With lead-acid batteries, a trick I've done a few times is to use them for jump-starting a car. Just make sure you're really using a ~12 V cell, as some UPS batteries are actually 2 wired in series.
I appreciate the review. Overall efficiency at different load levels and standby power consumption of the device itself would be of interest to me, with power prices developing as they are.
Also, would it be safe to replace the built in battery with a LiFePO4 battery of the same size? Many battery vendors now offer LiFePO4 in lead-acid form factors, advertised as drop-in replacements since they have a built in BMS that makes them compatible with regular chargers.
They are lighter, can't leak and supposedly last far longer (in both time and cycles). I noticed that they have much lower discharge values than lead acid/AGM batteries of the same size, so I doubt they would be safe to use at full load, though the BMS should cut off there. Perhaps something the author could investigate in a future content piece.
I've modified a few APC UPS to use LFP, and so far still good. BMS is a must for safety. And if one planning to use >50 AH battery for longer usage, remember to put a reliable fan that blow directly to the transformer.
A pity that there's almost no LFP battery in market.
There're a few UPSes with lithium batteries of some sort (never dug deep enough into the specs to find the exact chemistry). They're bafflingly expensive though.
ex APC's 1500VA model is $1800 on amazon.
But if you look at the specsheets for their conventional lead acid ones, the 1500VA model can provide 120W for 1 hour. 120Wh is only about 20% more capacity than a max size laptop battery; so UPSes using them shouldn't need to be crazy expensive. Once you factor in the far lower shipping cost I wouldn't be surprised if the delivered price actually ended up being cheaper.
Yeah, those "bafflingly expensive" UPS companies are going to be driven out of the market soon enough. By the likes of Ecoflow, Bluetti, and other endless "solar generator" brands (anker just got in on the game too)
They are selling products that instead of just giving you time to quickly shut down, will allow you to go on for hours of uptime if the power outage goes on for a long time, and they all allow you to charge the "UPS" with solar panels to further extend your uptime for a very long time. (with a big enough system, you can make your computer independent from the grid)
These companies are all using either ternary batteries, or LFP batteries. The ones with LFP batteries are monsters that you can cycle every day, for years without any noticeable degradation. (those would be the ones to use if you want an offgrid computer)
UPS companies are being disrupted and they don't even know it yet, I wouldn't be surprised to see many of these file for bankruptcy in the coming years, as these companies get their UPS capabilities more polished (some don't switch fast enough for critical loads yet) and up their marketing game as "UPS" instead of just "solar generators"
True that. For countries that have more day than night, you can go off grid if planned right using solar panel and LFP. My dream is having a DC home. So there's no need to convert DC to AC. Just have to use boost or buck converter to up or down the voltage. Better efficiency and fewer energy loss.
So you are planning to install new outlets? Or just with an armageddon of extension cords? I think it makes more sense to spend the money to have a electrician hook your battery up to the grid, if you live in a place where your output can get adjusted by grid demand automatically, it can pay for itself very quickly.
Sadly, almost anything that has electric motor inside, requires AC. Where you don't have access to gas line, induction is also a must. Do you really want to run 7-8 kW to induction stove or 2-3 kW to oven with arm-thick cables with DC? The best solution is to have dedicated circuits to both. I can picture a variation of UK type plug (because you DON'T want US-style ones for that... or for anything, really) that could handle both AC and DC from same outlet with complete separation of circuits and safe disconnection even from high current DC.
I absolutely see the worth of dedicated DC circuitry, however for it to be safe, the craftsmanship during installation would have to be impeccable, with highest quality materials, and let's face it, depending on the size of the house (or apartment) it would be very expensive.
I don't know what equipment from those brands you're looking at, but most of what I can find, while higher capacity and more feature-packed than this, are an order of magnitude higher in price than this is too. So while high-end UPSes might be in a process of being "disrupted" (god how I hate that nonsensical term), the one under review here certainly isn't.
Yes, there's lithium UPSes, but as far as I know, it used Li-Ion or Li-Po, not LFP. 500-800 Cycle vs 3.500 cycle. LFP has far more cycle, and safer too.
And yes, Those UPSes with Lithium is expensive as hell. Better modify the lead acid ourselves. For $1.800, you can get 24v 200AH LFP + BMS + connector + cable + fan + APC SMT750IC (pure sinewave) and still have changes. 24v 200AH = ~5KWH, translates to days when offline.
I wouldn't because LiFePO4 batteries differ dramatically from lead-acid batteries in charge requirements (current curve, voltage and such; for one, lead-acid is charged typically at 14.4V which is much too high for LiFePO4). Unless of course you plan to replace the charge circuit at which point you might just as well build it from scratch.
Since the writer makes the statement that UPS is making a comeback, and makes some fairly strong technical comments about the quality of the components and circuitry, it would be very interesting to see a review with a mix of new and better-known brands in their current incarnations pitted against each other, to indicate the current state of the market.
Another suggestion: my current UPS (APC Smart-UPS C 1500) had pretty bad coil whine, for the first few years I owned it. This happened when it was on normal A/C power, and I reproduced it both at home and where I work. I even returned it under warranty, but the replacement did the same thing.
So, I'd highlight this as an area to watch out for. In a quiet room, that can be a real annoyance.
Thanks for the feedback. I will surely point out such issues if they occur.
The most common cause of coil whine is poor quality power input. Harmonics are to blame 90% of the time. That clearly is not the case here though, since you had the same problem at two entirely different places. Someone probably just did not glue down an inductor that should have been glued down.
I own two Cyberpower CP1500PFCLCD PFC Sinewave UPS (1500VA, 1000W), one my Ryzen 7 2700X and the other for my ThreadRipper Pro 3955WX. And a friend of mine - on my recommendation - bought the same for his ThreadRipper Pro 3955WX.
So far, I have had no issues. I don't get many power outages, but when I do, they have always worked.
Same. Well, almost. Mine is a slightly older model with the only real changes being 2 x USB type A and a different LCD. It's worked great for me. 2 x 8.5AH batteries too not the usual 7AH. I'd replace them with 9AH ones but it seems like too much work for too little result.
Definitely like the UPS review(s) though, keep them coming!
Also, I wonder how feasible it'd be to test or at least teardown some Ethernet surge protectors. Right now, I've got my cable modem on its own small UPS, then a gig-E surge protector separates it from my router. If that surge protector turns out to be garbage, then I might as well just run everything on a single UPS.
FYI, the one I'm using is Tupavco TP302, purchased on Amazon.
Probably not, but most modern electronics have 100-240V PSUs; so all someone wanting to use it in the US would need is to wire a 240V outlet and then import some device powercords along with the UPS. So we could use it if we really wanted to.
Most of the UPS units I saw (that was a long time ago though) had detachable power cords of the "PC" style. If this is the case, powering up the UPS is no issue at all, as the PC power cords in local style are easy to find. As for output, there are "universal adaptors" (but they might not support higher amperage)
*A lot* of UPSes have the standard IEC 60320 C13/C14 connectors https://en.wikipedia.org/wiki/IEC_60320#C13/C14_co... I don't think I've seen any with a hardwired power cable (Europe, mind you). The more home oriented ones used a C14 connector for mains power and regular power outlets (instead of C13) for connecting the devices since most people want to connect a whole host of devices not just whatever you can connect with a "UPS/PC cable" https://notebookcentre.am/wa-data/public/shop/prod...
I used to just replace the cable of a multi outlet extension cord so it has a C14 plug at the other end and plug that into the UPS so things with a power brick would be easy to plug in.
Looking on Amazon, all the UPSes I see - including a few rackmount models - have hard wired power cords. I wonder if US regulations treats them as overgrown powerstrips (which I've also never seen with a detachable input cable).
The EU model reviewed here also has a hard wired cable for input power.
Glad to see a UPS from a vendor that isn't Schneider Electric.
Still I think there is very little point in a UPS with a 5-10 minute runtime on a desktop or server. I mean, you get 5 or 10 minutes of terror to save your files, whereas it is routine for a machine to come back after a power loss and not even have to fsck. If you skip the UPS, you might lose 30 minutes work of unsaved work in Word and Word will never stop pestering you about the files you didn't save, even 10 years from now, but you didn't pay $250, you passed go, you didn't have to deal with noise, acid fumes, disposing of lead-acid batteries, fire hazard, etc.
A more realistic mission for a UPS in 2022 is to support a small amount of equipment. Back when we had DSL from the central office we had a UPS set up to power one DSL modem and a cordless phone with an answering machine. This gave us 21st century comms that we could tap with a laptop or a tablet for a whole weekend.
They upgraded our DSL for a 10x performance boost but now it is dependent on power at the crossroads near us, probably beta(10,2) the power goes out at the crossroads if it goes out at our farm, so the UPS powered a DSL modem that didn't do anything so we donated the UPS to a reuse center and crossed our fingers we wouldn't get charged for "donating" e-waste.
Unfortunately keeping a few PCs running just long enough to shut down safely seems to still be the design target of all UPSes.
A half dozenish years ago I tested a 1300VA APC and 1500 VA Cyberpower UPS with new batteries. They self-discharged in 5-6 hours with no external load; with my cable modem and router attached they lasted between 4 and 5 hours.
That implies that a design made to be efficient at low load could keep my wired internet running as long as my phone/etc batteries last. If an efficient at low load design existed.
I have my router, gateway and a ye olden land line phone hooked up to an old APC Smart-UPS 620VA - it's enough to keep the internet going for a solid six hours when the power goes out, which it does fairly often (weather). I end up changing out the batteries every four years or so, but it's well worth the convenience to be able to check for power company updates, keep in touch or otherwise continue to while away the time on a tablet or phone.
My main desktop has its own APC Back-UPS 1500VA, enough for about 30-45 minutes of normal desktop usage, so often the power will be restored before I need to shut down (or set it to hibernate). If it's particularly nasty weather, I'll shut down the desktop ASAP and switch the UPS off, in case I need to plug something else in later (LED lights, phone/laptop charger, etc).
