New R9 280x making screeching noises in certain aplications.

Updated on 16-12-2014 in Graphics cards
14 on 15-12-2014

The issue is isolated to the card and mainly happens with V-Sync off. Catalyst forced V-sycn is broken (what isn’t broken in catalyst?)

The card is 2 days old, freshly upackaged and has not been toyed with in any shape or form. simply installed onto the pc and ran. No overclocking because my CPU is more likely to bottleneck it than the GPU working slowly.

I have a 750 watt power supply and that should not be the issue. It’s connected to via PCI-E, 6 and 8 pin respectively.

It usually happens in game menus, where the resource demand is low but without v-sync you can reach some crazy fps. (CS:GO produces moderate noise on default. Alt tabbing out of the game eliminates the noise. Rome 2 total war is particularly awful and produces really really loud noises. Troughout this, the fan speed remains the same. regardless of the noise and vsync fixes it somewhat again)

Reading up on the topic i noticed that it could be the capacitators vibrating at a specific frequence and that limiting FPS like i said can be an answer however not all games have this option and some that do, cause generaly cause awful input lag. As a long time gamer i can assure you that in multiplayer games this could be the equivalent of an extra 100 to 200 ms ping. The console type FPS limiting like the one in source engine is a bright exception to the rule and even then in main menu at 60fps i can still hear some audiable noise.

The noise is audiable like a mosquito flying around your head when you’re trying to sleep in a rural area on the games which are louder. Emphasis on rural, IE no sound pollution. I am afraid that such faults from the day 1 could result in a low lifespan and as a user that owned a GPU for 6 years without any issues what so ever i am disheartened. The same GPU still works on moderately demanding games (maxed out DOTA2 or CS:GO) but it’s overheating, probably due to thermal gel and poor contact in worse optimized games such as Rome 2, Chivarly etc.

In many of the games i play, main menu is a place where i spend quite a bit of time, waiting for teammates to join party or simply waiting for the 5 to 10 min solo queues because i’m on the higher end in matchmaking rating.

I’ve also used GPU-Z to monitor the temperature.

I get 3 temperatures on that tool, 1 being the GPU temperature and 2 Voltage regulators. One voltage regulator is overheating unproportionately to the rest of the hardware, exceeding temperatures of the other two by ~10 degrees celsius.(gpu temp: 76, voltage regulator 1: 89, voltage regulator 2: 80 degrees.

I have not overheated the GPU and have strictly stopped at 90 degrees on the voltage regulator when i tested settings and i used the tool the very first minute of using the GPU. Considering that this GPU has an operating temperature of ~75+ with stock cooling this should be well within the safe zone and cannot be responsible for the issues i have.

I’ve been avoiding games i’ve bought this card for and i will continue, waiting for some sort of response.

Currently i am dissatisfied with the product and i would like to inquire what my RMA options are and how does this happen. What are the wait times and who to contact. The retail i purchased the product from or is RMA direct between the customer and you as the product provider.

if there is anything i should know or i can do i would like to hear about it. I really, really don’t want a 2 month RMA.

Water cooling or a power supply upgrade, or even bouth is out of the question. I can get the 970 for the price and enjoy higher end hardware with lower power consumption which results in less heat. I am not interested in it however, because the 280x is strong enough.

A GPU should not overheat out of the box and i have experienced some of it, sadly. Unwilling to test further.

I can give dxdiag, cpu-z reports and gpu-z logs for gpu usage/temperature readings

could the asymmetric heating be due to the proximity to the motherboard? I don’t know what the infrastructure of the GPU looks like and where the temperature readings take place. The motherboard isn’t overheating but it could affect how easily the heat dissapates in a certain region.

What are safe operating temperatures, normal load temperatures. How should i test this futher to see if it’s indeed faulty.

 

Edit: Screeching noises happen on temperatures as low as 40 degrees, when i first open the game and is not subject to long periods of gaming.

The gpu is R9 280x Royal Queen

 
  • Liked by
Reply
0 on 15-12-2014

I may also have underestimated the noise with vsycn off. It’s completely unacceptable, having it checked again.

 

Should i remove the GPU and install the old one back?

  • Liked by
Reply
Cancel
11 on 15-12-2014

Hi Invisibleblue,
There are some capacitors that has a little gap in between that causes the noise. These unfortunately is very hard to resolve as those are components that AMD approves for manufacturing. We have found that if you cover the tiny gap with a special glue it does minimise the noise a bit, not entirely and it is a painstaking job.
Based on your description I do think is better to RMA the card, please send a mail to [email protected] and send us your serial number and the description of the problem. Proceed to RMA the card through the shop where you bought it and we will replace it accordingly. Noise beyond what we consider normal is not acceptable.
We do apologize for the inconvenience caused and hope we can resolve your problem as fast as we can.
Welcome to Insights.

on 15-12-2014

https://drive.google.com/open?id=0B3K4gEjB2jigTU1KLTVWcUtSV0MteVZya1lQQ3AzUHFhSG9j&authuser=0

I recorded it with iphone 4s by holding it close to the gpu.

