Contribute
Register

Ongoing Progress - Big Sur on Gigabyte B550 Vision D - AMD Ryzen 7 3700X - Thunderbolt 3

Status
Not open for further replies.
wow.. this thread is getting so much momentum.. hope AMD is not coming out with a new chipset for the high-end Zen3!
I guess @CaseySJ will lead us all the way to 3990X!
We're expecting AMD to announce the new Ryzen 5000 series of desktop CPUs next week on October 8. Recent leaks suggest they might be 25% or so faster in single core performance compared to the 3000 series, which would make them faster than Intel's i9-9900K and i9-10900K.

However, we are not expecting a new chipset. This is of course good for existing motherboard owners because AMD does not change sockets as wildly as does Intel.
 
This thread and the z490 vision D thread are where all the excitement is for me right now on tonymac. The Thunderbolt connection and high core count possibilities are big leaps for people wanting to build a system that is a little more future proof. I like the Z390 thread too (also Thunderbolt), but that board is a little older now (but still awesome). I'm looking forward to getting one of these boards and trying it out. So far, the quieter noise on the B550 is a pretty nice bonus. Anyway, thanks CaseySJ and keep it going guys!
 
This thread and the z490 vision D thread are where all the excitement is for me right now on tonymac. The Thunderbolt connection and high core count possibilities are big leaps for people wanting to build a system that is a little more future proof. I like the Z390 thread too (also Thunderbolt), but that board is a little older now (but still awesome). I'm looking forward to getting one of these boards and trying it out. So far, the quieter noise on the B550 is a pretty nice bonus. Anyway, thanks CaseySJ and keep it going guys!
I am organizing my thoughts now for a proper User Builds thread for the B550 Vision D. Coming soon! :)

I've been using this system as my daily driver for the past few days with everything working (including sleep, wake, WiFi, Bluetooth) -- with a Fenvi FV-T919 PCIe card and not the BCM94360NG. However the dedicated USB 2.0 ports have to be disabled. Hence the build guide will recommend components that avoid those ports as much as possible. It is really liberating to be able to keep the system running all day long and seeing that it sleeps and wakes properly every time.

I also find that the system (with Big Sur public beta) is more stable when AMDRyzenCPUPowerManagement is disabled. The fans remain more stable as well, as opposed to revving high and low in sporadic bursts. I do expect AMDRyzenCPUPowerManagement to improve once Big Sur is released.
 
Last edited:
I'm looking forward for the full guide/ This is so inline with where Ii want to go next. More cores for less. I use Logic Pro to make music and would love the extra cores for less mullah. Going to plan carefully and watch the development in the next few weeks. Lots of eyes on Zen3. I already built my Z390 Designare with @CaseySJ's excellent guide. I cant wait to build my next with this new Ryzen build! Thanks for the contribution @CaseySJ!! You are making history! :)
 
Formal build guide is now available here:
 
Last edited:
** How to Use Bluetooth and WiFi with Sleep and Wake **


The Gigabyte B550 Vision D comes with an Intel AX200 WiFi/BT module. Because the Bluetooth section connects itself to an internal USB 2.0 port at PO10, the system will not wake-from-sleep when this port is enabled. Even if we replace the Intel AX200 with a Broadcom BCM94360NG, it too will connect to PO10 and once again no wake-from-sleep.

If we physically remove the WiFi/BT module from the M.2 slot and instead install a Fenvi FV-T919 or similar PCIe card, we still need to connect the 9-pin Bluetooth data cable to a USB port. Unfortunately, if we connect that 9-pin cable to one of the internal USB 2 headers, the system will again refuse to wake from sleep.

However, if we:
  • Install a USB 3.0 PCIe card such as the Inateck KT4004
  • Install Fenvi FV-T919 or similar card
  • Physically remove BCM94360NG M.2 WiFi/BT module (or keep the Intel AX200 plugged in)
  • Connect Bluetooth data cable from Fenvi to the Inateck using a 9-pin to Type A cable
  • Use a USB SSDT and kext that disables USB 2.0 ports PO9, PO10, PO11 (simply use EFI zip file in this post)
View attachment 490053 View attachment 490054 View attachment 490055
...then we get full functionality. Namely:
  • Sleep and wake work every time.
  • WiFi and Bluetooth are fully enabled.
  • Although we lose on-board USB 2 ports, we gain extra USB ports on the Inateck
This does mean that all three PCIe slots are occupied.