I think for most people, the reality is that 4-5 minutes to be able to immediately save their work and shut down gracefully is pretty realistic. Just having that option is a bit of a psychological safety net.
> I think there is very little point in a UPS with a 5-10 minute runtime on a desktop or server.
I have occasional brown outs and outages lasting only a few seconds. A good UPS has saved my bacon so many times, it's not even funny.
I've even done a cute thing where I put my desktop in "sleep", powered off everything else, then called the utility company about the problem, and they managed to fix it before my 1500 VA UPS ran out of juice.
> it is routine for a machine to come back after a power loss and not even have to fsck.
For servers, you're supposed to have it automatically trigger shutdown, to avoid filesystem corruption. You can "play with fire" and just let them lose power, but even if storage setup has no "write hole" and is correctly using write barriers, there's always the potential for application-level data corruption.
> If you skip the UPS, you might lose 30 minutes work of unsaved work in Word
If you want to take that risk, then so be it. However, not all of us are simply editing Word files or using MS Windows desktops.
> now it is dependent on power at the crossroads near us
FWIW, I've never lost my cable internet when I've lost power. I'm sure if the outage were sufficiently widespread and long-lasting, it'd be a different story.
"I have occasional brown outs and outages lasting only a few seconds. A good UPS has saved my bacon so many times, it's not even funny."
Exactly. Where I live, the power blinks a few times a year. About a year ago, it happened a dozen times in a short period, including a double flicker that caused Windows to trash my user profile, losing all my stored passwords and stuff. I got a UPS after that, put the computer and one monitor on it, and now I can weather any number of these few-second outages, and there's plenty of time to shut the machine down if the outage lasts longer than that.
If you don't want lead-acid, still want to use something akin to UPS and if the office mainly using DC tools, you could buy 12v/ 24v LFP. Charge it using SMPS PSU (AC to DC) with passthrough to the tools with buck (step down) or boost (step up) converter. No noise, no acid fumes, small probability to caught fire. And a good LFP battery have ~3.500 cycle. Quite simple to built, but have to be careful.
For my office, I'm using 54W (depends on loads) 14v adaptor buck to 13.34v to charge my 100 AH LFP, powering 24" LG monitor, a modem, a router, and a 15w laptop (step up to 19v converter). Been ~ 2 years now, and no problem so far.
I'm probably in the minority, but I have an 8 hour whole house battery backup. The UPS is here at my desk to gap the moments it takes for that to kick in.
That said, I've got two computers, my network gear, and a 38" monitor plugged into a small-ish Back-UPS Pro 1000S and the load currently says 72W with an estimated backup time of 53 minutes.
Also, the majority of power outages in my area (maybe once a year) last less than a minute.
A friend of mine has a gas generator that takes a few seconds to kick in, so I told him to buy a Cyberpower CP1500PFCLCD PFC Sinewave UPS for his ThreadRipper Pro 3955WX. And it works.
Not sure where you are from but I took a battery just the other day to Home Depot. There was a bin just for batteries you could toss them in. There are other places that do that too. And yes, a UPS on a modem/router is very nice to have. Keeps your internet up and you can use a tablet/laptop for plenty of time.
I'm just one person (physically, mentally... that's still undetermined) but I liked this. With more people (like me) working from home hopefully there's enough interest to see more of these. I'd personally like a little more educational commentary, either as part of reviews or guides, but hope to see more of these.
Great review - thanks. It could have done with a mention of the USB socket around the back -- it should be there so attached PCs can power down safely when the unit goes to battery power. A few words on this and any provided software would have been good (can one PC send a shutdown signal to another?)
Also, you might want to mention that European grid frequencies vary a lot more than this unit.
I am interested in seeing some APC reviews, especially at 1k+. APC (Scheider Electric) still seems to be the market leader in North America, so would love to see their products reviewed. My favorite does seem to be the 1500VA spot, so I'm curious to see whether the savings on the cheaper CyberPower is worth it.
Same here. I've a few APC ones which I bought when CompUSA (brick and mortar) died and was selling them at a quarter of the price by the last week. Otherwise I have many more CyberPower. No problems with both brands. Both brands did their job several several time during power outages. Meaning: the power went down for a few seconds and the computers did not mind. Or the power was out for hours and I had time to do normal shutdowns (automatic or manual). Contrary to APC, the CyberPower Windows Server software app was free. I don't really care anymore because I switched to Linux years ago. I also got several additional APC and CyberPower for free from a company that did not want them anymore and was on the verge to e-waste them. Like my first generation of UPSes I just replace the batteries and they were like new. All the models I have from both brands use the same battery anyway: any equivalent to DURACELL DURA12-8F2 12v 8Ah AGM. I went to a local brick and mortar battery retailer. He smiled and gave me a good discount when I ask the price for 24 units :-)
I had a Cyber Power CP685AVRLCD fail pretty badly on me, after about 5-6 years of use. I seem to recall it emitted a strange noise and started emitting a funny odor, not prompted by any sort of power event that I could discern. I can't remember if it also cut power to the outlets, but I think so.
We've had a couple APC units fail, at work. The most common thing I see is that a unit appears completely bricked, but the issue is that its battery is simply too dead. As a result, it gives no signs of life when plugged in. After installing a fresh battery, they seem to return to fully working status. However, I'm not counting that as "failed" - I just mention it as a caution to anyone who thinks they need to dump an old UPS that *seems* dead, after not being plugged in for a long time.
Keep your UPS's plugged into A/C, even if you're not using them! Lead acid batteries DO NOT like deep discharges!
Anyway, we have some APC rack mount units are probably 20 years old and still kicking. They're not on anything mission-critical, so as long as they don't catch fire or something, there seems to be no need to replace them.
Thanks for the review. A couple of features I look for: - is it passive or fanned/what's the fan noise level? (thanks for that) - can alerts be silenced? High electronic beeping really freaks out my dog. - when it's running on batteries or in AVR mode is the output ugly enough to cause a PC's PSU to make ugly noises? - Is there a master/slave mode where slave outputs turn on/off when the UPS detects load/no-load conditions on the master? A really useful function on older APC units. Put the PC on the master and the monitor and audio amp on the slave.
> can alerts be silenced? High electronic beeping really freaks out my dog.
Although there's frequently a front panel button to disable the beeps, you'll often find that even models without this will allow you to disable the beep via software, after connecting to its USB or serial port. You can also get some useful and interesting information that way, as well.
That reminds me of a funny story about an older model of APC Smart-UPS. It has a 9-pin serial port, on the back. So, I install the APC Power Chute software and connect it to my PC. As soon as I complete the cable connection, the USP goes *dead*. I unhook it, and it turns back on (I forget if I also had to hit the power button... probably I did). Then, I check the docs and see that the friggin' thing turns out to have a nonstandard pin-out! WTH?? Now I get why their newer models with serial ports don't use a standard-looking 9-pin connector, but that's just *bad*.
tl;dr - Make sure to use the included cable, if you have an APC model with the 9-pin connector on the back!
I discovered this accidentally, when I turned off a co-worker's monitor as I was leaving the office, one Friday afternoon. I thought I heard something else turn off, but I thought nothing of it. After the next time, he figured out what was happening and changed which device was plugged into the master outlet.
I think I've seen some units with a button to disable this behavior. IMO, it should be disabled by default. I'd even previously *seen* the "Master" label, but had no idea what it meant.
Excellent article, expertly written by a knowledgeable author I hope that future articles will include graphs / pictures with all reviewed models included. Such a database might become dated, but that also allows us to see how far we've come.
One thing I did notice was the clipping of the output waveform of this particular model under load. Without comparisons, it is difficult for the average person to really understand what that means for his equipment, so feel free to continue to comment in an unbiased manner. Hope that AnandTech is paying you enough that you can avoid 'sponsership'.
All of these companies are going to be driven out of business by the "solar generator" market soon.
Those guys are making very cheap LFP battery powered, pure sine wave inverters at very competitive prices. (at least compared to an UPS)
The only thing holding back those "solar generators" is that a lot of them don't switch fast enough yet, but honestly, even with the current 15ms switching times I would take them over a lead acid UPS.
Sounds like this might do the job for me. I want it for powering a couple of guitar amps strapped to a motorcycle to play offensive music at people who have glued themselves to roads in order to "protest" insulation. They never do it in winter, only when it's warm, which is superb for battery life and riding.
Excellent idea. I'm deaf, if you find me glued to a road please play your music nice and loud so that I can enjoy it properly.
Alternately please ask your MP to ask the Govt to follow the recommendations of its own expert advisors and spend some money on insulation. If it had done this when the advice first came out, the UK would be in a much better position with regard to being able to reject buying Russian oil and gas. As it is, your taxes are going towards giving billions to *both* sides of a bloody and genocidical war. I think that's worth protesting about.
I read quickly but saw nothing about the fact that lead-acid batteries are permanently degraded by being discharged. The article mentioned a 12-hour recharge time for when the batteries have been completely depleted but didn’t mention the far more pertinent matter of lead sulphate. That is the chemical that forms when lead acid batteries are discharged, even partially. It is a byproduct of the degradation inherent in the tech.
Lead-acid batteries have no place in a device like this. People who wish to debate that should truly look at the big picture. This is not a device like a car which has a workload defined by the battery usually being mostly fully-charged (the one workload that is suitable for lead-acid tech). The article said the batteries in this unit can be depleted in a matter of minutes. This device is designed around much heavier depletion than a car, making it fundamentally unsuited to the battery tech.
Why do manufacturers choose to use lead-acid batteries for devices with the wrong workload? They’re counting on buyers looking at things like sine wave output and wattage — the same thing this review does.
Perhaps the business model here is to sell replacement batteries to make up for the reduced pricing of its partial sine wave ability.
The irony of consumer-grade UPS devices has been mainly about how prone they are to dying. A computer surviving multiple lightning storms with a simple surge blocker is more likely than these UPS devices lasting long enough.
A useful review will not only warn about the unsuitability of lead-acid batteries, it will be in two parts — the second being a long-term lifespan report. See if you can get a few years out of these products. See how much uptime lead-acid batteries manage after multiple discharges.
