Question Why is my PSU buzzing/crackling when on UPS battery power?

[gesseler555]

I just put together a new gaming PC with a Corsair HX1000i PSU that is hooked up to my APC Back-UPS BR1500G-IN UPS. As long as the UPS is simply transferring power from the wall (it's line-interactive), or when the PSU is plugged into the wall directly, it is fine. But whenever the mains power goes out & the UPS switches to battery mode, my PSU starts making crackling/buzzing/vibrating noises.

I've heard some PSUs don't like to work with the modified/stepped sine wave outputted by most UPSs and that this noise is just the PSU's circuitry compensating for that. Is that correct? Corsair officially recommends either a Pure Sinewave or Simulated Sinewave UPS for their PSUs (so I guess what my UPS outputs on battery which is what APC describes as "step-approximated sinewave" doesn't qualify as simulated?). What's surprising is that my previous PSU, a 80+ gold-rated Corsair RM850x, was also plugged into this same UPS and it never made a noise like this, and it had gone on battery dozens of times back then. Aren't both these PSUs supposed to have ActivePFC? Or is the HX1000i just more sensitive as it's supposed to be more efficient?

Despite the noise I'm not having any stability issues with my PC itself, even when running Cinebench R23 on battery (but the PSU does seem to make a bit louder noise of the same kind when under load). Under realistic use it would only have to be on battery power for a minute or so before I shut it down (or my building's diesel genset comes on, which typically happens about ~40 seconds following an outage).

Also, I'm using a 6A to 16A adapter to plug this PSU into my BR1500G-IN because this UPS only comes with 6A sockets and my PSU has a 16A plug. Where I live the voltage is ~240v so if I'm right the Amperage should never even hit 6A even if I'm pulling over a 1000W which my system really can't. That said, do you think the adapter is causing this? If so, why would the noise only happen when the UPS is outputting battery power instead of all the time?

Is this noise indication of any harm being caused to my PSU (and as a consequence, the rest of my system)? Should I just get a pure sine wave UPS or is the noise safe to ignore? Is this even due to the waveform? The system is otherwise perfectly stable but this noise coming on whenever the power goes out is making me nervous for my new PC.

P.S. I did talk to a Corsair support rep regarding the issue. While he did recommend getting a pure or simulated sinewave UPS (reading off Corsair's official recommendation I reckon), he also said that if the PSU won't be staying on battery for more than a few minutes the buzzing shouldn't be a big deal. But he also wasn't able to give me a straight answer regarding whether APC's step-approximated wave qualifies as "simulated" or not as per Corsair's own measure, so I'm not entirely sure if he fully knew what he was talking about.

P.P.S I also talked with a local UPS retailer and he said the problem was the voltage drop when the UPS goes on battery mode (city's input being ~240v but the UPS drops to about ~214v when on battery), and that I need to get an On-line Double Conversion UPS (like APC Smart-UPS SRC1KI-IN) to make this problem go away, rather than just a pure sinewave UPS (like APC Smart-UPS SUA1500I-IND) which is still line-interactive like my BR1500G.

But aren't these modern PSUs supposed to work on any voltage between 120-240v (HX1000i is indeed rated for that range)? How could the slight drop to ~214v cause it to make that noise?

 

[gesseler555]

Doesn't seem like I can upload a video here so I'm linking my post on LTT forums with a recording of the noise:

https://linustechtips.com/topic/157...attery-power/?do=findComment&comment=16435026

Use headphones. Wait for the UPS beeping to end, then turn up the volume.

 

[Ralston18]

This:

".But whenever the mains power goes out & the UPS switches to battery mode, my PSU starts making crackling/buzzing/vibrating noises."

That could be due to a number of reasons: Any power strips, surge protectors, other adapters,etc.?

How many devices are connected to the UPS and demanding UPS power support when line power is lost?

The purpose of a UPS is to provide end users enough time to gracefully shutdown the UPS supported system(s).

Not to continue gaming etc...

Power connections being (line diagram where ===> represents a power cable):

Incoming 240 volt electrical service ===> Wall outlet ====> UPS (BR1500G-IN) ===> Power to computer and ===> power to monitor.

Any thing else being powered directly by the UPS: speakers, NAS, printer, etc.? I noted a switched power bar in the video - is that being used to turn the PC on and off? Where is the switched powere bar connected with respect to the line diagram?