After exiting rome2 the left over vibrations recorded after vsync was turned on stopped.

on 15-12-2014

Tried to access the file but it needs permission. Requested permission.

on 15-12-2014

https://drive.google.com/file/d/0B3K4gEjB2jigTU1KLTVWcUtSV0MteVZya1lQQ3AzUHFhSG9j/view?usp=sharing Wrong link? Sorry about that!

 

 Anyone with this link should be able to view.

on 15-12-2014

That is the typical coil whine on these cards, it is very much in all vendor cards as we use the same components.

on 15-12-2014
That is the typical coil whine on these cards, it is very much in all vendor cards as we use the same components.From Paul

 

In certain applications it could easily disrupt sleep of someone in the next room and that’s slightly an issue. It’s bouth loud and has a high pitch beyond what i consider normal. Do you have any recomendations on how to bouth monitor and control the GPU directly. AMD Catalyst is a broken piece of software and nothing seems to work on it. (overdrive deletes itself after hitting accept on bouth old and new gpu – for manually controlling fan speed on more demanding games, Forced V-sync doesn’t work and i’m not aware of an FPS limiter in catalyst)

 

I assume that the whine scales with FPS and not GPU load?

 

Any other non warranty removing way to tackle this issue?

 

Thanks for the help and the time invested so far. Coil whine is a completely foreign concept. The old GPU’s Coil whine is barely audiable while in the same game, on the same fps settings, the new one is several times more audiable.

 

In rome 2 it goes past all boundries and is actually extremely loud. i really can’t put the intensity of the sound into words.

on 15-12-2014

I was able to get the coil whine under control. The rome 2 total war was rendering at 350 frames per second and with third party fps limiting tools (msi afterburner) it’s working within my expectations on 75 FPS.

 

The 2nd question is the weird temperature reading in GPU-Z. Since the main issue has been solved i don’t feel like i need an RMA but i do want to know whether overheating voltage regulators could significantly reduce the GPU lifespan.

The GPU temperature is significantly cooler than voltage regulator 1 reading in GPU-Z by about 10+ degrees. Similarly Voltage regulator 2 is also significantly cooler.

 

Should i constrain my use to absolutely limit the voltage regulator temperature to 95 degrees or is it fine if i simply ignore that and aim at a reasonable GPU temperature of 85 degrees under full load. (which may or may not overheat the voltage regulator beyond 95 degrees)

 

I’ve already made my own, modified fan speed curve which starts off slower but rises much faster to compensate.

on 15-12-2014

Okay i’d like to correct myself. Going into game (leaving the main menu) the coil whine is audiable enough to hear from anywhere in the entire house if there is no sound polution, with the PC room doors open at 100% gpu use.

at 60 fps and 100 gpu use in rome total war 2 it’s like having a bee hive in my computer.

on 15-12-2014

I think i am satisfied with the perfomance for now. Thanks for the help and informing me about “coil whine”. I’ve done a lot of MSI afterburner tweaking and it’s working within parameters that i am comfortable with, with no additional cooling expenses. I am able to sustain a fully loaded gpu under 80 degrees or so and that should leave enough room for the voltage regulator regardless of it’s design. The sound is annoying at times but i do not play a lot of games which fully use my GPU. I mainly play games like dota2 and cs:go which are not very demanding and well optimized.

Thanks again for the patience and your help. If i experience anything i feel unnatural, i’ll revisit this thread i guess!

Double thanks 🙂

on 16-12-2014

Welcome.
I think it is crazy to tune the card so it does not make the coil noise, that is not the way to go.
Please detail in the RMA the SN and the issue so we can check your card when we receive it. Also we will make sure that the one going back does not have the same issues. 
Take care

on 16-12-2014

I did not “tune” it, i simply limited the FPS because the games where coil noise was happening i got about 500 fps or more. Ontop of that i simply changed the fan speed curve (how fast it kicks in) to keep the temperature in the regions i feel okay with under full load.

The voltage regulator is at 79 degrees while the card is at 71 under 90% load. The fan speed is 55%. The 2nd voltage regulator is at ~70 too though.

The noise is ok after limiting the FPS and adjusting the fan speed and i haven’t touched any other settings. I assume these are all harmless. On low temperatures i reduced the fan speed it and on higher temperatures i made it scale a bit better. The perfomance was okay so far. Is the voltage regulator temperature something i shouldn’t worry about?

on 16-12-2014

I’ll do the RMA afterall, like you proposed. The coil noise is very specific to certain applications and in windows it runs cool and silent like one would expect and like i am used to. however in games it’s not bearable without software limitation. I can still hear the buzz trough headphones after software limitation if i use a volume setting i am comfortable with (louder = a no go)

I am reluctant to send it because it is the holiday season and i really don’t want to use the old gpu because it’s age is really showing on performance and stability but i’ll do it nontheless.

Show more replies
  • Liked by
Reply
Cancel
0 on 15-12-2014

thats an awful BUZZ – Yeah, I would RMA that asap

  • Liked by
Reply
Cancel