coast, and Hawaii, otherwise I'd be wondering, why not just open a window?
I wonder if the submitter has tried to reduce power usage? There is an 80 Plus program to design more efficient computer power supplies. Before 80 Plus began (around 2005), a typical computer power supply was atrociously inefficient. Might not even be 60%, and 70% efficient was exceptional enough that the manufacturer might even brag about it. Now, pretty much everything is at least 80% efficient thanks to the 80 Plus program. If the submitter has the base 80% efficient, might be worthwhile to spring for 80 Plus Gold, Platinum, or Titanium certified power supplies, which are 90%, 92%, and 94% efficient, respectively. If all 800W are used by 80% efficient power supplies, upgrading to 94% efficient will cut that power usage to 680W. Won't have to change a thing on the computing hardware, no compromising or sacrificing at all. However, bigger savings can be achieved by moving from power hungry computers to low power equipment.
But the biggest problem may be that window A/C unit that was mentioned. The inrush is what's killing things. Electric motors draw much more power when starting than when at full speed. That particular unit has an efficiency of only 10.7 EER, and it may not have good mitigation (or any at all) for the inrush. To achieve Energy Star certification, it should be at least 11, as I recall. There are units with EERs over 12. Also, an A/C that is going bad can draw a great deal more power than it should, and it absolutely will trip the breakers. The submitter might want to try an Energy Star rated unit, which should be much better about keeping the inrush down.
(Score: 0) by Anonymous Coward on Monday November 11 2019, @03:18PM (1 child)
This room is on the gulf coast. This is the power supply: https://www.amazon.com/EVGA-Supernova-Platinum-Crossfire-220-P2-1200-X1/dp/B00KYK1CKI [amazon.com]
Also, it is a rental so I really do not want to replace the window AC unit.
(Score: 2) by bzipitidoo on Tuesday November 12 2019, @03:21AM
I salute you for already having an 80 Plus Platinum power supply.
If it works, the simplest thing to do is put the A/C unit on a separate circuit. You may have to do a little checking to find out which outlets are on which circuit breakers, but that's not hard, and doesn't take too terribly long, maybe 15 minutes. Turn a breaker off, and plug a light into each outlet in turn to test if it's live or dead. Moving things to different outlets so the A/C has a circuit to itself may provide enough isolation.
But if the A/C makes lights throughout the entire building dim when it turns on, then you'll have to do something else. One problem is that many of the possible solutions may be more costly than a better window A/C unit.
You could do things manually. Cut power to the lights and electronic equipment, by unplugging them or flipping a switch, let them run off a UPS for a moment, turn the A/C on, wait 3 seconds for it to settle into operation, then restore power to everything else. Far from ideal for gadget lovers who'd much rather have that be automated, but it can be a quick, cheap stopgap until there's something better in place.
(Score: 3, Interesting) by Runaway1956 on Monday November 11 2019, @02:08AM (1 child)
Regarding loose screws and lugs - we routinely tighten them in our electrical equipment at annual PM's. I can go through the whole electrical cabinet, tightening 480 lugs, 220, 120, and even the AC terminals. Come back a year later, and I'll find loose lugs. The higher energy circuits seem to be worse than lower energy, but you can expect to find loose wires in house wiring. Especially since no one does an annual preventive maintenance on their household wiring!
(Score: 2) by RS3 on Monday November 11 2019, @03:26AM
Yes, same here, I'm always finding and tightening loose lugs and wire terminal screws. Occasionally they're already really tight.
For those who don't do this stuff, on bigger wires you use a torque wrench, which measures how much rotational force you're putting into the screw / lug. It's always a lot more than I think it should be, but you follow the spec. I've been stunned at how loose I've found some. In one case, 208 3-phase 350A breaker, the resistance due to the loose lug heated up the wire and breaker so much it melted back the high-temp insulation (it was at least THHN if not higher temp rating) and tripped the breaker. Breakers will trip for both short-term peak current, and also longer-term heating.
I think some people just don't tighten terminal screws / bolts enough the first time. I'll do a bunch of them, circle back to the first one, and find it can tighten some more.
"Cold flow" is a term I've heard for metal deformation, and it makes sense.
All that said, I did a service call in someone's house where many circuits were intermittent. I found many of the neutral (returns) in the main breaker panel had been over-tightened so as to cut the wire off. So somewhere there's a happy medium.