Check the UPS User Guide/Manual regarding the use of surge protectors/ power strips. Most manufacturers warn against using them.

Feel free to edit and correct my line diagram as necessary to match your environment.

The PSU may be making noise because it is simply not receiving the power necessary to support the system.

Could be a loop of some sort: Device A connected to Device B connected to Device C connected to Device D and connected to Device A. Any sort of electrical loop is problematic.

More information needed.

 

[gesseler555]

This:

".But whenever the mains power goes out & the UPS switches to battery mode, my PSU starts making crackling/buzzing/vibrating noises."

That could be due to a number of reasons: Any power strips, surge protectors, other adapters,etc.?

Like I said, I'm using a 6A to 16A adapter to plug my PSU into the UPS. Nothing besides that.

How many devices are connected to the UPS and demanding UPS power support when line power is lost?

The PC in question + one LCD monitor.

The purpose of a UPS is to provide end users enough time to gracefully shutdown the UPS supported system(s).

Not to continue gaming etc...

Yes, and that's indeed my use-case. The noise from the PSU when it goes to battery however is immediate, not something that builds up over time or anything.

Power connections being (line diagram where ===> represents a power cable):

Incoming 240 volt electrical service ===> Wall outlet ====> UPS (BR1500G-IN) ===> Power to computer and ===> power to monitor.

Mine would be:

Incoming 240 volt electrical service ===> Wall outlet ====> UPS (BR1500G-IN) ===> 6A to 16A adapter ===> Power to computer and ===> power to monitor.

Any thing else being powered directly by the UPS: speakers, NAS, printer, etc.? I noted a switched power bar in the video - is that being used to turn the PC on and off? Where is the switched powere bar connected with respect to the line diagram?

Nothing else is being powered by the UPS besides the PC & Monitor. Even my home router is powered by a different UPS in a different room.

Not sure what you mean by switched power bar...if you mean the white thing that the UPS itself is connected to (where I threw the switch to shut off utility power) that's just the wall outlet.

There's no power bars or surge protector strips involved. Except, like I said, my 6A to 16A plug adapter that I'm using to connect my PSU's 16A plug into my UPS's 6A socket.

Check the UPS User Guide/Manual regarding the use of surge protectors/ power strips. Most manufacturers warn against using them.

Feel free to edit and correct my line diagram as necessary to match your environment.

The PSU may be making noise because it is simply not receiving the power necessary to support the system.

Could be a loop of some sort: Device A connected to Device B connected to Device C connected to Device D and connected to Device A. Any sort of electrical loop is problematic.

More information needed.

Do you suppose the plug adapter is causing the noise? If so, why would it only happen when the UPS is outputting battery power instead of all the time?

The only thing that changes when it goes to battery as opposed to just transferring power from the wall would be that it starts outputting a different sine wave form. That, and a slight drop in voltage to ~214V as opposed to ~240V.

Do you suppose the drop in voltage is causing a problem in pushing the current through my plug adapter?

 

[Ralston18]

Here are two generic links (not a product recommendation or endorsement) to help understand pure sine wave and square waves.

https://suvastika.com/square-wave-vs-sinewave-ups-for-computers/

https://hilelectronic.com/ups/

You can easily find other similar links and tutorials.

It does not surprise me that "support techs" do not provide direct answers. Much of what is advertised for specs, performance, etc. is based on ideal circumstances - likely in a laboratory environment.

I would not expect a plug adapter to be the problem - however, if of poor quality and condition then it may be partially involved. If possible replace or swap in another known working adapter from another system.

And the voltage drop is certainly of concern. Most devices have some tolerance for voltage variations, Others do not. The drop (26 volts) is about 11% and may be in excess of the PSU's design. 214V likely being at the lower end of the range. Another factor is the test equipment or other source that is providing the voltage reading(s). Most have some + or - accuracy which means that, for example, 214V could actually be 210V or 218V. 210V being even worse from the PSU's point of view.

Two things: voltage itself and the quality/stability of that incoming voltage.

Not sure about the switching. If you are using a power bar simply to simulate loss of external power that is one thing. Having the power bar continually in use to turn components on and off is another, Check the UPS's manufacturer's documentation for allowed or supported devices that can be used with the UPS.