I can see using lead-acid batteries in a device that has an optional ‘emergency mode’ where the user can decide to engage them when the lithium batteries are nearing depletion. This could be a cost-saving way to give the unit longer uptime at a lower price point — suitable only for workloads that are unlikely, but not completely, to need the extra runtime.
what needs regulation for advertisement and technical description for standardized consumer support and information (and consumers understanding their usage profile, planed/scheduled repairs/upgrades, replacement parts available and costs including wear/maintenance/refirmation, before purchase) (?)
> Lead-acid batteries have no place in a device like this.
The typical usage model is that the user saves their work, shuts down their PC, and then switches off the UPS. You're not necessarily supposed to run until the battery is 100% depleted.
> A computer surviving multiple lightning storms with a simple surge blocker
Surge suppression won't save you from lost work and possible file or filesystem corruption due to abrupt reboots.
> the second being a long-term lifespan report.
Here's my lifespan report. All of my co-workers have APC UPS units at their desk. Sometimes, we lose power when people aren't around to switch them off, and most haven't gone to the trouble of configuring auto-shutdown. Yet, we still typically get about 5 years out of these batteries (possibly more). At home, I've gotten over 7 years out of mine, which I've run down probably at least 3 times.
Yes, Lithium Ion would be great. However, as noted in other comments, there's a cost differential. And Lithium Ion have other tradeoffs.
You make some fair points, especially with your reminder about minimizing depletion of lead acid batteries. I definitely want to acknowledge that.
I generally get 4 years/battery pack at home; OTOH my high power desktops (running distributed computing 24/7) mean I need a larger than normal chunk of the nameplate capacity to avoid an immediate shutdown on power loss.
And with maybe 2-6 interruptions longer than a few seconds/year decay due to discharge cycles is basically irrelevant. (I find the person upthread ranting about the use of lithium batteries only good for several hundred rather than several thousand cycles upthread hilariously divorced from how typical UPSes are used.)
Defining that lifespan is also fraught since it varies wildly by the circumstances and how each person defines the lifespan. One may claim the product is still working when the battery can only manage 15% of its original charge. Another will wait until the product simply doesn’t power up or whatever.
I define "lifespan" as "when the UPS tells me to replace the battery". I have no idea what that actually means, in terms of raw capacity, because my experience is that a UPS in that state won't even *try* to give you any runtime.
I would advise people not to buy UPS batteries in advance, because they have some degree of self-discharge. And, as we've discussed, you don't want to them to have any significant amount of discharge.
For the same reason, I advise people to keep every UPS plugged-in (at least some of the time), even if they're not using it.
The numerous units we (myself and colleagues) have purchased have failed. It hasn’t merely been the batteries. That includes models claiming sine wave output. I am thoroughly unimpressed with what I have seen so far in terms of ‘affordable’ UPSs.
‘The typical usage model is that the user saves their work, shuts down their PC, and then switches off the UPS. You're not necessarily supposed to run until the battery is 100% depleted.’
This device can be depleted within minutes. How quickly is someone going to prevent the batteries from draining at all, let alone completely? People are not always at their PC. People take bathroom breaks, get snacks, et cetera.
Lead-acid tech is fundamentally incompatible with this usage model. It is that simple. The only usage model that is at all justifiable is having lead-acid batteries as an additional optional emergency capacity mode, to supplement a typically-adequate (capacity) lithium primary battery.
There is no justification for polluting the planet with batteries that are intentionally lifespan-shortened. The biosphere is in bad shape already, without being intentionally more reckless.
Even for those who don’t have a conscience and thus care not about ecological product justification, they need to grant that basic logic shows there is a severe mismatch between the workload and the solution chosen.
Runtime is determined by capacity / load. That it *can be* doesn't mean it *always is*. Even if you buy the minimum capacity needed for your load, that would be *peak* load. As we all know, modern computers can burn a lot of power at full load, but not much at idle. And, unless you're gaming or doing other heavy-duty computation, PCs are typically running closer to idle than peak.
> People are not always at their PC. People take bathroom breaks, get snacks, et cetera.
Yes. As I mentioned, my workplace typically gets about 5 years out of a battery, even though there's probably an average of 1 full discharge per year.
> Lead-acid tech is fundamentally incompatible with this usage model.
It's worked well enough for 3-4 decades. True, it's not ideal, but the batteries aren't a major cost item for the amount of time they last. I mean, relative to the cost of what's plugged into the UPS and its rate of depreciation.
> There is no justification for polluting the planet with batteries
And here's where I figured we'd end up. Like all things, it's a matter of cost/benefit. People should dispose of their batteries properly, but of course that process isn't perfect and not everyone is going to do it. So, I agree that lead acid is a dirty technology, but Lithium Ion also has environmental impacts. As and when better technologies become competitive, I'd like to get away from the tremendously heavy & dirty lead acid tech.
> basic logic shows there is a severe mismatch between the workload and the solution chosen.
No, it's not severe. You're overstating your case, which distracts from your core message.
Seems a lot of people think a UPS is equivalent to a power wall and don't really understand the designed use cases. They're for short term use so you can switch from one power source to another, prevent brownouts or enable automated clean shutdowns. I run a 3K VA UPS at hone that's attached to the mains and an LPG/Petrol generator. You'd be surprised how often brownouts occur, even last night it protected my machines:
Date : 04/19/2022, Time : 02:36:12, Code : 0x0109 Warning - UPS: On battery power in response to distorted input.
Date : 04/19/2022, Time : 02:36:13, Code : 0x010A Informational - UPS: No longer on battery power.
Given the fact that lead acid batteries are the top recycled consumer product in the US with 99% recycling rates, one could argue that lead acid is far cleaner than any Lithium based alternative. The batteries are also standardised and you don't get locked into using a vendor's specific replacements.
‘Seems a lot of people think a UPS is equivalent to a power wall and don't really understand the designed use cases.’
That explains the use of lead batteries when people can need to do things like use a restroom or answer the door. Lead batteries are permanently degraded within minutes with these devices.
> Lead batteries are permanently degraded within minutes with these devices.
Your characterizations simply don't match my experiences. Your warnings sound very dire, but the 5-7 years I've gotten out of these batteries would argue otherwise.
I've long used AGM lead-acid batteries in my cars. I've even treated them somewhat roughly, especially during the pandemic, and yet still averaged about 5 years out of them.
If you want to keep pressing this case, then please provide some data on the specific battery tech used in these UPS units.
You don’t understand that lead acid batteries don’t work like lithium batteries. They are degraded when they are even partially discharged. The only workload they’re suitable for is one where the battery is only slightly discharged, like with starting a car.
Very nice and technical review, as always from mr Fylladitakis. The only thing I believe is missing is a measure of the UPS power consumption while on standby with the batteries charged. In my experience, this consumption can range from 10W (very good "green" UPS) up to 50W (typical APC UPS). This consumption can contribute significantly to the Total Cost of Ownership of the UPS.
Not unless it's charging. Common sense should tell you that if it's burning 50 W, its fan must run to keep it from overheating. The only time mine runs is while charging or discharging. And I have a 1500 VA model, which is at the upper end of capacity for a desk-side UPS and even has a backlit front panel LCD. Also, it's a model with Sine Wave output.
APC publishes Load/Efficiency specs, on their website. For my SMC1500, they claim:
I had this 50w while on standby figure on 2 Mustek 2000VA that I used to own. I had a powermeter in the wall socket and was amazed when I noticed this. This was constant, not while charging.
I think this is a great idea/category, but I'm surprised you chose a European-market focused brand that is not really available here in the West.
Perhaps you could consider doing reviews on the top two UPS companies, APC and Cyberpower? Expand this review to include comparisons with Cyberpower's CP1500PFCLCD and APC's BR1500MS true sinewave systems?
If you have any contacts with those companies that could help convince them to send a sample to Greece I'm certain Ryan Smith and E. Fylladitakis would be grateful for your assistance. 🙄
Nice review. Yes, UPS reviews are definitely useful. And the noise test/commentary Ishould definitely be include for any UPS that may find its way to a home office /setting. One thing toe consider perhaps, speaking from experience, is the not all surge/power cut events register trigger a UPS (standby or interactive) to take over. For example I had a power cut trigger it while a milliseconds "on/off" type of thing did not and the PCs simply lost power. So you may want devise a couple of tests that are not instrument based or linear but more brutal ( pull the plug momentarily /tripping over the power cable for example, switch power on off, vary the timings, etc ) that are closer to what may happen to some of us. Still, UPS reviews: nice keep them coming..:-)
> I had a power cut trigger it while a milliseconds "on/off" type of thing did not and the PCs simply lost power.
I've never seen a UPS fail to intervene in a power event that resulted in a PC rebooting. I'm not saying it doesn't happen, but not with the APC UPS units that I've used at home and at work.
Could you perhaps determine what the trickle\standby voltage of the battery charging circuit is? I've seen a few models of APC UPS's lately that seem to overcharge the batteries to 12.8-13.1v, dramatically shortening their lifespan. One set of batteries from an APC XS1000 were so swollen they had to be pryed out and the unit\batteries were only 3 years old!
For funzies YEARS ago when li-ion battery banks first started to become available for about $250, i found one with a pure sine wave and used it as a battery for a cart in a warehouse with 20" monitor, laptop, and label printer.
That worked fantastic, then i got another one for a business with super flaky power that went in and out and at times read 90v from the wall, and computers and printers would have issues working. This time not only did it work as a UPS but also a line conditioner, you see these on servers all the time (there not just power strips sometimes). Basically the wall would charge/power whatever it could and the inverter would always be the one sending the power as a true "uninterruptable" power supply. It would correct, say 90v or 130v to a very stable 120v at the cost of loosing a bit of efficiency and also act as a far better surge protector. Surge suppressor should be your minimum when protecting your equipment, it ensures you don't get over voltage or "dirty" power. Anywho, a li-ion powered ups should be the norm for spending $200. Not only does it have more power, also smaller, charges much faster (recovery times between outages). Specifically LiFePO4 (lithium iron) batteries would be best, about 2-3 times the duty cycle of li-ion and higher amperage (can be smaller/less capacity and still have enough power for devices).
