Sometimes I hear that you shouldn't plug (UPS brand X / any UPS) into (power strip brand X / any power strip) because of some interaction leading to poorly conditioned power, reduced battery life, massive explosions spattering the room with battery acid, and so on. Sometimes I hear that it's the power strip that you shouldn't plug into the UPS. What I haven't gotten is a clear idea of how reliable these recommendations are or how generally/specifically they apply.
Can anyone speak precisely and non-urban-legendfully on these UPS and power strip interactions, if there are in fact ones worth thinking about?
Having had some 'discussions' with the inspector that comes around our offices once a year to make sure we're not being bad, I have a better idea as to what code says about this. Paraphrased from said inspector:
The inspector wasn't kind enough to elucidate what, exactly, constitutes the 'fire-hazard'. We get dinged on the power-strip in power-strip commandment every other year or so. This necessitated the purchase of a bunch of long-tail power-strips (power strips on a 15' cord), and a few long extension cords with 3 outlets on the ends of them.
Edit: Regarding rackmount UPS's and PDU's. I believe they're OK so long as the PDU plugs into a locking outlet of some kind, such as an L5-20 or L5-30.
I have seen some very, very bad server interactions when you plug a UPS into another UPS, and run a server off it.
In our specific case, the server had a clock that ran ridiculously fast, as in, it would gain 5+ secs per hour. Removing the "double UPS" fixed this.
Granted this is not exactly what you're describing, but I would say based on my experience that nothing should be run "upstream" of any quality UPS. Plugging a power strip in "downstream" of the UPS may not be up to fire code (per @sysadmin1138 excellent response) but it's probably not going to hurt anything.
I just spent 6 hours trouble shooting an application issue on two systems due to this very reason.
Customer reports application Ultra Tax 2010 software taking 8 minutes to open on Workstation 1 and Workstation 3. This same software application is working correctly on Workstation 2 and the server.
System changes being reported by client leading up to this issue.
Actions take to troubleshoot this issue:
We decided before spending $250.00 to call Microsoft we should rule out the new UPS. Disconnected the APC power Strip which had the Sonic Wall, Netgear, phone system, and the cable modems AC adapters plugged in, from the UPS and connected directly in the wall. We then tested the Ultra Tax software on workstation 1 and 2. Ultra Tax software is now responding as it use to. We went through the process of opening and closing the software multiple times each time having the same positive results.
We then decide to verify the UPS was truly the issue by reconnecting the APC power strip. Again the Ultra Tax application began to hang. We adjusted the sensitivity setting on the UPS to low. The issue with Ultra tax was still there. We then disconnected the SonicWall and the Netgear Switch AC adapters from the APC Power Strip and connected them directly in to the UPS.
We then tested the Ultra Tax software on workstation 1 and 2. Ultra Tax software is now responding as it use to. We went through the process of opening and closing the software multiple times with the same positive results. We believe this issue is resolved.
By Connecting the SonicWall, Netgear Switch, directly in to the APC UPS with the sensitivity level set to low has resolved the Ultra Tax issue of taking 8 minutes to open.
Do not hang a surge protector off your UPS, as it will waste a big % of your power, it has to do with the way the cheap surge protector interacts with the non sine wave power from the UPS.
Strictly speaking, don't hang a surge protector that uses MOVs for protection off a UPS that doesn't put out a real sine wave (most are kinda squarish, line power is a sine wave).
http://www.codinghorror.com/blog/archives/000632.html
We've had cheap power strips blow up when connected directly to a wall (with poorly conditioned power coming from a bad transformer), and we've had the same happen to a UPS. When our transformer was on the fritz, the bad power would flow through the power strip, but wouldn't flow through the UPS (we had printers die that were connected to a power strip, but nothing die connected to the UPS, other than the UPS itself).
That said, I'd never plug a UPS directly into a power strip, but probably wouldn't hesitate to plug a power strip into a UPS if the power strip was decent and powering lower wattage items (lights/switches/etc).
I am no expert, but...I grew up working for my electrical-contractor father and, having seen the inside of power strips, I dont see what the harm would be in plugging 'just' a UPS into one. Now, if you plug in a UPS and a space heater there may be issues, but just a UPS and maybe something else that doesnt pull a lot of current, should be fine.
Also, I dont think I would plug a powerstrip into a UPS. They are designed to provide a certain amount of temp power. So plugging in a loaded strip may overload the UPS.
Coincidentally, we just made our first UPS purchase. We bought APC and tried to buy one for every critical workstation and every server. I had no idea how expensive they were!
The only thing other than the wall that it's acceptable for a UPS to be plugged into is a line filter or conditioner, as not every UPS is designed to provide that kind of protection. This device should be completely passive and shouldn't have any sort of battery or intelligence for the reasons given in other answers, but it may be a necessity in an area with poor or erratic power quality or one prone to surges, especially if you have particularly sensitive equipment (e.g., medical devices) attached.
The problem might not been the power strips per se. It might be the number of devices you can now hook into one source for electricity. Increasing the number of parallel circuits increases the amount of current being drawn through that power source. The more current, the more heat produced. If you are lucky the breaker will trip before it gets to be too much for the outlet and the devices on the daisy chained strips/UPS. If you are unlucky and/or the wiring is not up to code you will start an electrical fire.
You must also remember that the flow of electrons is limited by the size of the pipe/wire it is going through. Even if you have great wiring there is always a chance you will reach the maximum capacity of electrons the wire can carry. This means devices will be fighting to grab the flow of electrons. Many will have their digital electronic components fall into indeterminate power levels on the chip power pins (Digital is on or off - it needs input power with a certain voltage on the power pin, if it falls to less power than tolerance it becomes indeterminate). This means the electronics will begin acting erratically or stop functioning.
Well here's what I think. There are three issues at hand. UPS into UPS, UPS into a strip and a strip into a UPS. If you consider how a UPS works, it seems that it would not make technical sense to plug one UPS into another, here's why. A UPS basicaly takes (multiple) 12 volt DC lead acid batteries, and chops ups the available direct current, and coverts it into (typically 115vac) alternating current. BUT the output-ed "sine wave" is not a true sine wave, at last not equal to the quality of the electric company supplied ac current. However some UPS high end suppliers do advertise "true sine wave output", at less than 3% total harmonic distortion. I think its asking for problems to supply a UPS with a product producing a sine wave to another product producing a sine wave. A "power factor" problem could exist also, and increased distortion as the load demand goes up. The uPS should be plugged into the ac wall outlet directly, unless its a really small unit. The isuue of plugging strips into a ups is not probalmatic. Just keep an eye on the leds on the front pannel indicating total load of the devices. I would like to hear from an electronics engineer regarding the above, as I am a technician.