[SUCCESS] Gigabyte Designare Z390 (Thunderbolt 3) + i7-9700K + AMD RX 580

[hussar]

Good question. The Designare, unfortunately, has but a single USB 2 header. We can very effectively negate this disadvantage by using the following 9-pin USB splitter.
https://www.amazon.com/dp/B07FKJ1CW9/?tag=tonymacx86com-20
This model is particularly effective because if you look at the highlighted connector, you'll see that it uses 8 pins, four on each row. Each row defines one USB port (+5V, GND, DATA+, DATA-). One row corresponds to USB port HS11, and the other to HS12.

And you can see that there are 2 splitters in the picture (blue printed circuit boards or PCBs). Each splitter accommodates 2 USB connections for a total of 4.

Your internal connections (Bluetooth data and CPU cooler, for example) would be connected to the same blue PCB. And your external USB ports located on the front of your PC case would be connected to the other blue PCB.

Each PCB provides a 2-port USB hub so even though there are 4 USB connections, they will still count as only two (HS11 and HS12).

Let's say we connect the internal devices (Bluetooth and CPU cooler) to HS11. We can set its type to 255 (Internal).

And let's say we connect the external USB ports (on PC case) to HS12. We will set its type to 0.

So using this particular splitter provides the level of control we need.

View attachment 408877

oh this is so good!
i have a NZXT 500i
and a
Corsair H115i Pro
they both need the usb header on the z390!

French Amazon doesn't have the RINGBUU.
But I did find something similar in case anyone is interested: BGNING 9-pin USB.
Find it here: https://www.amazon.fr/gp/product/B079CBQ11Q/?tag=tonymacx8607-21
\
Can't wait to install it!

Though do I have to modify anything in the config file to get it to work?

 

[CaseySJ]

oh this is so good!
i have a NZXT 500i
and a
Corsair H115i Pro
they both need the usb header on the z390!

French Amazon doesn't have the RINGBUU.
But I did find something similar in case anyone is interested: BGNING 9-pin USB.
Find it here: https://www.amazon.fr/gp/product/B079CBQ11Q/?tag=tonymacx8607-21
\
Can't wait to install it!

Though do I have to modify anything in the config file to get it to work?

When you install this device and connect your NZXT and Corsair coolers, please do this:

• Download and run IORegistryExplorer
• Then select File --> Save As... and upload the IOReg file

Based on this I can check/update the USB SSDT for your system.

 

[CaseySJ]

Dear all, I switched to the Z390 as I was having issues with TB3 AIC + IGP, and I need QSV to work.

I have tried many described methods to activating the 9900K HD630 in headless mode, besides the Vega64.

Closest I got is it showing up in many places, and a flat green line in Intel Power Gadget, that doesn't activate while testing BruceX.

Your iGPU is perfectly configured and activated in headless mode (platform ID 0x3E980003). However:

• You have Lilu 1.3.1. Please update to 1.3.6.

Besides that, one of the LAN ports isn't showing, I believe only the I211 does, according to DPCIManager.

• You have IntelMausiEthernet 2.3.0. Please update to 2.4.0.

I tested changing the Intel WiFI chip for the BCM94352Z (WiFi/BT) which I have used in multiple builds, tried two chips, but neither get recognised in Windows as well. Does the Designare have a WiFi whitelist? I'll have to test and see whether th Intel chip even works in Windows. I'm aware that with this guide the M.2 WiFi slot gets disabled, hence my Windows testing.

Unfortunately this motherboard has an Intel CNVi slot. It is not pin-compatible with BCM94352Z (M.2 Key-E) and even if it were, the Gigabyte firmware will only support Intel CNVi devices in that slot. For WiFi and Bluetooth we must use a PCIe x1 card such as Fenvi FV-T919.

 

[AlBeast]

Thanks CaseySJ, much appreciated Didn't know the CNVi slot was so specific. I have a PCI-E 1x with a working chip laying around, so I'll resort to the correct SSDT to disable the CNVi slot, after I put the unit in.

The LAN port is up now, I was a bit stupified when I saw that version mistake. Just started trusting Kext Updater a few days ago.

What do you reckon is the reason I can't see the green graph on Intel's gadget? It stays flat. I think I have virtualsmc and the like set up correctly?

Is it normal BruceX 5K test ends around 14 seconds both with ProRes422 and H.264, with perhaps 15.5 sec on 4444XQ? it seems to me the difference would have to be larger between a ~20MB and ~670MB file, and the first two shouldn't be identical? Perhaps you have no experience with this test and the QuickSync specific, but hopefully someone else does then. I'm not sure if it's related to that IGP graph flatlining.

 

[NoiseCoalition]

How often does everyone update their Clover EFI bootloader? Can you update via https://sourceforge.net/projects/cloverefiboot/ ? Previous to my latest UniBeast build, I had a vanilla build which I kept semi updated. I'm just unsure if the same update steps can be used with my current UniBeast build.

