OT: UPS manufacturer

Nov 26, 2025 Last reply: 2 months ago 134 Replies

The USB interface on it is USB1.1 .

formatting link
# "Network-UPS-Tools"

If you're on Linux, then perhaps that "nut" will work ?

*******

USB3 is XHCI. But the hardware consists of two parts. A USB2 part and a USB3 part. The USB2 part can be used for negotiation and "stepping up" to USB3 mode.

The USB2 part, should support both USB1.1 and USB2.

It is USB4 or so, which no longer supports USB1.1 . That's a voltage issue. A USB4 may involve mostly low voltage differential and the I/O pads may not be suited for the

5V logic of USB1.1 . While there are a few motherboards now with an Asmedia USB4, the penetration of USB4 is not all that great yet.

Even on a modern PC (like mine), that has USB2 ports and USB3 ports on the back, my USB2 is just the USB2 "portion" of a USB3 port. The super high speed signals are being thrown away, so that only the USB2 part connects.

In the old days, there was USB1.1

------ Black box -------

When USB2 came along, there was a mux and negotiation. In terms of bandwidth, the USB2 block was shared by six muxes, the USB1.1 block by two muxes. So while USB2 might be 60MB/sec at wire speed, in the old scheme, if six USB2 ports worked at the same time, they would get little bandwidth each on average.

USB2 block -------|\ | \_____ Eaton UPS (negotiating for only a USB1.1) | / USB1.1 -----------|/

Whereas on a modern PC, the USB2 interface is no longer a mux, it is a hub, and the hub changes modes to one interface standard or the other. Because it is a hub, it reduces by one, the max number of hubs in a row.

Normally, if you wanted to attempt an experiment, you would plug a PCI card with a NEC USB2 chip on it, into the PC. But as modern PCs had the PCI slot removed (there was one PCI for a period of time, but they've stopped that now), I can no longer insert a PCI card to test for "better behavior with authentic/ancient kit".

I even have a USB1.1 PCI card :-) But again, I can only fit that in the 12 year old Test Machine, and not in any of the newer AMD machines. If I need to go backwards, I think I have just one card to use to do it.

# Add-on USB1.1 chip in the junk room... Card made by Keyspan.

formatting link
Your ID of the device then, uncovers a bit of unsavory behavior, but we must not give up and "science marches on".

USB1.1 has two speeds. The 12Mbit/sec speed (wire speed 1.5MB/sec), does USB Mass Storage at 1.0MB/sec. But the lower speed on that port, talks to keyboards and mice. You don't need 12Mbit/sec to talk to a keyboard. The second speed is low enough, unshielded (thinner) cable is used to keyboards of ancient times. You might find a captive cable on a USB keyboard that is thinner than normal, and there might not be a ground shield on it.

Summary: I just *barely* still have the hardware to test your device for potential issues. I could install the ancient USB1.1 card on the PCI bus, plug in the Eaton, then see if a booted Linux finds a stable signal. The computer in question, still runs most everything (it has Win11 on it now). It has had WinXP running on it (requires turning off AHCI on the SATA port). And that's one of the reasons it is called the Test Machine.

Paul

Yikes! That article is above my comprehension. It seemed to conclude that there is no answer, and it's Eaton's fault.

As I still do not have a functioning unit, I will leave it until that happens. I could ship you one of the bad ones, the one I have still set up seems as though it wants to talk, but they weigh a ton, and will cost as much for shipping.

All in all, I shan't be too worried if this 3S exercise dies, then I would look for a different non-APC, non-Eaton, UPS. One which is easy to set up to talk to my PC.

Thanks for thoughts.

It doesn't work that way. The USB3 interface will probe the link with a special pulse train to see whether there is a USB3 device at the other end of the cable. It looks at the termination impedance. If there is, they connect. If not, then the USB2 interface connects instead. This means that you can disconnect the USB2 D+ and D- connections and the USB3 will still negotiate a link which works fine. I have proved this by making cables with missing D+ and D-. They work if both ends support USB3.

John

Getting ready to abandon Eaton, I looked up same UPSs, and found the CyberPower BR1000ELCD-UK. Does anybody have any comments on this brand and model?

A customer replaced about 30 APC with Cyberpower a few years ago; they seem to have replaced them all with something else already ...

