The z97 has 2 chipset have 2 ehci controllers(usb.2.0) and one xhci usb.3.0 controller.
The intel specification leaves the motherboard developers to run usb.2.0 devices from ehci or xhci.
The usb 2.0 devices in asrock run from xchi controller (usb 3.0) (pic 1) and not use the ehci controller.(pic 2)
If u have the usb3.0 on (enabled ,auto or smart auto)
1
2
There exist an option in bios that is locked that Allows you to manually rout each USB
3.0/2.0 port to xHCI or EHCI?
Iwant the usb.2 ports to rout from echi and usb3.0 ports from xchi is this possible?<br />
And i found this:
USB 2.0 Host Controller Selector Mask (USB2HCSELM) — R/WL.
This bit field allows the BIOS to communicate to the OS which PCH USB 2.0 ports can be switched from the EHC controller to the xHC controller.
When set to 1, the OS may switch the PCH USB 2.0 port between the EHCI and xHCI host controllers by modifying the corresponding USB2HCSEL bit (D20:F0:D0h, bit 3:0).
When set to 0, The OS shall not modify the corresponding USB2HCSEL bit.
If i provide the uefi bios can anyone check it?
EDIT by Fernando: Thread title customize to make it more specific
@boombastik - Link to BIOS, or FPT dump of your BIOS region please.
forget what i have say above.
I have 2 motherboards from asrock that after installing gpus the onboard sound card quitting playing.
in photo that option should be enabled?
I attached also my bios.
Z97 L2.14.rar (5.38 MB)
@boombastik - Uninstall all GFX Driver, run DDU (Display Driver Uninstaller), then reinstall driver again and be sure to uncheck anything that says HDMI Audio.
Sorry, I don’t know about those settings you selected in image, but I doubt they have anything to do with audio disabling after install GPU’s, this often caused by GPU trying to use it’s own audio (HDMI audio for AMD, unsure what Nvidia one is called)
All these options in that folder i circle (red color) is all the uefi codes that running during boot?
I understand this correctly?
That is debug folder, some are used/live and some are duplicates of settings visible to you in other sections. When duplicates they may not be used, some may not ever be used nor apply to your current hardware
When not duplicates and do apply to your hardware they are usually live/in-use (setting shown at optimal column is what is currently applied)
I am tracking an old bug with windows timer resolution that stack randomly at 0.5 ms after a pc restart.
At first I thought that is a driver bug, but this exist in windows 8.1 and 10.
I changed my psu, gpu, I removed all expansion cards and it continues.
But I found the culprit .
The problem is the option boot performance mode that have the value turbo performance.
This setting need to be max-non turbo performance and then the os will then control properly the turbo boost.
It is possible this option to be added in CPU sub-menu?
Also the on board sound car that stops playing is hardware related at least 3 gens that have similar PCB.
the internet is full of them.
-https://forums.tweaktown.com/asrock/5562…s-z77-pro3.html
-https://www.reddit.com/r/techsupport/com…me4_rear_audio/
-https://forums.tomshardware.com/threads/…-sound.1420185/
The problem is that the speakers are detected the windows shows that the music plays, but i dont listen anything.
With more serach i found this from hwinfo64:
at restarts the sound codec recognized like this:
It is like its firmware rewritten by other sound cards.
Good find on that bug, thanks for sharing, hopefully it will help some user in the future 
On the audio issue, does the same happen without NVidia graphics drivers installed? Uninstall them with driver cleaner/sweeper etc.
I’m wondering if it’s a Nvidia/HDMI audio type issue, I see this sometimes with AMD/ATI drivers if audio driver from GFX package allowed to be installed during driver setiup
It happens the same without GPU installed (software + hardware).
In an another site @Lost_N_BIOS unlocked the menu of the same motherboard that i have. (Asrock z97 anniversary)
In my mind i knew some bugs so i take the opportunity to play with the unlocked options.
I found the critical options:
1) USB keyboard and mouse don’t work on safe mode.
Solution:Change intel usb 3.0 mode from smart auto to auto.
2) Some usb sticks and wifi stick dont detected properly in windows 10 but works ok under windows 7.
Solution: change xchi hand off from enabled to disabled.
3)Some legacy pci sound cards are detected as VGA.
Solution: Before u install the pci legacy cards change the Link Training Timeout (uS) from 100us to a value like 1000 us. Asus server motherboards use the value 400 and above with 1000 maximum. Asrock x99 motherboards use the value 1000us.
4)Windows timer resolution that stack randomly at 0.5 ms after a pc restart
Solution: change the option boot performance mode from turbo performance to —> max non turbo performance.