As should product truthfulness, rather than claiming sine wave output when that begins to fail beyond 25% load. Sadly, humanity is full of falsehoods for ‘profit’.
You have yet to give one modern example representing a viable alternative. Every modern attempt at true socialism thus far has failed. On the other hand, there are examples of regulated capitalism that continue to work well to this day.
tangential topic - generators. It seems inverter generators are raved about as being superior, but no one ever seems to mention whether they have a true sine wave output or a modified sine wave output
A flippant statement like that implies a fairly substantial amount of reverse-engineering that must've gone into such a conclusion! And it would be more persuasive if the intermediate steps were shared -- how was the reverse-engineering performed? What topology does the unit use? What's outdated about it? How does it compare to others on the market?
I think the UPS market very much needs technically-competent reviews, and this is an encouraging step in that direction. But it could stand to add some rigor and detail. Thank you for the promising start!
Nice work - an nice new section - keep them UPSs reviews coming. Speaking form experience, line-interactive UPSs may sometimes not respond consistently to power interruptions. I've had one (MGE Eclipse) that for example you would pull the plug and it would work as expected. But for minor power interrupts it wouldn't... so maybe its a nice idea to devise a few power-off scenarios just to make sure the UPS works as expected in a range of scenarios. It wasn't clear tom me in the review if already do so.
Another +1 for these types of reviews, and additional suggestions:
1. What is the UPS power draw from the wall when not charging/under load (as mentioned by others).
2. What status monitoring / remote shutdown capabilities are present. In addition to running the local "Powerchute Personal Edition" type software connected via USB, are there remote shut down agents when connecting multiple devices? Are there fleet management capabilities if I have one under every desk, etc.
3. Low voltage protection (also alluded to, was another interesting question). IE: In the event of an outage, does the battery run to 0% (not grate for Lead Acid), or is there an x% cut-off to protect the battery
4. Parasitic draw when the unit is powered off. Can I keep the unit offline (ie: not in production) but with batteries internally connected without damaging the batteries? There is a natural discharge rate for lead-acid, but if there is parasitic draw even when off, those batteries could be toast when I pull the unit from the cabinet.
5. Do the USB ports function when the UPS (inverter) is powered down. It's an edge case, but not always well identified in specs. In the event of a hurricane/wildfire/[insert longer-term-outage] occurs, this can be a lifesaver. This also may have some IoT applications.
To those wondering about adding more sockets . . . For those wondering what to do if you lack enough power ports on the UPS, I install passive mini extension cords. The "squid" style ( https://www.amazon.com/dp/B075ZRB14L) and mini-extension (https://www.amazon.com/dp/B09TC7ZPCJ/) are my go-to's. To address potential criticism of this approach, I always use grounded, passive cords and keep the loads in line with the UPS specs. We work with what we've got!
I don't own one, but UPSes could save the backsides of many in my country, where there are scheduled power cuts all the time, and that we can't always keep track of. Euphemistically called "load shedding," it is sometimes a daily occurrence.
I assume laptops are particularly popular, there? The built-in batteries give you the same effect, if you don't mind not using any external monitors during an outage.
If I had frequent power outages and lived somewhere that I could install solar panels, that'd be where I'd put whatever money I could afford to set aside for reliability. It also scales much better across a large population than charging batteries from wall power, which will only make the outages longer-lasting and more frequent. And, unlike petrol-based generators, you don't have to keep spending money on fuel.
Very! But it's more because of the cost, I would say, since cheap Dell and HP laptops abound. Recently, I stayed with my grandparents for about a fortnight, and was surprised to see that his Core 2 Duo laptop, sporting Windows 7, ran quite all right, and it's about 11 years old.
Here in Cape Town, it's particularly sunny, comparable to southern California, and most newer houses come with solar panels. Unfortunately, most are only assisting the water heater (or geyser as we call it), which draws a lot of power. We've got one of those panels, but the pump packed up recently and we haven't replaced it as yet. I agree it would be great if people's panels could assist with actual power and perhaps send some back to the grid.
How does this UPS interact with a backup generator? This is one of the key things I need to have tested.
I have a home/office set-up and need reliable power to work, so I have a natural gas fired whole-house backup generator. But for many of the UPS systems I've tried, the system goes into the following loop:
Power goes off.
UPS comes on (battery-power mode).
generator comes on
LOOP: UPS detects acceptable input voltage and goes to input-line mode Equipment comes on (3 computers + monitors + router) Power demand causes backup generator voltage to drop temporarily UPS detects low input voltage and switches back to battery-power mode Lowered demand on backup generator allows voltage to rise
I'd say either your generator is under-spec'd or you need some sort of buffer - like a big capacitor that can charge up and absorb the immediate surge - between it and your UPS (if not the house, as a whole).
Some UPS models are less sensitive to voltage dips and have built-in voltage regulators to help shield plugged-in devices from them. You could look into that angle, as well.
The other thing that might help break the loop is to go around the house and make sure everything else is turned off.
As for the output voltage waveform, it would also be interesting to see the waveform of the current that the load is pulling. The type of load wasn't mentioned in the article.
If the load consists of an AC-DC PSU with bridge rectifier and capacitor, I would imagin when fed by a relatively weak inverter it could distort the voltage waveform in the way shown? On the other hand, such a PSU doesn't care at all whether the voltage waveform is sine or square..
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
107 Comments
Back to Article
techjunkie123 - Wednesday, April 13, 2022 - link
UNLIMITED POWAAAAHHHHHsupercaliber - Wednesday, April 13, 2022 - link
Love this category of reviews--keep them coming! (also intrigued about mentioned Li-ion based as I need small form factors)mode_13h - Wednesday, April 13, 2022 - link
With lead-acid batteries, a trick I've done a few times is to use them for jump-starting a car. Just make sure you're really using a ~12 V cell, as some UPS batteries are actually 2 wired in series.ydeer - Wednesday, April 13, 2022 - link
I appreciate the review. Overall efficiency at different load levels and standby power consumption of the device itself would be of interest to me, with power prices developing as they are.ydeer - Wednesday, April 13, 2022 - link
Also, would it be safe to replace the built in battery with a LiFePO4 battery of the same size?Many battery vendors now offer LiFePO4 in lead-acid form factors, advertised as drop-in replacements since they have a built in BMS that makes them compatible with regular chargers.
They are lighter, can't leak and supposedly last far longer (in both time and cycles).
I noticed that they have much lower discharge values than lead acid/AGM batteries of the same size, so I doubt they would be safe to use at full load, though the BMS should cut off there. Perhaps something the author could investigate in a future content piece.
t.s - Wednesday, April 13, 2022 - link
I've modified a few APC UPS to use LFP, and so far still good. BMS is a must for safety. And if one planning to use >50 AH battery for longer usage, remember to put a reliable fan that blow directly to the transformer.A pity that there's almost no LFP battery in market.
DanNeely - Wednesday, April 13, 2022 - link
There're a few UPSes with lithium batteries of some sort (never dug deep enough into the specs to find the exact chemistry). They're bafflingly expensive though.ex APC's 1500VA model is $1800 on amazon.
But if you look at the specsheets for their conventional lead acid ones, the 1500VA model can provide 120W for 1 hour. 120Wh is only about 20% more capacity than a max size laptop battery; so UPSes using them shouldn't need to be crazy expensive. Once you factor in the far lower shipping cost I wouldn't be surprised if the delivered price actually ended up being cheaper.
Kamus - Thursday, April 14, 2022 - link
Yeah, those "bafflingly expensive" UPS companies are going to be driven out of the market soon enough. By the likes of Ecoflow, Bluetti, and other endless "solar generator" brands (anker just got in on the game too)They are selling products that instead of just giving you time to quickly shut down, will allow you to go on for hours of uptime if the power outage goes on for a long time, and they all allow you to charge the "UPS" with solar panels to further extend your uptime for a very long time. (with a big enough system, you can make your computer independent from the grid)
These companies are all using either ternary batteries, or LFP batteries. The ones with LFP batteries are monsters that you can cycle every day, for years without any noticeable degradation. (those would be the ones to use if you want an offgrid computer)
UPS companies are being disrupted and they don't even know it yet, I wouldn't be surprised to see many of these file for bankruptcy in the coming years, as these companies get their UPS capabilities more polished (some don't switch fast enough for critical loads yet) and up their marketing game as "UPS" instead of just "solar generators"
t.s - Thursday, April 14, 2022 - link
True that. For countries that have more day than night, you can go off grid if planned right using solar panel and LFP. My dream is having a DC home. So there's no need to convert DC to AC. Just have to use boost or buck converter to up or down the voltage. Better efficiency and fewer energy loss.dersteffeneilers - Saturday, April 16, 2022 - link
So you are planning to install new outlets? Or just with an armageddon of extension cords?I think it makes more sense to spend the money to have a electrician hook your battery up to the grid, if you live in a place where your output can get adjusted by grid demand automatically, it can pay for itself very quickly.
Vatharian - Saturday, April 16, 2022 - link
Sadly, almost anything that has electric motor inside, requires AC. Where you don't have access to gas line, induction is also a must. Do you really want to run 7-8 kW to induction stove or 2-3 kW to oven with arm-thick cables with DC? The best solution is to have dedicated circuits to both. I can picture a variation of UK type plug (because you DON'T want US-style ones for that... or for anything, really) that could handle both AC and DC from same outlet with complete separation of circuits and safe disconnection even from high current DC.I absolutely see the worth of dedicated DC circuitry, however for it to be safe, the craftsmanship during installation would have to be impeccable, with highest quality materials, and let's face it, depending on the size of the house (or apartment) it would be very expensive.
