I look forward to hearing if you actually get the unit in and how it ends up performing.
I don’t think there’s any question that I’ll get the unit; there are too many user reviews on Amazon to make me think it’s a complete rugpull (although I did notice that fulfilment does not seem to be via Prime).
The real question is how it behaves once I’ve got it on hand, and there really are questions to answer there. I know for a fact the Anker Solix units match their specified watt-hour ratings, because I ran them dead flat to check. I know their sine wave output looks good and the handover is clean, because I made them do it repeatedly while watching over an oscilloscope. Those are some of the things we still need to see out of the Goldenmate unit.
There are also questions about heat and fan noise. I saw that the units do have cooling fans in them; it’s acceptable to me if those kick on while the units are switched for battery output, but NOT if they’re passing through power.
There is some potential for chicanery there, too: I tested a Bluetti powerbank/UPS similar to the Anker Solix ones, and I saw that they spin up the fans every hour for a couple minutes if there’s anything plugged into their 120VAC out. This strikes me as, in addition to annoying as hell, likely evidence that they might not BE switching the outputs when the power goes out… they might (I have not actually disassembled one to try to look at circuitry) just be running the 120VAC out off the battery ALL the time, and simul-charging while doing so.
The Anker Solix, by contrast, doesn’t ever spin the fans up unless it’s either charging at 200+W, or discharging at 200+W without shore power available.
Like I said. Plenty to test. 
I can confirm this same behavior with the Bluetti unit I have. If it is always powering the load off of the battery side, does that in essence make it a “double conversion” setup, which to some could be a bonus feature?
So, that would make it one of what we used to call true “Uninterruptible” power supplies. Technically speaking, what we usually refer to as a “UPS” isn’t–or at least wasn’t always considered that. Standard "UPS"es are actually switching power supplies, meaning that they switch over from shore power to battery power faster than the connected devices can fail.
Way back in the day, the really big iron guys kinda sneered at switching power supplies; they used genuinely uninterruptible power supplies that always powered devices from battery, but simultaneously charged the battery from shore power. This was a plus because those connected devices neither had to worry about switchover times, nor about a sudden change in the power environment as the supply switched.
I haven’t seen a true UPS being talked about in a couple decades, though. Modern switching power supplies both switchover rapidly enough, and feature high enough quality pseudo-sine output (cheap switching power supplies used to output plain square waves!) that there’s just not much point in a truly “uninterruptible” supply, which will need constant maintenance due to the neverending charging.
So. Technically, you might argue that if my suspicions are correct and Bluetti is merely simultaneously charging, it’s a “great feature” because there’s no switchover at all. But in practice, that would mean much greater power inefficiency, more heat generated, and a much shorter battery life.
Again, I don’t know for a fact that’s what Bluetti is doing. I haven’t taken either of mine apart. Maybe they’re just little idiots that insist on running their fans once an hour whether they need it or not!
But either way, in 2024 I would not really call a true uninterruptible power supply an improvement over modern switching power supplies.
Assuming clean power I agree, as I have had this discussion many times at work. We do find them useful still in areas with sketchy unreliable mains power and/or “cheap” generators, as I have seen the constant switching quickly trash some very expensive equipment. The “switching” part of the UPS seems to wear out the more it is used and the switching time starts to get longer. At some point it fails completely, or the time gets too long and the connected equipment shuts down due to lack of power. What surprised me is how bad some of the power is in many areas of the US still.
1 Like
I am way more comfortable with LFP than standard Li-ion, at least from no-name brands. My VRLA battery on my APC 1500S just failed, so I will look for those. Those ANKER powerbank things are interesting, because they make quality stuff, but the 300W doesn’t sound useful for desktop PCs, so I would probably stick to the larger sizes when I get one.
A typical desktop PC will rarely pull more than 100W. For reference, I tested my C300 on an i7-6700 desktop PC plus a 24" monitor, running a non-stop load pulling about 100MiB/sec off the SSD in the PC. The PC and the monitor together pulled around 70W IIRC.
A super-leet gaming PC with a $600 GPU is potentially a different story, but that’s a “gaming PC” not a “desktop PC” and it’s toward the higher end of that spectrum anyway. You might want to break out a Kill-A-Watt or similar power meter and actually watch power consumption before jumping to conclusions, if you haven’t already.
You are right. I tend to think of my usage as typical, but it is probably not. I have a fairly high end GPU and also plug in multiple 32 inch monitors, so I’m probably using more power than a standard corporate desktop with one monitor. That said, the main reason I tend towards larger UPSs is longer runtime. I once finished a WOW raid on battery power after my lights went out ten minutes before we finished. 
I will bust out the Kill-a-Watt after I replace my failed battery. Yes I realize the irony in that statement. This (APC BR1500MS) is likely the last VRLA UPS I will buy.
I grabbed one of the C300 units on a holiday sale and replaced the UPS on one of my main systems. Older i7-6700 with a 32in 4k monitor and two 20in 1080p monitors. This complete setup doesnt exceed 200w when running full tilt, and the Solix handles it fine. From my initial testing it also is showing to be very accurate in its runtime estimates, and the switchover time is so quick that the screens do not even flicker at all.
Thanks agan @mercenary_sysadmin for the suggestion, and hopefully the “UPS” version you ordered from Amazon works out just as well.
