Updating realtek audio driver fixed usb mouse freeze.

[Adelin]

Hello,

I thought my motherboard had usb 3 ports problem, or a uefi bug / windows driver bug.
In fact, my Roccat pure kone military mouse, was freezing every two minutes

6.0.8809 driver from Realtek for audio chipset alc 892 fixed the problem.

I don't understand how an audio driver is able to interfer with usb mice, i looked for different forums, and it's not the first time, realtek driver update or downgrade fixed the issue.

I would like to share this expêrience, so players with mouse freezing fro one second every two seconds, may have one from multiple solutions available :)

[EmotionallyBroken]

That is indeed weird.
Someone wired something VERY far out of circuit, to paraphrase.
Thats kind of like chaning your tire and suddenly your alternator works again.

[Adelin]

Mouse freeze, can be a conflict from software or hardware.
I remember before, users, was checking there was as little as possible or no irq sharing :

https://www.tldp.org/HOWTO/Plug-and-Play-HOWTO-10.html

About computers, everything may happen lol

https://www.overclock.net/forum/18-sound-cards-computer-audio/1619627-windows-10-realtek-hd-audio-drivers-causing-issues-solutions.html

The first message posted, is interesting, even if different than mouse freezing for me every minutes or two for one second, while mouse cursor is responsive the rest of the time

[Snow]

Originally posted by Adelin:

I don't understand how an audio driver is able to interfer with usb mice

Lots of hardware including audio work by sending interrupts through south bridge to CPU. This means whenever your PC has a sound to play, sound chip tells CPU to immediately stop everything it is doing and process the audio right now. This is indeed an important feature to have, as this way your audio will always be processed without delays. For better or worse, USB mice, while also working via south bridge, do not send interrupts, they work by polling, basically waiting for PC to ask them "whatchadoin?" up to 1000 times per second in a modern PC. Some simple error in driver might lead to some command being processed for too long or repeating itself over and over until the job gets done or until error is simply skipped, hogging your SB/CPU and resulting in USB devices simply not receiving polls for some time, which is perceived as a stutter. I also had this kind of problems with onboard audio/lan, but the worst I seen was a drive running in PIO mode instead of IDE - just imagine every slow sector resulting in stutter, it's basically only usable as long as you don't launch any apps at all.

[Adelin]

Hello,

Thanks snow, this information is interesting.


It was horrible when hard drive was runing in pîo mode when i had a standard bios with only raid, and no ahci support.
With sata, i read that even if emulating ide mode, it won't have pio, that is an ide thing.



With old usb devices like a still series shift, just usind hid driver coming from win 10 install work / or win 10 update one.

i guess not adding extra software without 10 compatibility reduce risk of bug from a added layer.

But i already checked without lan, keyboard, and just gpu + ram + cpu + powersuply connected.
The interrupt issue was still present, and i never thought about checking audio by disabling it in UEFI :)

[Adelin]

I had the problem again after one day.

Opening the mouse ( unswcrewing from under mousefeets) cleaning prism and sensor with distilled water, checking all connections , removing dust, fixed the issue for sure.

Three days from now, and no more stuttering.

I will never use again cotton compress under my hands, and i'm used now from the feelingcontact of plastic.

I tryed swaping corsair ram with g skill and so on, one barret at a time,
Putting hd 7870 instead of rx570 8gb,

Just putting mouse on one port one at each time from asmedia, intel and p400s case.

I bought a i219 network pciexpress card ( not to fix this problem but for other reasons) and disabled on board lan.

Trying the mouse on another new windows install on another hard disk.

And only cleaning each components and ssensor worked :)

As i removed the glue from mousefeets, ( an error, as mousefeet lenght increased, and teflon was not sticking to the mouse with glue ( no teflon glue at home)

Roccat is sending me free mousefeet.

I'm happy, right now, it will be like a new mouse lol


Edit 2019 december 18:

Mouse freezed a little time to time, one week after i uninstalled and reinstalled the windows 1909 enablement package.

Why id did that :(

Anyway, over time i fully cleaned the steel series shift keyboard, removing all screws ( + one under the sticker "do not remove".

I had lot of fur from my previous cat, then i rmeembered all the good time we paste, it on my knees, me at computer, almost feeling sleepy, every time my cat purred :(

It was the good time.