View attachment 490047
I know you have fixed sleep by disabling the 2.0 hub, but what if you go back and make AppleUSB20InternalHub attach instead of AppleUSB20Hub? I'm saying the default driver may be incompatible and causing sleep/wake issues and you may need to choose the correct driver for working 2.0 hub and sleep/wake.
/S*/L*/E*/IOUSBHostFamily.kext/C*/AppleUSBHub.kext/C*/Info.plist has the configuration of USB Hub drivers.
스크린샷 2020-10-16 오전 8.08.49.png

USB hubs will normally default to AppleUSB20Hub because their bProperties match as non-Apple USB hub or external USB hub. Yours also matches as AppleUSB20Hub in your IORegistry.
I had sleep issues with a HackBook when the AppleUSB20Hub was attached. I did not have any sleep/wake problems when I correctly configured USB and AppleUSB20InternalIntelHub was attached. Below is my USBPorts.kext/C*/Info.plistView attachment 491958
If you look into /S*/L*/E*/IOUSBHostFamily.kext/C*/AppleUSBHostPlatformProperties.kext/C*/Info.plist, you see a list of legacy Macs' USB port configuration.
스크린샷 2020-10-16 오전 7.50.49.png

Some models appear to have internal hubs and their specific driver is AppleUSB20InternalHub.
The way a USB2 device is recognized on EHC is EHCx-InternalIntelHub-ThumbDrive (e.g. [iMac13,1-EHC1]-[iMac13,1-InternalHub-EHC1's PRT1]-[a thumb drive is connected to PRT1])
or EHCx-InternalIntelHub-InternalHub-ThumbDrive (e.g. [iMac13,1-EHC1]-[iMac13,1-InternalHub-EHC1's PRT8]-[iMac13,1-InternalHub-EHC1-InternalHub's PRT1 is connected to PRT8(locationID 488,112,128=0x1D1'8'0000)]-[a thumb drive is connected to a port under iMac13,1-InternalHub-EHC1-InternalHub's PRT1 hub]). Most hacks fall in the first category.
I think your board may fall in the second category. Not exactly the same because your USB controller is not EHC but XHC, but the registry can be something like AppleUSBXHCIABCD-AppleUSB20InternalHub-PO9_,PO10,PO11 if you inject IOProviderClass.
View attachment 491962
If you use a kext to inject USB ports and driver information, you also need to have all the ports specified under IOProviderMergeProperties. I'm not sure if you need to do the same if only driver information is being injected into one controller.
I'm just curious if this would work. I wish I had a similar board so I can test, but I don't have one lol.
 

Attachments

  • 스크린샷 2020-10-16 오전 7.09.21.png
    스크린샷 2020-10-16 오전 7.09.21.png
    278.7 KB · Views: 64
@whatnameisit,

Thank you for the detailed description and suggestion! I’ll take a deeper look into this. It would certainly be wonderful to solve the root problem.
 
I think the 4 USB ports should be attached to PO14. I don't see PO14 on Hackintool, so please check in the SSDT if PO14 is enabled.
Also,
1. USBInjectAll does not support AMD systems.
2. I helped headkaze add support for EHCxInternalIntelHub-PortsAttachedHub by providing him debug logs on laptops with EHCx, but not for EHCxInternalIntelHub-InternalHub-PortsAttachedHub or XHC-InternalHub-PortsAttachedToHub because I didn't have such a machine, still don't. So I don't think Hackintool will show ports attached to PO14. Still, you or your friend should see PO14 on Hackintool.
3. If you don't see PO14 on IOReg, the 2.0 hub isn't being recognized, and none of the 4 ports will be recognized on macOS. You first need to make macOS see PO14 with the SSDT.
 
Status
Not open for further replies.
Back
Top