1 Like
Funny you posted today, I just unboxed the GoldenMate unit and got it set up, finished reading the manual, etc. First impressions: it’s very, VERY basic. Makes the Anker units seem like absolute Cadillacs. You don’t get so much as a numeric readout for the battery charge level–just primitive bars–let alone a runtime estimate.
There’s also no external comms with the GoldenMate Unit–neither Bluetooth nor Wi-Fi nor anything else. There’s a USB port, labeled “diagnostic interface” in the user manual, with stern warnings that it is not consumer accessible and will not enable “communication with the unit.”
There’s also a faint but audible capacitor or coil whine when the unit is powered on. Not horrible, and a lot of people wouldn’t be able to hear it at all, but if you’re one of those folks with an excellent upper frequency hearing range… you’ll be able to hear it, in a reasonably quiet room, although the whine will disappear entirely if there’s a quiet conversation or kid noisily crunching his way through a bowl of pretzels nearby.
(Yes, that’s a very specific example. Well, I stopped being able to hear the whine from the GoldenMate unit ten feet or so behind my head, so… that very specific example is a very accurate example, too!)
I haven’t had a chance to properly put this thing through run-flat tests yet, but I suspect it will do essentially as-advertised on those. It weighs about 30% more than the Anker Solix C300 that was on the bench before it, and there are those positive Amazon reviews, so I suspect that part will be okay.
All things being all things, I prefer the Anker so far.
I guess it will depend on how cheap it ends up being (it is now $160 on Amazon US) and if that is enough to make up for the missing extra features the Anker has. The C300 appears to have a $60 coupon on Amazon US right now, which brings the price of it down to $190. The $30 difference makes it very easy to choose the C300 for sure.
I do wonder if a unit like this could be used in essence as an extended battery for the C300 to increase its runtime?
Do these also protect the hardware from - potentially unclean - wall voltage?
I’ve seen so much variance in what should be a steady 120V.
@mercenary_sysadmin How did your testing go on the Goldenmate device you received?
Thoroughly “meh.” It doesn’t seem like a scam device, but it’s… Basic. Way too basic for me. The biggest issue is that it doesn’t give either runtime estimates OR charge percentage at all… You get a stupid little segmented battery icon like what you’d see on a phone. (Except phones usually also show you the percentage in numbers next to the icon. This does not. At all.)
There’s also no over the network communication, so there’s a lot of stuff you can configure on Anker or Bluetti (not that I recommend Bluetti when Anker is an option) that you can’t here. You also can’t even make airy promises about tinkerers being able to gin up their own auto shut down system–no exposed metrics over either LAN or USB or Bluetooth==>no way to either tie it into NUT or just script your own relatively simple check.
I’d be willing to test future, upgraded versions of the goldenmate, and I’ll find a solid use for the one I already have, but I’m not buying any more of this particular model.
To be fair, I say this because I have other LiPO options available. If it was either this or lead acid, I’d personally be buying this, warts and all.
It’s still early days, and I slacked off on the real world load testing in this one once it was clear I wouldn’t recommend it. But AFAICT, it should live roughly up to its stated capacity (and the expected runtime that gives you), and my test server didn’t mind its power switchover times a bit. No flicker on the monitor, clean sine output as observed on an oscilloscope, etc.
Any thoughts to using these as just a battery to extend the runtime of a smaller UPS or power bank like the C300? At least the Goldmate unit has a built in mechanism for charging, unlike hooking up a bare battery to a power bank using the solar input to extend runtime.
Hopefully we will see more options in the space at these lower price points, and with more traditional UPS features.
Risky. When the power is off, you need to worry about whether the equipment-side UPS thinks something looks shady about the wall-side UPS’s output, in which case it would ignore the wall-side UPS and immediately engage its own battery instead.
That’s bad enough, but the charging cycle is potentially worse. You have to make VERY certain that your wall-side UPS is capable of delivering the maximum power demanded by the equipment-side UPS. And you have to make sure that it’s capable of that WHILE THE POWER IS OUT as well as while it’s on; it’s not exactly a once-a-century event to have a long outage (that drains both batteries flat) followed by a short one (that leaves the wall-side UPS having to supply the charge rate demanded by the equipment-side UPS, while still on battery itself).
That’s if you hook them up in series, of course. You can’t hook them up in parallel, because we’re talking AC, not DC, and the odds of both devices producing precisely in-phase sine waves during an outage is essentially zero, so you’d have a never-ending and potentially MASSIVE current flow ping-ponging between the units. I’d pretty much EXPECT that to cause a fire, very likely the first time it happened. DON’T bridge separate AC sources together!
Yes it would be in series, with my thought being the goldmate plugged into the wall and the C300 plugged into it. This way during a power failure the C300 would not recognize anything had happened as the goldmate would continue to provide it power. Concerning charging, luckily the Anker units do let you set the maximum it will pull from the outlet to charge. This combined with knowing the equipment load should allow one to keep the total draw under what the goldmate can reliably provide.
Good points about the extra planning needed in sizing things, and an interesting thought exercise at least.
Just test, test, test, test, test. You’re literally (potentially) playing with fire!
Just picked up an EcoFlow river 3 plus and it has a usb-b port for communication. According to the quick start guide they have a “power manager” software that can be installed on the connected computer for automatic shutdown functions. This appears to be the piece missing from the Anker units keeping them from being a true UPS replacement.
It also advertises less than 10ms switchover time which also puts it in the UPS category as opposed to just a battery baxkup.