Chaser - Sunday, April 17, 2022 - link
I don't believe Cyberpower and APC's home consumer marketed UPS products are losing any sleep.Valantar - Friday, April 22, 2022 - link
I don't know what equipment from those brands you're looking at, but most of what I can find, while higher capacity and more feature-packed than this, are an order of magnitude higher in price than this is too. So while high-end UPSes might be in a process of being "disrupted" (god how I hate that nonsensical term), the one under review here certainly isn't.t.s - Thursday, April 14, 2022 - link
Yes, there's lithium UPSes, but as far as I know, it used Li-Ion or Li-Po, not LFP. 500-800 Cycle vs 3.500 cycle. LFP has far more cycle, and safer too.And yes, Those UPSes with Lithium is expensive as hell. Better modify the lead acid ourselves. For $1.800, you can get 24v 200AH LFP + BMS + connector + cable + fan + APC SMT750IC (pure sinewave) and still have changes. 24v 200AH = ~5KWH, translates to days when offline.
bananaforscale - Thursday, April 21, 2022 - link
I wouldn't because LiFePO4 batteries differ dramatically from lead-acid batteries in charge requirements (current curve, voltage and such; for one, lead-acid is charged typically at 14.4V which is much too high for LiFePO4). Unless of course you plan to replace the charge circuit at which point you might just as well build it from scratch.asmian - Wednesday, April 13, 2022 - link
"nagual UPS review" -> "inaugural", perhaps...Since the writer makes the statement that UPS is making a comeback, and makes some fairly strong technical comments about the quality of the components and circuitry, it would be very interesting to see a review with a mix of new and better-known brands in their current incarnations pitted against each other, to indicate the current state of the market.
Ryan Smith - Wednesday, April 13, 2022 - link
Thanks!mode_13h - Wednesday, April 13, 2022 - link
Another suggestion: my current UPS (APC Smart-UPS C 1500) had pretty bad coil whine, for the first few years I owned it. This happened when it was on normal A/C power, and I reproduced it both at home and where I work. I even returned it under warranty, but the replacement did the same thing.So, I'd highlight this as an area to watch out for. In a quiet room, that can be a real annoyance.
E.Fyll - Wednesday, April 13, 2022 - link
Thanks for the feedback. I will surely point out such issues if they occur.The most common cause of coil whine is poor quality power input. Harmonics are to blame 90% of the time. That clearly is not the case here though, since you had the same problem at two entirely different places. Someone probably just did not glue down an inductor that should have been glued down.
Mikewind Dale - Wednesday, April 13, 2022 - link
I own two Cyberpower CP1500PFCLCD PFC Sinewave UPS (1500VA, 1000W), one my Ryzen 7 2700X and the other for my ThreadRipper Pro 3955WX. And a friend of mine - on my recommendation - bought the same for his ThreadRipper Pro 3955WX.So far, I have had no issues. I don't get many power outages, but when I do, they have always worked.
So that's one more to test.
Thunder 57 - Friday, April 15, 2022 - link
Same. Well, almost. Mine is a slightly older model with the only real changes being 2 x USB type A and a different LCD. It's worked great for me. 2 x 8.5AH batteries too not the usual 7AH. I'd replace them with 9AH ones but it seems like too much work for too little result.Definitely like the UPS review(s) though, keep them coming!
mode_13h - Wednesday, April 13, 2022 - link
Also, I wonder how feasible it'd be to test or at least teardown some Ethernet surge protectors. Right now, I've got my cable modem on its own small UPS, then a gig-E surge protector separates it from my router. If that surge protector turns out to be garbage, then I might as well just run everything on a single UPS.FYI, the one I'm using is Tupavco TP302, purchased on Amazon.
Einy0 - Wednesday, April 13, 2022 - link
"The UPS only outputs at a nominal 230V, and similarly, is only designed to accept voltages around that range (sorry, Americans!)."I'm going to pretend you mentioned the frequency difference as a reason for Americans not to use this or the outlet form factors.
Eletriarnation - Wednesday, April 13, 2022 - link
Unless there's a model with North American standard outlets, I don't think it will be very popular there anyway.Eletriarnation - Wednesday, April 13, 2022 - link
Ah sorry, I see you mentioned the outlets. Entirely correct though, the 230V in/out isn't necessarily an issue as much as the other things.DanNeely - Wednesday, April 13, 2022 - link
Probably not, but most modern electronics have 100-240V PSUs; so all someone wanting to use it in the US would need is to wire a 240V outlet and then import some device powercords along with the UPS. So we could use it if we really wanted to.Calin - Thursday, April 14, 2022 - link
Most of the UPS units I saw (that was a long time ago though) had detachable power cords of the "PC" style. If this is the case, powering up the UPS is no issue at all, as the PC power cords in local style are easy to find. As for output, there are "universal adaptors" (but they might not support higher amperage)mode_13h - Thursday, April 14, 2022 - link
I don't recall ever seeing a UPS with a detachable power cord, in the USA. Most I've seen are APC, but I've seen a couple Cyberpower and other brands.at_clucks - Friday, April 15, 2022 - link
*A lot* of UPSes have the standard IEC 60320 C13/C14 connectors https://en.wikipedia.org/wiki/IEC_60320#C13/C14_co... I don't think I've seen any with a hardwired power cable (Europe, mind you). The more home oriented ones used a C14 connector for mains power and regular power outlets (instead of C13) for connecting the devices since most people want to connect a whole host of devices not just whatever you can connect with a "UPS/PC cable" https://notebookcentre.am/wa-data/public/shop/prod...I used to just replace the cable of a multi outlet extension cord so it has a C14 plug at the other end and plug that into the UPS so things with a power brick would be easy to plug in.
DanNeely - Friday, April 15, 2022 - link
Looking on Amazon, all the UPSes I see - including a few rackmount models - have hard wired power cords. I wonder if US regulations treats them as overgrown powerstrips (which I've also never seen with a detachable input cable).The EU model reviewed here also has a hard wired cable for input power.
PaulHoule - Wednesday, April 13, 2022 - link
Glad to see a UPS from a vendor that isn't Schneider Electric.Still I think there is very little point in a UPS with a 5-10 minute runtime on a desktop or server. I mean, you get 5 or 10 minutes of terror to save your files, whereas it is routine for a machine to come back after a power loss and not even have to fsck. If you skip the UPS, you might lose 30 minutes work of unsaved work in Word and Word will never stop pestering you about the files you didn't save, even 10 years from now, but you didn't pay $250, you passed go, you didn't have to deal with noise, acid fumes, disposing of lead-acid batteries, fire hazard, etc.
A more realistic mission for a UPS in 2022 is to support a small amount of equipment. Back when we had DSL from the central office we had a UPS set up to power one DSL modem and a cordless phone with an answering machine. This gave us 21st century comms that we could tap with a laptop or a tablet for a whole weekend.
They upgraded our DSL for a 10x performance boost but now it is dependent on power at the crossroads near us, probably beta(10,2) the power goes out at the crossroads if it goes out at our farm, so the UPS powered a DSL modem that didn't do anything so we donated the UPS to a reuse center and crossed our fingers we wouldn't get charged for "donating" e-waste.
DanNeely - Wednesday, April 13, 2022 - link
Unfortunately keeping a few PCs running just long enough to shut down safely seems to still be the design target of all UPSes.A half dozenish years ago I tested a 1300VA APC and 1500 VA Cyberpower UPS with new batteries. They self-discharged in 5-6 hours with no external load; with my cable modem and router attached they lasted between 4 and 5 hours.
That implies that a design made to be efficient at low load could keep my wired internet running as long as my phone/etc batteries last. If an efficient at low load design existed.
Slash3 - Wednesday, April 13, 2022 - link
I have my router, gateway and a ye olden land line phone hooked up to an old APC Smart-UPS 620VA - it's enough to keep the internet going for a solid six hours when the power goes out, which it does fairly often (weather). I end up changing out the batteries every four years or so, but it's well worth the convenience to be able to check for power company updates, keep in touch or otherwise continue to while away the time on a tablet or phone.My main desktop has its own APC Back-UPS 1500VA, enough for about 30-45 minutes of normal desktop usage, so often the power will be restored before I need to shut down (or set it to hibernate). If it's particularly nasty weather, I'll shut down the desktop ASAP and switch the UPS off, in case I need to plug something else in later (LED lights, phone/laptop charger, etc).
I think for most people, the reality is that 4-5 minutes to be able to immediately save their work and shut down gracefully is pretty realistic. Just having that option is a bit of a psychological safety net.
Threska - Wednesday, April 13, 2022 - link
That's one reason to having one of the smaller ones because UPSs can be HEAVY.mode_13h - Wednesday, April 13, 2022 - link
> keeping a few PCs running just long enough to shut down safely> seems to still be the design target of all UPSes.
No, there are certainly more capacious UPS models, if you want more uptime badly enough.
mode_13h - Wednesday, April 13, 2022 - link
I forgot to add that you might do well to investigate medical-grade UPS models.Also, I wonder if true sinewave output places a higher residual load on these units.
mode_13h - Wednesday, April 13, 2022 - link
> I think there is very little point in a UPS with a 5-10 minute runtime on a desktop or server.I have occasional brown outs and outages lasting only a few seconds. A good UPS has saved my bacon so many times, it's not even funny.
I've even done a cute thing where I put my desktop in "sleep", powered off everything else, then called the utility company about the problem, and they managed to fix it before my 1500 VA UPS ran out of juice.
> it is routine for a machine to come back after a power loss and not even have to fsck.
For servers, you're supposed to have it automatically trigger shutdown, to avoid filesystem corruption. You can "play with fire" and just let them lose power, but even if storage setup has no "write hole" and is correctly using write barriers, there's always the potential for application-level data corruption.
> If you skip the UPS, you might lose 30 minutes work of unsaved work in Word
If you want to take that risk, then so be it. However, not all of us are simply editing Word files or using MS Windows desktops.
> now it is dependent on power at the crossroads near us
FWIW, I've never lost my cable internet when I've lost power. I'm sure if the outage were sufficiently widespread and long-lasting, it'd be a different story.