Now i have two poodles :)

Ok so there was lot of furs, i wouldn't imagine, it would go even under silicon pcb.
Cleaned everything with distilled water + cotton pad, used compressed air before.

It helped a lot with mouse stuttering, reducing it by 50% at last.

Then i reopened mouse and dismantled it to clean it again with disstilled water, compress air, even the sensor.

it helped again by maybe 10%

I think i fixed again 20 % of mouse stuttering later by:

Disconnecting the white led on shift keyboard, that illuminate STEELSERIES letters.

I think it's possible even if led does not need lot of power, something went wrong, so mouse is not stuttering again after all those steps.

New adrenalin 2020, even if buggy for most people, fixed stuttering a bit, same with 6.0.8858 realtek driver.

As well i disabled onboard lan,installed i 210 ethernet pcie hp , but it did not help at this time.

Snow answer is 100% valid :)

From now, i still think it's possible every steps to fix stuttering was hazardous, and maybe i owuld have to replace the kone pure military usb cable, to be sure everything work well.

There is another theory:

The kpm, doesn't work well on vynile wood like surface.
I tryed it, and sensor went off and on, mouse wouldn't work half of the time.

It's possible the corsair mouse pad bee, black, or maybe the foam surface, doesn't help the kpm either.

But the sensor would have fail at start of this new year, when i got the foam pad.

So maybe there is a sensor problem or usb problem.

But it would not explain 100% why previous steps removed stuttering two days ago.

So Snow answer, is my favorite explaination.

[Adelin]

Ok i'm furious lol

I installed amd driver december 18 or 19 version adrenalin 2020.
I choosed the remove all settings, from adrenalin installer, and after reboot, i found all my usb , ahci , audio and lan drivers, where reinstalled, and there was a double item for each of thelm + hdd, while checking ghost devices in devices manager.

It's not a problem, but something went wrong, and now my mouse is stuttering like it did before all the fix i tryed.

I guess i will have to uninstall all devices and reinstall them one by one, as audio driver broke, my realtek devices got no sound, and when trying to change sample rate, it sayd there was an error.

I fixed it by reinstalling realtek audio driver from device manager, but still has mouse stuttering.

I was about to buy the kone emp, at not expensive price, but it's not sold anymore, so i will wayt my birthday, to get the aimo or aimo remastered, that has the right shape for me.

Anyway that was funny in a sense and show me, what ever fix i thought was working, was not the true answer to mouse stuttering^^


Edit december 30:

I found the culprit:

I tested with a small asus mouse, and got no stutter on mouse.

Put the military back it stuttered.

I opend the mouse, as mouse wheel scrolling wasn't working since few weeks, and saw three very small cables with 3 points soldering.

One of the cable was touching the others, but it was not the cause of the stuttering.

If i put pressure on the cable, the mouse scrolling work again, so i tryed to use iron soldering kit, and failed with cables i was not able to put again.

Anyway, even with all cables cut, the stuttering was present.

I found also i had middle mouse click not working, and putting pressure to the omron switch, sometime fixed it, sometime not.

At the end, i can't say if the usb cable was faulty, or a component isinde the mouse ( probably sensor) was faulty.

I just went to buy a kone aimo remastered, as it's actually at same price on roccat shop as emp and old aimo from a big known online store as evil as voldemort,

At last, the hardware diagnostic is done, i thanks one friend who told me to test with another mouse, a thing i never tryed all this time .


January 6 2020:

I received a kone aimo remastered, and now, i can access uefi, without mouse that need to be plug and unplug again to work.

Same after a complete shutdown, in windows, the mouse is already working, without connect disconnect required.

I thought my motherboard needed to be repaired, and had usb ports not working, but it was wrong.


ALso, the kone aimo remastered, is better fit for me, my wrist is not slowing down the mouse anymre, as the mouse is larger, i don't need to put as much pressure to maintain it.

I thought the heavy 130g weight would be an issue, but if you don't lift it often it's perfect accuracy.

A low weight mouse, would be required for players in need of fast movement, as i can do 2 time more moves from way less inertia from a very small asus mouse.

But more weight is not entirely bad, as micro movements won't translate to mouse cursor .

Thing is i loved the weight of 90g, from ko,ne pure military, but i have an increased precision with theaimo remastered.

And more important as i sayd, wrist as some sort of support from the mouse dimension.