PC is buzzing on battery and UPS is absurdly loud under specific load

[lm-]

Hello, recently I have bough used SMT3000RM2U and it behaves differently than I would have expected (that does not mean my expectations are correct). I have 2 weird cases.

First case: PC is buzzing when running on a battery. What is strange, it seems that the higher the load - the lower the buzz. Check a video (sound): https://www.youtube.com/watch?v=B3w09ArUqz4

1. At first, PC is running a game, so there is like 400W of load - PC silently buzzes and you cannot hear it on a video, but the buzzing is there.
2. At 6 second mark, I have "tried to force an unexpected reboot" by hitting the reset button - that is when you can hear the buzzing going louder and audible on video.
3. at 19 second mark, I turned off the PC, and yet you can still hear buzzing. At this point there is almost no load on UPS.

Second case: I connected low load things - laptop, guitar amplifier, audio mixer, some guitar pedals, all taking probably about 20-30W. When I unplug power from UPS, it starts to be ABSURDLY LOUD. And the sound gives me chills. Check the "music video": https://www.youtube.com/watch?v=t1C5FacO0wY

I also have SUA3000RMI2U, but it does not produce that kind of sounds.

* UPS is grounded.
* My PC power supply does have Active PFC, but UPS is supposed to output "pure sine wave". I checked with my multimeter - that states it has no true RMS, and voltage reading is 230V.

Should I be worried? Should I service that UPS maybe? This does not seem like a normal operation, but surely I'm no expert.

[Teken]

Appreciate the time you took to provide a video of the problem you see. Generally speaking almost all UPS make some kind of noise when either online or when on battery only power.

If a unit has a fan it comes down to the size, type, quality, and if it’s managed. Almost no UPS maker offers a PWM Fan on their units?!? 👎

As it relates to the noise you hear from this specific model UPS it’s the inverter / transformer that is making the sound. Some models the noise is very quiet vs others it’s very noticeable in a quiet room.

The larger enterprise units are normally intended to be installed far away from the end user in a closet or server room. So noise isn’t a major factor for these environments.

At first blush the noise does seem excessive but it can be more noticeable in a small enclosed space. 

Can you elaborate more on what you mean by using the none True RMS meter? 🤔

NOTE: There are several standards to understand and accept. APC calls out what the maximum sound level (dB) is at a specific distance normally at 3 meters distance. The user manual will say something like 40-53 dB. If you had a sound meter and measure say 54 - whatever - it’s over the rating.

Keeping in mind the human ear is easily fooled to thinking something is louder / softer when it’s just the pitch / frequency being outputted differently. 

A perfect example is a car horn, smoke alarm, microwave timer. 

In your video it’s obvious there are high frequencies present. You may very well find the sound level is within the stated dB level. But a sensible person would also know regardless of the overall (output noise) the sound is NOT acceptable or normal! ☝️

The unit is clearly out of warranty so there’s nothing for you to do from a (APC) technical support point of view. Regardless, there are basic and common things you could do if qualified or simply hire it out.

Fan: Confirm the fan is operating correctly and spins freely because it’s common to see dirty bearings to gummed up blades. If you hear ticking, humming, and obvious squealing replace the fan.

Capacitors: Almost every unit produced by APC has seen bad capacitors. As such it makes sense to replace as many of them where possible. Capacitors provide filtering, power flow, isolation, and regulation in a typical power supply. It can / may reduce the noise you hear.

Transistors: A failed transistor can cause random noises. Unless you’re really committed it doesn’t pencil out to replace them due to the cost benefit ratio. Keeping in mind you only replace a defective part not just whole sale replace them unless it’s for something better.

Transformer: There is nothing for you to do except replace the entire UPS. You can not buy this proprietary transformer part. 

Grounding: This is extremely important and typical noise can be due to ground loops, loose, dirty grounds. All you can do is ensure the outlet to the service panel wiring is firm and tight.

On the UPS insure the AC plug is clean and all of the pins are solid. If you open the cover verify the other end of the cord is secure and corrosion free. Inside of the unit you can clean all ground points and ensure they are securely fastened to their respective ground points. On the main board you will see lots of screws holding it down. Each one is a chassis ground point. 

This would be a good time to remove any dust / dirt from the entire unit. From the video you provided it’s clear this unit is very dirty inside and this results in bad airflow which equals hotter temperatures which equates to shorter service life.