I would simply unplug the UPS from the wall to simulate power loss.

You mentioned another UPS: That is okay (I have two myself). However, you must be careful about creating any ground loops with respect to electrical systems. Does not seem to be an issue as I understand the posts thus far. Just take a look around and sketch out devices and connections....

From APC:

https://www.apc.com/ca/en/faqs/FAQ0...ccur in,places with varying ground potentials.

If the PSU does not normally buzz when external power is present then it is likely okay. If the buzzing begins only when the UPS is providing power then the problem is likely PSU related. Either low output voltage and/or the quality of the "sine wave".

What is that second UPS - can they be swapped for testing purposes?

 

[gesseler555]

Here are two generic links (not a product recommendation or endorsement) to help understand pure sine wave and square waves.

https://suvastika.com/square-wave-vs-sinewave-ups-for-computers/

https://hilelectronic.com/ups/

You can easily find other similar links and tutorials.

It does not surprise me that "support techs" do not provide direct answers. Much of what is advertised for specs, performance, etc. is based on ideal circumstances - likely in a laboratory environment.

Thanks for the links!

I would not expect a plug adapter to be the problem - however, if of poor quality and condition then it may be partially involved. If possible replace or swap in another known working adapter from another system.

I have tried a different adapter of the same make & model (I had bought a 2-pack) and there was no difference in the noise.

I've ordered a different brand of adapter which my friend uses on his APC Easy-UPS BVX2200 (also a stepped-sinewave model) and which is known to not present any issue with his Asus ROG THOR 1200W PSU.

I'll see if it solves the problem.

And the voltage drop is certainly of concern. Most devices have some tolerance for voltage variations, Others do not. The drop (26 volts) is about 11% and may be in excess of the PSU's design. 214V likely being at the lower end of the range. Another factor is the test equipment or other source that is providing the voltage reading(s). Most have some + or - accuracy which means that, for example, 214V could actually be 210V or 218V. 210V being even worse from the PSU's point of view.

Two things: voltage itself and the quality/stability of that incoming voltage.

So I checked the HX1000i's spec sheet (linked below) and it's rated to work with 100-240V input. Presumably so it can work in any part of the world.

So I don't see how the relatively small drop to ~214V could give it a problem.

https://www.corsair.com/us/en/p/psu...tt-pc-power-supply-cp-9020259-na#tab-overview

Not sure about the switching. If you are using a power bar simply to simulate loss of external power that is one thing. Having the power bar continually in use to turn components on and off is another, Check the UPS's manufacturer's documentation for allowed or supported devices that can be used with the UPS.

I would simply unplug the UPS from the wall to simulate power loss.

There's no power bar at all. I'm just throwing the switch on the wall outlet to simulate power loss. But I'm not unplugging the UPS from the wall though, just so it continues to remain properly grounded.

You mentioned another UPS: That is okay (I have two myself). However, you must be careful about creating any ground loops with respect to electrical systems. Does not seem to be an issue as I understand the posts thus far. Just take a look around and sketch out devices and connections....

From APC:

https://www.apc.com/ca/en/faqs/FAQ000259682/#:~:text=Ground loops can occur in,places with varying ground potentials.

If the PSU does not normally buzz when external power is present then it is likely okay. If the buzzing begins only when the UPS is providing power then the problem is likely PSU related. Either low output voltage and/or the quality of the "sine wave".

Yeah it definitely looks like a sinewave issue.

In addition to all that's been said before, do note that I've been using this same UPS in the same location with my older PC and its Corsair RM850x PSU for over 6 months and it never made any noise even when it went on battery.

What is that second UPS - can they be swapped for testing purposes?

It's an APC Back-UPS BX1100C-IN. It's also a stepped sinewave, but the battery is really weak and needs replacement. I wouldn't trust it to run my PC for more than a few seconds and even then I'm risking it just immediately shutting down.

I've talked with a local UPS dealer and told them to bring over a pure sinewave UPS for a demonstration.

I'll report back with any updates.

 

[gesseler555]

@Ralston18

I'm pleased to announce - I went ahead and tested it with a Pure Sinewave UPS and voila! The buzzing issue is gone!

Turns out these high-efficiency PSUs can indeed be quite picky regarding the kind of waveform they like to be fed.