We've been getting a few more small power dropouts than usual recently, but have just had half a dozen (one of a minute, the rest of ~ 1 second) while I was trying to watch something on Netflix. It typically takes a minute or two to get back the internet connection and re-establish my wired and wireless LAN.
So I am tempted to get a small UPS just to run the Plusnet box (oh, and probably the doorbell camera since it is on the same spreader).
Any particular things to look out for? I was thinking of a conventional UPS with a 13A socket, but I see you can also get Li-Ion power banks that feed the router with 12V dc.
(I have a woodburner and small generator to deal with longer interruptions, but have not needed the latter in 30 years).
I think some kit can be fussy about the waveform of UPS generated power (but I've not found such yet) and the switching time of offline compared with online UPS's.
We have several APC BackUPS CS's here (350-600VA) and they seem to have done what we expected them to do every time they have been needed. ;-)
The good thing about them is they only take one, fairly small 12V battery so whilst they won't support much for long, they will support most of the stuff we have plugged into them long enough and certainly long enough to shut down safely when required (low battery etc). Also therefore, the batteries are relatively cheap and easy to find.The units themselves are also fairly cheap s/h.
(I've collected some in person with dead batteries and offset some of their cost (and certainly the petrol cost) on the scrap batteries) ;-)
It's labelled as obsolete on Elmdene website but available on CPC. I think this is intended for CCTV applictaions but might also work with modemrouters - with the usual caveat that the plug might be the wrong size. Given Elmdene's usual market of professional intruder alarm kit it might be better quality than Lucky Golden Hedgehog.
OK ... let's call that 5 years bar a month, they've just started to fail whenever the UPS switches in/out of buck mode (several times a day here) they report as 83% capacity, but die instantly when I trigger a self test, at £25 each I think I'll go for another set, Yuasa being about £40 each and I doubt they'd last any longer.
Yup. My APC 1400VA needs the batteries replaced every five years or so, its on its third round now. When I left it six years, I had to lever the battery tray out of the chassis as one of the cells had swollen...
You'll also need a mains to DC supply for it as well, voltage matched to your router. Hopefully one of the preset volatges the unit can maintain... This PSU will also need to be capable of powering the connected kit *and* charging the battery after an outage. The datasheet and manual quietly ignore the charging current demand but it says the 4400 mAHr battery charges in 12 to 16 hours so allowing
500 mA ought to be safe.
APC UPS's have a habit of cooking their batteries. The charge voltage is too high for "standby" use (rarely discharged), it's more that required for "cyclic" use (frequent discharge). There may also a be a balance betwwen the time it takes to get recharge the batteries from cut off to a level where it is sensible to switch the power back on. You need the battaies to be recharged enough to main the kit for safe shutdown should the mai power disappear again.
My APC Smart UPS 700 used to kill a pair of 12 V 7 AHr SLA's in three/four years. Got fed up with that, modified the charge voltage and fitted a fan, when I replaced the last set(*) in Mar 2014. It provided power as long as expected about 6 months ago.
(*) Noticed a "funny smell", took a day or two to find until I put my hand on the UPS casing, <ouch>... batteries so hot the plastic was starting to char...
The switching part of a SMPS power supply, is not fussy about waveform. This circuit, for example, would run off a square wave input. Only the input filter might object a bit, and as long as the square wave was a bit trapezoidal it wouldn't be too bad.
formatting link
What did get upset, was the Active PFC (not present in that example item). Early versions did not deal well with square wave input. This may have been fixed to some extent, on newer designs.
Many 80%+ ATX supplies have Active PFC on the front end.
formatting link
"In the case of a switched-mode power supply, a boost converter is inserted between the bridge rectifier and the main input capacitors. The boost converter attempts to maintain a constant DC bus voltage on its output while drawing a current that is always in phase with and at the same frequency as the line voltage. Another switched-mode converter inside the power supply produces the desired output voltage from the DC bus. This approach requires additional semiconductor switches and control electronics, but permits cheaper and smaller passive components. It is frequently used in practice."
You can get "pure sine" UPS outputs, but it'll cost a bit more. A pure sine UPS can run everything, without quibble. (Only motors with high stall current requirements, won't work on anything but very expensive setups. Tesla powerwall maybe. As they're intended as a whole house power source. A typical house apparently needs two of them.)
In this example, "square wave" is green in color. And that's what some battery inverters for camping used to do. Make a pure square wave. The "stepped sine" in blue, is a square wave with the sides moved a bit to more closely mimic a sine wave. The older active PFC didn't really like the stepped sine either. My UPS has that stepped sine output.
formatting link
A pure sine UPS comes a lot closer to smooth and curvy, and even an older active PFC front end would like it. You can see in the scope waveform here, how close they come to a sine wave. A bit fuzzy.
formatting link
Unless you can find evidence a particular appliance takes exception to crappy UPS designs, I'd ignore the details and just buy something cheap.
As for the notion of Lithium Cobalt as an "energy source" for a UPS, *only* if I knew for a fact it was being charged to 80% capacity. A LiCo battery should not be charged to 100% capacity over the long term, as it shortens the lifespan. Charging to 80%, leaves a slightly lower cell voltage and does less damage. If you're going to be using the energy right away, there's nothing wrong with charging to 100% and using the energy and bleeding it off. But just keeping a pack charged to 100% forever "because you might need it", is bad for the battery. Where 80% comes from, is if you stop charging after Phase1, that leaves it at about 80%. It's convenient to design such a thing, as you disable Phase2 charging and your job is done (this assumes the device is programmable and supports this as an option).
Some laptops have the 80% charging option.
The last time I checked, only some 100V UPS in Japan used LiCo inside. Everywhere else, it's reliable Lead Acid SLA instead (10 year life if you take care of it and don't drain the batteries all the way on an outage).
Summary: If one of your appliances makes a "funny noise" when the UPS is on battery, that likely means you were unlucky regarding the sine wave issue. A few select devices may switch off, or even smoke in exceptional cases. Since the industry has not switched to pure sine, this is "obviously" a non-issue.
They have a reputation for that, not sure it's entirely deserved ...
Last time I changed my batteries I discovered the "hidden" commands to read/alter the float voltage, so compared it to a multimeter reading, it was 54.55 V, so only a fraction over the 54.05 volts that yuasa recommends (after compensating for temperature), so I knocked it down one notch.
I wondered about that. I suppose that's the merits of buying a UPS rather than a regular power bank - they're designed for spending most of their life on charge. Dropping the charge termination voltage to 4.1V or 4.05V gives the cells a much easier life.
I expect there's a certain degree of lying as to what 100% actually means. Although I pulled this Macbook Pro off the charger at 100% and it's saying the pack voltage is 12747mV which works out to be 4.249V per cell, so that's the top of the pack. At 98% battery the cell voltage was 4.026V per cell.
Newer Macs have a battery life optimising feature that presumably has a lower charge cutoff. I recorded the reported voltage at 2% battery and it was about 3.5V per cell, so they do cut the cells a lot of slack at the bottom end.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.