Larger units like yours (depending upon model) had plastic air baffles. You’ll want to clean them if present and insure they are properly seated. As they can cause noise if the baffle is flapping around or is cut which makes a whistling noise. 

Let us know what you find or plan to do! Upload some photos of the inside before you do anything. Show the fan, main board, and the transformer. 

Cheers! 👍🍺

[lm-]

Hi, thank you for detailed reply.

All my assumptions are basically a comparison to my other APC UPS (older unit, but also smartups) unit that I have, and it behaves differently (yeah, it does make sound, but not as nearly as the newer unit).

By non true rms meter I mean, that if UPS outputted sinewave approximation, meter would show reding lower than 230V, but since it shows 230V, that's kinda proof that USP outputs "pure sinvewave", which should be compatible with APFC power supply.

I also made another test. And now I have absolutely 0 clue what is happening there.

Only 1 PC connected to the UPS. PC is turned off, and only switch on PSU is on. I put clamp current meter on L. While on line power - it showed 0A, as expected. But as soon as I unplug the UPS from outlet, current flow on that L line goes to 4-5A values. And UPS (built in screen) reports whooping >1500W of energy consumption. Where does all that power go?! Nothing looks really broken or burned...

[Teken]

Would it be possible for you to make a video of this where the power increases to 1500+ watts? 

Im not quite clear as to what is still connected to the UPS and powered on when AC Mains is removed from the wall?

(Switch PSU) <- What is this you’re speaking about? 🤔

NOTE: Just to be clear I totally agree the unit is making excessive noise which isn’t normal. I wanted you to know this in case my previous reply was unclear. 👍

[lm-]

Here's a video

https://www.youtube.com/watch?v=32PGagerrV8

By switch on PSU i mean, near the plug on the PSU you have on/off switch, I think it's physically disconnects the L line.

Ok, let's make it more clear. On video I posted above, there is only 1 thing connected to the UPS. It's a PC, you can turn it on with a button, but in above example the PC is off (but there is still power on the motherboard, and some LEDs are turned on). But this draws negligible current from the grid - so little that my clamp meter shows 0A.

And this is enough for UPS to behave strangely - high power consumption, and high noise. UPS reports 1500W, and I clamped my DC current meter onto battery, and it showed 30A. With 50V on battery, these two readings match.

[Teken]

What do you measure with nothing plugged into the UPS? Where are you making the measurements from? Also, on a related note that sound is just wow! 😮

[lm-]

Exactly! What do I measure when there is no load? 😛 But really, the load is there, it's PCs PSU and motherboard (and whatever electronic is working on turned off PC).

And I measured current in two places.

1) On the outputs of the UPS - between UPS and PC. I stripped some isolation to expose L, N and P, and clamped meter on L cable.

2) Second meter (DC) I clamped right after batter - so between battery and UPS. Cables for + and - are easily exposed.

When UPS is working on utility power - everything is as expected, AC lines to PC shows 0A, and on the battery it's close to 0 as well. But when I unplug the cord, AC line shows 4-5A (I wouldn't really trust that too much as output from UPS might be very distorted and meter might show invalid data, but current is flowing allright). And DC meter shows 30A draining from the battery.

So it looks clear. Power is being transferred from battery to... somewhere. Anyway, would you agree with me that something is terribly wrong with that UPS? I'm afraid we won't really know what is going on there without some magician with an oscilloscope 😞 But I don't have a scope, and surely am not a magician 😞

[lm-]

Hi, little update. Store agreed to replace faulty unit, but unfortunately I don't know what was the reason of such behavior.

I have a small question about new unit. It seems to be working fine - but I've noticed 2 things that don't match to my expectation (which may be very wrong) and it would be nice to have some clarification to put my mind at ease.

1. On network power, on UPS AC output I see about 0.35-0.40A of current. When I go onto battery, it goes up to 0.6A. Now it doesn't seem normal for AC current to go this high? During testing, network voltage was 230.4V, so current increase cannot be explained by lower voltage.

2. On battery, when I turn on the PC, I can hear, a very silent... buzz? It sounds just like when a cable very lightly touches running fan and you hear that silent 'tyk tyk tyk'.

Are those behavior expected in UPS? Or did I get another faulty unit?:)