Threska - Wednesday, April 13, 2022 - link
Depends upon ISP. Power's out, internet is out.1_rick - Monday, April 18, 2022 - link
"I have occasional brown outs and outages lasting only a few seconds. A good UPS has saved my bacon so many times, it's not even funny."Exactly. Where I live, the power blinks a few times a year. About a year ago, it happened a dozen times in a short period, including a double flicker that caused Windows to trash my user profile, losing all my stored passwords and stuff. I got a UPS after that, put the computer and one monitor on it, and now I can weather any number of these few-second outages, and there's plenty of time to shut the machine down if the outage lasts longer than that.
t.s - Wednesday, April 13, 2022 - link
If you don't want lead-acid, still want to use something akin to UPS and if the office mainly using DC tools, you could buy 12v/ 24v LFP. Charge it using SMPS PSU (AC to DC) with passthrough to the tools with buck (step down) or boost (step up) converter. No noise, no acid fumes, small probability to caught fire. And a good LFP battery have ~3.500 cycle. Quite simple to built, but have to be careful.For my office, I'm using 54W (depends on loads) 14v adaptor buck to 13.34v to charge my 100 AH LFP, powering 24" LG monitor, a modem, a router, and a 15w laptop (step up to 19v converter). Been ~ 2 years now, and no problem so far.
sor - Wednesday, April 13, 2022 - link
I'm probably in the minority, but I have an 8 hour whole house battery backup. The UPS is here at my desk to gap the moments it takes for that to kick in.That said, I've got two computers, my network gear, and a 38" monitor plugged into a small-ish Back-UPS Pro 1000S and the load currently says 72W with an estimated backup time of 53 minutes.
Also, the majority of power outages in my area (maybe once a year) last less than a minute.
Mikewind Dale - Wednesday, April 13, 2022 - link
A friend of mine has a gas generator that takes a few seconds to kick in, so I told him to buy a Cyberpower CP1500PFCLCD PFC Sinewave UPS for his ThreadRipper Pro 3955WX. And it works.t.s - Thursday, April 14, 2022 - link
Very nice electricity there. Here, at least in a month (average), there's 1 time power outages and usually goes for 2 hours.Thunder 57 - Friday, April 15, 2022 - link
Not sure where you are from but I took a battery just the other day to Home Depot. There was a bin just for batteries you could toss them in. There are other places that do that too. And yes, a UPS on a modem/router is very nice to have. Keeps your internet up and you can use a tablet/laptop for plenty of time.Lezmaka - Wednesday, April 13, 2022 - link
I'm just one person (physically, mentally... that's still undetermined) but I liked this. With more people (like me) working from home hopefully there's enough interest to see more of these. I'd personally like a little more educational commentary, either as part of reviews or guides, but hope to see more of these.jamesindevon - Wednesday, April 13, 2022 - link
Great review - thanks. It could have done with a mention of the USB socket around the back -- it should be there so attached PCs can power down safely when the unit goes to battery power. A few words on this and any provided software would have been good (can one PC send a shutdown signal to another?)Also, you might want to mention that European grid frequencies vary a lot more than this unit.
romrunning - Wednesday, April 13, 2022 - link
I am interested in seeing some APC reviews, especially at 1k+. APC (Scheider Electric) still seems to be the market leader in North America, so would love to see their products reviewed. My favorite does seem to be the 1500VA spot, so I'm curious to see whether the savings on the cheaper CyberPower is worth it.tyger11 - Wednesday, April 13, 2022 - link
I went with CyberPower units after an APC one died on me, and I've never looked back,domih - Wednesday, April 13, 2022 - link
Same here. I've a few APC ones which I bought when CompUSA (brick and mortar) died and was selling them at a quarter of the price by the last week. Otherwise I have many more CyberPower. No problems with both brands. Both brands did their job several several time during power outages. Meaning: the power went down for a few seconds and the computers did not mind. Or the power was out for hours and I had time to do normal shutdowns (automatic or manual). Contrary to APC, the CyberPower Windows Server software app was free. I don't really care anymore because I switched to Linux years ago. I also got several additional APC and CyberPower for free from a company that did not want them anymore and was on the verge to e-waste them. Like my first generation of UPSes I just replace the batteries and they were like new. All the models I have from both brands use the same battery anyway: any equivalent to DURACELL DURA12-8F2 12v 8Ah AGM. I went to a local brick and mortar battery retailer. He smiled and gave me a good discount when I ask the price for 24 units :-)mode_13h - Wednesday, April 13, 2022 - link
I had a Cyber Power CP685AVRLCD fail pretty badly on me, after about 5-6 years of use. I seem to recall it emitted a strange noise and started emitting a funny odor, not prompted by any sort of power event that I could discern. I can't remember if it also cut power to the outlets, but I think so.We've had a couple APC units fail, at work. The most common thing I see is that a unit appears completely bricked, but the issue is that its battery is simply too dead. As a result, it gives no signs of life when plugged in. After installing a fresh battery, they seem to return to fully working status. However, I'm not counting that as "failed" - I just mention it as a caution to anyone who thinks they need to dump an old UPS that *seems* dead, after not being plugged in for a long time.
Keep your UPS's plugged into A/C, even if you're not using them! Lead acid batteries DO NOT like deep discharges!
Anyway, we have some APC rack mount units are probably 20 years old and still kicking. They're not on anything mission-critical, so as long as they don't catch fire or something, there seems to be no need to replace them.
casteve - Wednesday, April 13, 2022 - link
Thanks for the review. A couple of features I look for:- is it passive or fanned/what's the fan noise level? (thanks for that)
- can alerts be silenced? High electronic beeping really freaks out my dog.
- when it's running on batteries or in AVR mode is the output ugly enough to cause a PC's PSU to make ugly noises?
- Is there a master/slave mode where slave outputs turn on/off when the UPS detects load/no-load conditions on the master? A really useful function on older APC units. Put the PC on the master and the monitor and audio amp on the slave.
mode_13h - Wednesday, April 13, 2022 - link
> can alerts be silenced? High electronic beeping really freaks out my dog.Although there's frequently a front panel button to disable the beeps, you'll often find that even models without this will allow you to disable the beep via software, after connecting to its USB or serial port. You can also get some useful and interesting information that way, as well.
That reminds me of a funny story about an older model of APC Smart-UPS. It has a 9-pin serial port, on the back. So, I install the APC Power Chute software and connect it to my PC. As soon as I complete the cable connection, the USP goes *dead*. I unhook it, and it turns back on (I forget if I also had to hit the power button... probably I did). Then, I check the docs and see that the friggin' thing turns out to have a nonstandard pin-out! WTH?? Now I get why their newer models with serial ports don't use a standard-looking 9-pin connector, but that's just *bad*.
tl;dr - Make sure to use the included cable, if you have an APC model with the 9-pin connector on the back!
mode_13h - Wednesday, April 13, 2022 - link
> Is there a masterI discovered this accidentally, when I turned off a co-worker's monitor as I was leaving the office, one Friday afternoon. I thought I heard something else turn off, but I thought nothing of it. After the next time, he figured out what was happening and changed which device was plugged into the master outlet.
I think I've seen some units with a button to disable this behavior. IMO, it should be disabled by default. I'd even previously *seen* the "Master" label, but had no idea what it meant.
watersb - Wednesday, April 13, 2022 - link
I am really pleased to see a UPS review the same level of expertise as the remarkable power supply articles. Excellent!espressomakerdeluxe - Wednesday, April 13, 2022 - link
Brilliant idea! Please do rack mount units. Pretty please.croc - Thursday, April 14, 2022 - link
Excellent article, expertly written by a knowledgeable author I hope that future articles will include graphs / pictures with all reviewed models included. Such a database might become dated, but that also allows us to see how far we've come.One thing I did notice was the clipping of the output waveform of this particular model under load. Without comparisons, it is difficult for the average person to really understand what that means for his equipment, so feel free to continue to comment in an unbiased manner. Hope that AnandTech is paying you enough that you can avoid 'sponsership'.
Kamus - Thursday, April 14, 2022 - link
All of these companies are going to be driven out of business by the "solar generator" market soon.Those guys are making very cheap LFP battery powered, pure sine wave inverters at very competitive prices. (at least compared to an UPS)
The only thing holding back those "solar generators" is that a lot of them don't switch fast enough yet, but honestly, even with the current 15ms switching times I would take them over a lead acid UPS.
philehidiot - Thursday, April 14, 2022 - link
Sounds like this might do the job for me. I want it for powering a couple of guitar amps strapped to a motorcycle to play offensive music at people who have glued themselves to roads in order to "protest" insulation. They never do it in winter, only when it's warm, which is superb for battery life and riding.Tomatotech - Saturday, April 23, 2022 - link
Excellent idea. I'm deaf, if you find me glued to a road please play your music nice and loud so that I can enjoy it properly.Alternately please ask your MP to ask the Govt to follow the recommendations of its own expert advisors and spend some money on insulation. If it had done this when the advice first came out, the UK would be in a much better position with regard to being able to reject buying Russian oil and gas. As it is, your taxes are going towards giving billions to *both* sides of a bloody and genocidical war. I think that's worth protesting about.
DigitalFreak - Thursday, April 14, 2022 - link
Question is, are these reviews going to be frequent or only one or two per year?Ryan Smith - Wednesday, April 20, 2022 - link
It depends on reader interest, and how eager manufacturers are to sample. There's no reason we can't do 4-5 a year if the supply and demand are there.Oxford Guy - Thursday, April 14, 2022 - link
I read quickly but saw nothing about the fact that lead-acid batteries are permanently degraded by being discharged. The article mentioned a 12-hour recharge time for when the batteries have been completely depleted but didn’t mention the far more pertinent matter of lead sulphate. That is the chemical that forms when lead acid batteries are discharged, even partially. It is a byproduct of the degradation inherent in the tech.Lead-acid batteries have no place in a device like this. People who wish to debate that should truly look at the big picture. This is not a device like a car which has a workload defined by the battery usually being mostly fully-charged (the one workload that is suitable for lead-acid tech). The article said the batteries in this unit can be depleted in a matter of minutes. This device is designed around much heavier depletion than a car, making it fundamentally unsuited to the battery tech.