I have an Eaton 5P1150i which has a front panel for setting most of the relevant parameters and a USB connection which works with NUT for status info and shutdown quite seamlessly on several different motherboard USB ports.

On the negative side, I had another similar machine which has just melted most of its battery charging circuits after nine years = not sure if that is within the range of acceptable after 9 years. Remains to be seen if it is repairable as a spare.

My mobile failed on its way to a strange destination in a strange car with no charge socket. (long story).

Battery was 6 years old give or take. Now on its third battery...

Do you have any experience, actual or anecdotal, with the Power Smart brand?

I saw Amazon suggesting Amazon Basics, not sure if I would trust them.

Vertiv looks good, but not priced for the home PC user.

The whole consumer UPS world seems a bit of wild west.

Sadly... correct.

The progeny of foreign lands. Never buy from the bottom tier. That's one rule.

formatting link
When we had NiCd powered electric screwdrivers, those kinda sucked. They went through batteries like crazy. That's because there is no precision charger for that chemistry, worth a damn. The batteries had to be fast charge type, to design a charger for them.

Maybe lead-acid UPS are similar. A technology with no design progress. Each company just keeps doing cost reductions on their design. That's probably how Cyberpower got its reputation.

formatting link
"3. Eaton 9SX 5000i RT3U A flexible UPS for businesses

The Eaton ABM battery management system uses a three-stage charging technique to expand the battery life by up to half. "

Page 2 here, shows their charge cycle design. It doesn't float all the time, but it does use a (short) float cycle.

formatting link
Somebody knows that we out here, hate the way we're getting shafted on "new" lead-acid batteries. Since lead-acid cells in the past, lasted longer, we know there is something wrong with the supply chain. Is it calcium in the plates (via recycling) ? Who knows.

A Tesla Powerwall, based on Lithium chemistry, might function as a UPS, but it is a relatively expensive toy. And not all competitors to the thing, have the same feature set.

There is a competing product on wheels, that charges, or provides inverter based power, but... it can't do both at the same time. When the battery runs down, you have to charge it first, before plugging in a load. And that's a completely worthless feature set.

In Japan, they have UPS with lithium cells in the unit. But that practice may not be common elsewhere, quite yet (due to transportation rules). The current draw from the lead acid, is rather high, and the battery puts up with that. Tiny lithium cells running at mega-amps, there's a limit to how hard you can push those, and keep the rated service life. At least in cars, there is liquid cooling for the Lithium pack.

Paul

Actually there is.

All NiCd are inherently fast charge, and fast discharge is what you need for power tools and fast charge is the flip side of the same coin

APC now sell small (1000VA) UPSs using Lithium batteries, but not at the bottom end of the market.

formatting link

Well, this one has legs. I am having some medical treatment, so I informed Eaton that they should ensure that the courier should: a. Let me know in advance about the delivery, and b. Ensure that the unit was NOT left unattended on the doorstep, as my house is right on the road. Yesterday, I asked what was happening. I finally got the report that it had been delivered the previous day, by FedEx, while I happened to be at my local clinic getting treatment. Not only was there no sign of it, my CCTV showed nobody approaching my front door at around the specified time of delivery. I followed the FedEx Proof of Delivery trail, and it just says that it was delivered to my village, with no other address details, not even the Postcode, and no photo. It also says that FedEx has my signature on file, which is news to me. So I have thrown this back at Eaton, and am waiting for news and hopefully a resolution, in the form of the missing UPS.

Watch this space..............

And the news is: I managed today to get delivery information from Fedex. No photo, but a map with the GPS co-ordinates of where it was delivered. It is about three miles away. It gives my correct address, but the pin on the map shows it in the next village, it gives a set of GPS co-ordinates which are also in that village, but not where the pin is! The co-ordinates imply that it was delivered to a church, the pin implies that it was delivered to the local school. So somebody is sitting on a brand new, and hopefully working, UPS that he/she/it/they (choose your pronoun) didn't order. I have passed this information back to Eaton, and am waiting for Fedex to inform me when they are going to deliver a unit to me. That would be a first, to actually be told in advance that they are on their way.

Don't hold your collective breaths.

Ye gods! I just took a look at NUT, and it is a whole lot of work that I need to do understand it. I am not a programmer by trade, and much of that is Greek to me. Maybe later.... But at least, At Last, I have a working Eaton UPS. It's only taken 5 months.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required