If u need ultra fast boot with gop don’t enable it because it bypass this option. Workaround: Disable csm (compatibility mode) first and then enable fast boot (no ultra fast boot).
5)My keyboard start my PC even if i have usb power on to disable.
Workaround: Dont exist. Probably your keyboard is usb1.1. This is platform errata. USB keyboard 1.1 will usually start the PC from s5. U can try with a keyboard that is usb2.0
I asked Asrock for these options and their support provided me a new bios that have all these four options user selectable.
I uploaded it here:
Z97 Anniversary L2.15.zip (5.43 MB)
Also in all asrock motherboards from haswell (chip set 8-9) the deep s4 and s5 don’t work. The motherboard dont shutdown when u enable deep s4,s5.
The reason is this board capability. The deep s4, s5 to work properly need the board capability to be deesx but asrock select the other value.
@Lost_N_BIOS is it possible to find out where the id of the hardware audio definition is stored?
This is the beta bios for z97 anniversary with fixed s4,s5 deep sleeps from asrock after i point to them the bug.
change log:
Provides selectable 1) option for echi hand off 2) option for xchi hand off 3) option for Link Training Timeout 4) option for boot performance mode
Fixed the deep s4,s5
Z97 Anniversary L2.16.zip (5.43 MB)
I have found an interesting bug. In Haswell and Sky-lake motherboards.
I can recreate it with multiple motherboards and maybe it exist until now in Ry-zen motherboards.
In the asrock forum there exist multiple threads.
After you put the pc in s4 (hibernation or hybrid shutdown) many times the CPU sensor stuck and
then if u restore the pc from hibernation the CPU temperature is stuck in the same temp.(the last temp when u put the CPU to s4)
This bug is well known but they cant recreate it because you need specific options.
To recreate this bug u need only the motherboard to work in uefi mode with cms on.
In pure uefi (CMS off) and in pure legacy mode this bug don’t exist.
Sorry @boombastik - I don’t know what audio thing you are asking about.
I checked the mod BIOS you sent me to inspect, it’s bad mod, I would not recommend flashing it due to missing pad file before the last (actual) microcode module.
It might be OK, but without flash programmer in your hand, and backup made and confirmed valid and ready to reprogram I would not test that BIOS. Every time I’ve seen someone flash BIOS like that = brick
Here is correctly modified BIOS - (BIOS was brick, do not use, link nulled! See post #26)
http://s000.tinyupload.com/index.php?file_id=NULL
@Lost_N_BIOS
Basically i use the UEFI BIOS Updater v1.74.0.1 with MMTool v5.0.0.7 as mmtool_a4.exe
And it automatically updated the 4 microcodes. This means that it is no so simple.
Thanks for your time again.
Yes, I thought so, UBU is having issue with some BIOS sometimes at microcode updating, I’m not sure why it’s not solved yet. It seems it should be a simple fix to me but I don’t know how to do all that coding.
I do mod by hand, without UBU. I used UEFITool 25 and hex editor (only to fix FIT). I extracted the microcode body you already had, nothing wrong there, replaced body with V25 UEFITool in new stock BIOS copy, then saved and then fixed FIT via hex, reinserted with UEFITool 25 - done
And you’re welcome, happy to help save you from possible bad BIOS flash
I flashed it and now it does not do anything. I clear the bios also. So i think that i need a new bios chip. 
I was unlucky
You flashed the file I sent, or the other one I said was probably not good? 
No matter which file, sorry to hear this!
You don’t need new BIOS chip, only flash programmer like CH341A and SOIC8 test clip w/ cable to fix this. Here is cheap examples on ebay, shipped slow (3-5 weeks to arrive), but you can find other sellers on ebay that may ship faster, or are closer to you etc.
You can also find these items at other places like Newegg or Amazon type places, faster shipped or closer to you always means more $$
https://www.ebay.com/itm/201316582787
https://www.ebay.com/itm/382486015977
Sorry, I forgot your BIOS is in a socket. You don’t need the SOIC8 test clip cable above it’s for BIOS soldered to board type (maybe pickup anyway for future boards), you only need the flash programmer and this U Type Flat IC Chip extractor
https://www.ebay.com/itm/123167465408
@Lost_N_BIOS i removed the bios and i program it with ch314a the latest stable version( i fully erased it an then program it , the bios version 2.10) but it does not do anything. Any tip?
Maybe i need to change something in the bios that download it from as-rock site? I simply reprogram it with the bios that i download without change anything.
The only message before reprogram it writes: the uper range will be ignored.
Btw i have flashed yours in the fixed folder.
Maybe i need to change the downloaded bios to .bin before the reprogram?