Why do manufacturers choose to use lead-acid batteries for devices with the wrong workload? They’re counting on buyers looking at things like sine wave output and wattage — the same thing this review does.
Perhaps the business model here is to sell replacement batteries to make up for the reduced pricing of its partial sine wave ability.
The irony of consumer-grade UPS devices has been mainly about how prone they are to dying. A computer surviving multiple lightning storms with a simple surge blocker is more likely than these UPS devices lasting long enough.
A useful review will not only warn about the unsuitability of lead-acid batteries, it will be in two parts — the second being a long-term lifespan report. See if you can get a few years out of these products. See how much uptime lead-acid batteries manage after multiple discharges.
I can see using lead-acid batteries in a device that has an optional ‘emergency mode’ where the user can decide to engage them when the lithium batteries are nearing depletion. This could be a cost-saving way to give the unit longer uptime at a lower price point — suitable only for workloads that are unlikely, but not completely, to need the extra runtime.
Oxford Guy - Thursday, April 14, 2022 - link
It’s also bait and switch fraud to claim sine wave output for a device that does not produce sine waves over most of its output range.That should not be hand-waved with excuses about what most people might use the device for. The product is being sold on a lie.
back2future - Friday, April 15, 2022 - link
what needs regulation for advertisement and technical description for standardized consumer support and information (and consumers understanding their usage profile, planed/scheduled repairs/upgrades, replacement parts available and costs including wear/maintenance/refirmation, before purchase) (?)mode_13h - Thursday, April 14, 2022 - link
> Lead-acid batteries have no place in a device like this.The typical usage model is that the user saves their work, shuts down their PC, and then switches off the UPS. You're not necessarily supposed to run until the battery is 100% depleted.
> A computer surviving multiple lightning storms with a simple surge blocker
Surge suppression won't save you from lost work and possible file or filesystem corruption due to abrupt reboots.
> the second being a long-term lifespan report.
Here's my lifespan report. All of my co-workers have APC UPS units at their desk. Sometimes, we lose power when people aren't around to switch them off, and most haven't gone to the trouble of configuring auto-shutdown. Yet, we still typically get about 5 years out of these batteries (possibly more). At home, I've gotten over 7 years out of mine, which I've run down probably at least 3 times.
Yes, Lithium Ion would be great. However, as noted in other comments, there's a cost differential. And Lithium Ion have other tradeoffs.
You make some fair points, especially with your reminder about minimizing depletion of lead acid batteries. I definitely want to acknowledge that.
DanNeely - Friday, April 15, 2022 - link
I generally get 4 years/battery pack at home; OTOH my high power desktops (running distributed computing 24/7) mean I need a larger than normal chunk of the nameplate capacity to avoid an immediate shutdown on power loss.And with maybe 2-6 interruptions longer than a few seconds/year decay due to discharge cycles is basically irrelevant. (I find the person upthread ranting about the use of lithium batteries only good for several hundred rather than several thousand cycles upthread hilariously divorced from how typical UPSes are used.)
Oxford Guy - Friday, April 15, 2022 - link
Mr. Neely:Lead-acid batteries die partially whenever they are depleted, even partially. That is a the opposite of irrelevance unless the device is never used.
Being happy with a four-year lifespan is also quite debatable.
Oxford Guy - Friday, April 15, 2022 - link
Defining that lifespan is also fraught since it varies wildly by the circumstances and how each person defines the lifespan. One may claim the product is still working when the battery can only manage 15% of its original charge. Another will wait until the product simply doesn’t power up or whatever.mode_13h - Friday, April 15, 2022 - link
I define "lifespan" as "when the UPS tells me to replace the battery". I have no idea what that actually means, in terms of raw capacity, because my experience is that a UPS in that state won't even *try* to give you any runtime.I would advise people not to buy UPS batteries in advance, because they have some degree of self-discharge. And, as we've discussed, you don't want to them to have any significant amount of discharge.
For the same reason, I advise people to keep every UPS plugged-in (at least some of the time), even if they're not using it.
Oxford Guy - Saturday, April 16, 2022 - link
The numerous units we (myself and colleagues) have purchased have failed. It hasn’t merely been the batteries. That includes models claiming sine wave output. I am thoroughly unimpressed with what I have seen so far in terms of ‘affordable’ UPSs.Oxford Guy - Friday, April 15, 2022 - link
‘The typical usage model is that the user saves their work, shuts down their PC, and then switches off the UPS. You're not necessarily supposed to run until the battery is 100% depleted.’This device can be depleted within minutes. How quickly is someone going to prevent the batteries from draining at all, let alone completely? People are not always at their PC. People take bathroom breaks, get snacks, et cetera.
Lead-acid tech is fundamentally incompatible with this usage model. It is that simple. The only usage model that is at all justifiable is having lead-acid batteries as an additional optional emergency capacity mode, to supplement a typically-adequate (capacity) lithium primary battery.
There is no justification for polluting the planet with batteries that are intentionally lifespan-shortened. The biosphere is in bad shape already, without being intentionally more reckless.
Even for those who don’t have a conscience and thus care not about ecological product justification, they need to grant that basic logic shows there is a severe mismatch between the workload and the solution chosen.
mode_13h - Friday, April 15, 2022 - link
> This device can be depleted within minutes.Runtime is determined by capacity / load. That it *can be* doesn't mean it *always is*. Even if you buy the minimum capacity needed for your load, that would be *peak* load. As we all know, modern computers can burn a lot of power at full load, but not much at idle. And, unless you're gaming or doing other heavy-duty computation, PCs are typically running closer to idle than peak.
> People are not always at their PC. People take bathroom breaks, get snacks, et cetera.
Yes. As I mentioned, my workplace typically gets about 5 years out of a battery, even though there's probably an average of 1 full discharge per year.
> Lead-acid tech is fundamentally incompatible with this usage model.
It's worked well enough for 3-4 decades. True, it's not ideal, but the batteries aren't a major cost item for the amount of time they last. I mean, relative to the cost of what's plugged into the UPS and its rate of depreciation.
> There is no justification for polluting the planet with batteries
And here's where I figured we'd end up. Like all things, it's a matter of cost/benefit. People should dispose of their batteries properly, but of course that process isn't perfect and not everyone is going to do it. So, I agree that lead acid is a dirty technology, but Lithium Ion also has environmental impacts. As and when better technologies become competitive, I'd like to get away from the tremendously heavy & dirty lead acid tech.
> basic logic shows there is a severe mismatch between the workload and the solution chosen.
No, it's not severe. You're overstating your case, which distracts from your core message.
Oxford Guy - Saturday, April 16, 2022 - link
‘You're overstating your case, which distracts from your core message.’That is your opinion.
bunkle - Tuesday, April 19, 2022 - link
Seems a lot of people think a UPS is equivalent to a power wall and don't really understand the designed use cases. They're for short term use so you can switch from one power source to another, prevent brownouts or enable automated clean shutdowns. I run a 3K VA UPS at hone that's attached to the mains and an LPG/Petrol generator. You'd be surprised how often brownouts occur, even last night it protected my machines:Date : 04/19/2022, Time : 02:36:12, Code : 0x0109
Warning - UPS: On battery power in response to distorted input.
Date : 04/19/2022, Time : 02:36:13, Code : 0x010A
Informational - UPS: No longer on battery power.
Given the fact that lead acid batteries are the top recycled consumer product in the US with 99% recycling rates, one could argue that lead acid is far cleaner than any Lithium based alternative. The batteries are also standardised and you don't get locked into using a vendor's specific replacements.
Oxford Guy - Thursday, April 21, 2022 - link
‘Seems a lot of people think a UPS is equivalent to a power wall and don't really understand the designed use cases.’That explains the use of lead batteries when people can need to do things like use a restroom or answer the door. Lead batteries are permanently degraded within minutes with these devices.
mode_13h - Friday, April 22, 2022 - link
> Lead batteries are permanently degraded within minutes with these devices.Your characterizations simply don't match my experiences. Your warnings sound very dire, but the 5-7 years I've gotten out of these batteries would argue otherwise.
I've long used AGM lead-acid batteries in my cars. I've even treated them somewhat roughly, especially during the pandemic, and yet still averaged about 5 years out of them.
If you want to keep pressing this case, then please provide some data on the specific battery tech used in these UPS units.
Oxford Guy - Thursday, May 12, 2022 - link
You don’t understand that lead acid batteries don’t work like lithium batteries. They are degraded when they are even partially discharged. The only workload they’re suitable for is one where the battery is only slightly discharged, like with starting a car.kpp - Thursday, April 14, 2022 - link
Very nice and technical review, as always from mr Fylladitakis. The only thing I believe is missing is a measure of the UPS power consumption while on standby with the batteries charged. In my experience, this consumption can range from 10W (very good "green" UPS) up to 50W (typical APC UPS). This consumption can contribute significantly to the Total Cost of Ownership of the UPS.mode_13h - Thursday, April 14, 2022 - link
> up to 50W (typical APC UPS).Not unless it's charging. Common sense should tell you that if it's burning 50 W, its fan must run to keep it from overheating. The only time mine runs is while charging or discharging. And I have a 1500 VA model, which is at the upper end of capacity for a desk-side UPS and even has a backlit front panel LCD. Also, it's a model with Sine Wave output.
APC publishes Load/Efficiency specs, on their website. For my SMC1500, they claim:
Load | Efficiency
--------------------
25% | 96.1%
50% | 97.6%
75% | 97.9%
100% | 98.0%
mode_13h - Thursday, April 14, 2022 - link
Translating that into Watts, I think it'd be:Load | Overhead
---------------------
225 W | 9 W
450 W | 11 W
675 W | 14 W
900 W | 18 W
kpp - Wednesday, April 20, 2022 - link
I had this 50w while on standby figure on 2 Mustek 2000VA that I used to own. I had a powermeter in the wall socket and was amazed when I noticed this. This was constant, not while charging.ranran - Friday, April 15, 2022 - link
I think this is a great idea/category, but I'm surprised you chose a European-market focused brand that is not really available here in the West.Perhaps you could consider doing reviews on the top two UPS companies, APC and Cyberpower? Expand this review to include comparisons with Cyberpower's CP1500PFCLCD and APC's BR1500MS true sinewave systems?
DanNeely - Friday, April 15, 2022 - link
If you have any contacts with those companies that could help convince them to send a sample to Greece I'm certain Ryan Smith and E. Fylladitakis would be grateful for your assistance. 🙄https://twitter.com/RyanSmithAT/status/15142159668...
nikitas589 - Friday, April 15, 2022 - link
Nice review. Yes, UPS reviews are definitely useful. And the noise test/commentary Ishould definitely be include for any UPS that may find its way to a home office /setting. One thing toe consider perhaps, speaking from experience, is the not all surge/power cut events register trigger a UPS (standby or interactive) to take over. For example I had a power cut trigger it while a milliseconds "on/off" type of thing did not and the PCs simply lost power. So you may want devise a couple of tests that are not instrument based or linear but more brutal ( pull the plug momentarily /tripping over the power cable for example, switch power on off, vary the timings, etc ) that are closer to what may happen to some of us. Still, UPS reviews: nice keep them coming..:-)mode_13h - Tuesday, April 19, 2022 - link
> I had a power cut trigger it while a milliseconds "on/off" type of thing did not and the PCs simply lost power.I've never seen a UPS fail to intervene in a power event that resulted in a PC rebooting. I'm not saying it doesn't happen, but not with the APC UPS units that I've used at home and at work.
swizznowt - Friday, April 15, 2022 - link
0.6 power factor, line interactive, no isolation, no common mode/normal mode protection. It's an APC by another name. Junk!Samus - Monday, April 18, 2022 - link
Could you perhaps determine what the trickle\standby voltage of the battery charging circuit is? I've seen a few models of APC UPS's lately that seem to overcharge the batteries to 12.8-13.1v, dramatically shortening their lifespan. One set of batteries from an APC XS1000 were so swollen they had to be pryed out and the unit\batteries were only 3 years old!Samus - Monday, April 18, 2022 - link
Specifically, could you test the measured float voltage of the batteries in UPS reviews?NeatOman - Monday, April 18, 2022 - link
For funzies YEARS ago when li-ion battery banks first started to become available for about $250, i found one with a pure sine wave and used it as a battery for a cart in a warehouse with 20" monitor, laptop, and label printer.That worked fantastic, then i got another one for a business with super flaky power that went in and out and at times read 90v from the wall, and computers and printers would have issues working. This time not only did it work as a UPS but also a line conditioner, you see these on servers all the time (there not just power strips sometimes). Basically the wall would charge/power whatever it could and the inverter would always be the one sending the power as a true "uninterruptable" power supply. It would correct, say 90v or 130v to a very stable 120v at the cost of loosing a bit of efficiency and also act as a far better surge protector. Surge suppressor should be your minimum when protecting your equipment, it ensures you don't get over voltage or "dirty" power. Anywho, a li-ion powered ups should be the norm for spending $200. Not only does it have more power, also smaller, charges much faster (recovery times between outages). Specifically LiFePO4 (lithium iron) batteries would be best, about 2-3 times the duty cycle of li-ion and higher amperage (can be smaller/less capacity and still have enough power for devices).
Oxford Guy - Thursday, April 21, 2022 - link
‘should be the norm’As should product truthfulness, rather than claiming sine wave output when that begins to fail beyond 25% load. Sadly, humanity is full of falsehoods for ‘profit’.
mode_13h - Friday, April 22, 2022 - link
> Sadly, humanity is full of falsehoods for ‘profit’.Why is everything bad somehow an indictment of capitalism, yet you never credit it with anything good that's achieved or made accessible to consumers?
Oxford Guy - Thursday, May 12, 2022 - link
Too bad.mode_13h - Wednesday, July 13, 2022 - link
You have yet to give one modern example representing a viable alternative. Every modern attempt at true socialism thus far has failed. On the other hand, there are examples of regulated capitalism that continue to work well to this day.RU482 - Monday, April 18, 2022 - link
tangential topic - generators. It seems inverter generators are raved about as being superior, but no one ever seems to mention whether they have a true sine wave output or a modified sine wave outputmyself248 - Wednesday, April 20, 2022 - link
> the circuitry layout is fairly outdated.A flippant statement like that implies a fairly substantial amount of reverse-engineering that must've gone into such a conclusion! And it would be more persuasive if the intermediate steps were shared -- how was the reverse-engineering performed? What topology does the unit use? What's outdated about it? How does it compare to others on the market?
I think the UPS market very much needs technically-competent reviews, and this is an encouraging step in that direction. But it could stand to add some rigor and detail. Thank you for the promising start!
Leone018 - Wednesday, April 20, 2022 - link
Is equal to Tripplite ECO1500LCDnikitas589 - Thursday, April 21, 2022 - link
Nice work - an nice new section - keep them UPSs reviews coming.Speaking form experience, line-interactive UPSs may sometimes not respond consistently to power interruptions. I've had one (MGE Eclipse) that for example you would pull the plug and it would work as expected. But for minor power interrupts it wouldn't... so maybe its a nice idea to devise a few power-off scenarios just to make sure the UPS works as expected in a range of scenarios. It wasn't clear tom me in the review if already do so.
nikitas589 - Thursday, April 21, 2022 - link
Oh and the noise level report is 'critical' for home users at least. :-) The quietest I've found on spec is an NJOY's Echo Pro (Online UPS).JFish222 - Thursday, April 21, 2022 - link
Another +1 for these types of reviews, and additional suggestions:1. What is the UPS power draw from the wall when not charging/under load (as mentioned by others).
2. What status monitoring / remote shutdown capabilities are present.
In addition to running the local "Powerchute Personal Edition" type software connected via USB, are there remote shut down agents when connecting multiple devices? Are there fleet management capabilities if I have one under every desk, etc.
3. Low voltage protection (also alluded to, was another interesting question). IE: In the event of an outage, does the battery run to 0% (not grate for Lead Acid), or is there an x% cut-off to protect the battery
4. Parasitic draw when the unit is powered off.
Can I keep the unit offline (ie: not in production) but with batteries internally connected without damaging the batteries? There is a natural discharge rate for lead-acid, but if there is parasitic draw even when off, those batteries could be toast when I pull the unit from the cabinet.
5. Do the USB ports function when the UPS (inverter) is powered down.
It's an edge case, but not always well identified in specs. In the event of a hurricane/wildfire/[insert longer-term-outage] occurs, this can be a lifesaver. This also may have some IoT applications.
To those wondering about adding more sockets . . .
For those wondering what to do if you lack enough power ports on the UPS, I install passive mini extension cords.
The "squid" style ( https://www.amazon.com/dp/B075ZRB14L) and mini-extension (https://www.amazon.com/dp/B09TC7ZPCJ/) are my go-to's. To address potential criticism of this approach, I always use grounded, passive cords and keep the loads in line with the UPS specs. We work with what we've got!
GeoffreyA - Sunday, April 24, 2022 - link
I don't own one, but UPSes could save the backsides of many in my country, where there are scheduled power cuts all the time, and that we can't always keep track of. Euphemistically called "load shedding," it is sometimes a daily occurrence.mode_13h - Tuesday, April 26, 2022 - link
I assume laptops are particularly popular, there? The built-in batteries give you the same effect, if you don't mind not using any external monitors during an outage.If I had frequent power outages and lived somewhere that I could install solar panels, that'd be where I'd put whatever money I could afford to set aside for reliability. It also scales much better across a large population than charging batteries from wall power, which will only make the outages longer-lasting and more frequent. And, unlike petrol-based generators, you don't have to keep spending money on fuel.
GeoffreyA - Wednesday, April 27, 2022 - link
Very! But it's more because of the cost, I would say, since cheap Dell and HP laptops abound. Recently, I stayed with my grandparents for about a fortnight, and was surprised to see that his Core 2 Duo laptop, sporting Windows 7, ran quite all right, and it's about 11 years old.Here in Cape Town, it's particularly sunny, comparable to southern California, and most newer houses come with solar panels. Unfortunately, most are only assisting the water heater (or geyser as we call it), which draws a lot of power. We've got one of those panels, but the pump packed up recently and we haven't replaced it as yet. I agree it would be great if people's panels could assist with actual power and perhaps send some back to the grid.
mode_13h - Thursday, April 28, 2022 - link
> water heater (or geyser as we call it)Interesting!
cjcoats - Friday, May 13, 2022 - link
How does this UPS interact with a backup generator? This is one of the key things I need to have tested.I have a home/office set-up and need reliable power to work, so I have a natural gas fired whole-house backup generator. But for many of the UPS systems I've tried, the system goes into the following loop:
Power goes off.
UPS comes on (battery-power mode).
generator comes on
LOOP:
UPS detects acceptable input voltage and goes to input-line mode
Equipment comes on (3 computers + monitors + router)
Power demand causes backup generator voltage to drop temporarily
UPS detects low input voltage and switches back to battery-power mode
Lowered demand on backup generator allows voltage to rise
mode_13h - Wednesday, July 13, 2022 - link
I'd say either your generator is under-spec'd or you need some sort of buffer - like a big capacitor that can charge up and absorb the immediate surge - between it and your UPS (if not the house, as a whole).Some UPS models are less sensitive to voltage dips and have built-in voltage regulators to help shield plugged-in devices from them. You could look into that angle, as well.
The other thing that might help break the loop is to go around the house and make sure everything else is turned off.
shadowjk - Tuesday, May 31, 2022 - link
As for the output voltage waveform, it would also be interesting to see the waveform of the current that the load is pulling. The type of load wasn't mentioned in the article.If the load consists of an AC-DC PSU with bridge rectifier and capacitor, I would imagin when fed by a relatively weak inverter it could distort the voltage waveform in the way shown? On the other hand, such a PSU doesn't care at all whether the voltage waveform is sine or square..