 

[CaseySJ]

...
What do you reckon is the reason I can't see the green graph on Intel's gadget? It stays flat. I think I have virtualsmc and the like set up correctly?

Some additional comments after a closer look:

• In your ACPI section there is a patched named "ACPI PATCH". What is that for?
• In SMBIOS section, the Trust checkbox should be turned on.
• In Kernel and Kext Patches, all 4 of the USB port limit patches should be disabled. The bottom 2 are still enabled.

I've made these changes in the attached config.plist. I've checked-on the "disabled" flag for "ACPI PATCH", but if this patch is absolutely necessary, please uncheck that column.

Additionally, if you're running 10.14.5 you may delete SSDT-Z390-VEGA64.aml from CLOVER/ACPI/patched.

When you run the BruceX test, maybe the Vega is doing 100% of the work?

Is it normal BruceX 5K test ends around 14 seconds both with ProRes422 and H.264, with perhaps 15.5 sec on 4444XQ? it seems to me the difference would have to be larger between a ~20MB and ~670MB file, and the first two shouldn't be identical? Perhaps you have no experience with this test and the QuickSync specific, but hopefully someone else does then. I'm not sure if it's related to that IGP graph flatlining.

Unfortunately my video editing adventures are limited to iMovie! So if anyone reading this can offer any input, please do so.

 

[CaseySJ]

How often does everyone update their Clover EFI bootloader? Can you update via https://sourceforge.net/projects/cloverefiboot/ ? Previous to my latest Unibeast build, I had a vanilla build which I kept semi updated. I'm just unsure if the same update steps can be used with my current Unibeast build.

I tend to update Clover concurrently with each macOS update. I use the same link you provided and it's okay to run this installer on top of an existing UniBeast or MultiBeast installation of Clover.

 

[3Dman]

I have noticed that when I arrive home my PC wakes up; turns out that proximity wake feature works. I don't know when it started to work. I don't know if it's related to the latest iMac SMBIOS or not.

But this is what I found in pmset log:
2019-05-31 01:31:08 AppWakeReason AppWoke:com.apple.bluetoothd Reasonroximity Wake
2019-05-31 01:31:08 Assertions PID 56(powerd) Created InternalPreventSleep "Assertion to change proximity monitoring state" 00:00:00 id:0x0xd00009d6b [System: PrevIdle SRPrevSleep kCPU]
2019-05-31 01:31:08 Assertions PID 56(powerd) Created InternalPreventSleep "Holding in darkwake for user proximity for up to 600 secs" 00:00:00 id:0x0xd00009d6c [System: PrevIdle SRPrevSleep kCPU]

Very nice to have this feature working on a hackintosh.

 

[bsholbrook]

Please try the attached CLOVER-bsholbrook folder (rename it of course). Changes made:

• Duplicate Aptio Memory fix driver (OsxAptioFix3drv.efi) deleted. Only one memory driver is allowed at a time. OsxAptioFix2Drv-free2000.efi is already present.
• Checked-on "AppleIntelCPUPM" in Clover Configurator --> Kernel and Kext Patches. If your MSR 0xE2 register is locked (which is the default case on Designare Z390), then this should be checked on.
• Deleted non-standard FakeSMC sensor kexts from CLOVER/kexts/Other.
• Deleted framebuffer "Orinoco" from Clover Configurator --> Graphics.
• Set platform ID to 0x3E980003 by specifying the bytes in reverse order: 0300983E.

Appreciate your look at this. I've finally had a chance to swap your recommended Clover folder into place. Post reboot headless mode is still not working. I compared your written details to zipped Clover director and found a few small variances:

• the appleintelcpupm was not checked; I have a Desigare board so I checked that.
• the platform id in the config.plist was 0300923E vs the 0300983E listed above. I've tried both without either working.

Any other thoughts on what may be going on? What other details can I share that would help identify where things go off the rails?

Here is a view of the Intel Power Gadget. You can see the CPU rates jump during H265 playback, and the GFX graph remains flat. Also, you already said you were no familiar with the VDADecoderChecker. So this may not help matters. It says there is a problem. It will normally report SUCCESS when headless mode is operating properly.

 

[CaseySJ]

Appreciate your look at this. I've finally had a chance to swap your recommended Clover folder into place. Post reboot headless mode is still not working. I compared your written details to zipped Clover director and found a few small variances:

• the appleintelcpupm was not checked; I have a Desigare board so I checked that.
• the platform id in the config.plist was 0300923E vs the 0300983E listed above. I've tried both without either working.

Any other thoughts on what may be going on? What other details can I share that would help identify where things go off the rails?

Here is a view of the Intel Power Gadget. You can see the CPU rates jump during H265 playback, and the GFX graph remains flat. Also, you already said you were no familiar with the VDADecoderChecker. So this may not help matters. It says there is a problem. It will normally report SUCCESS when headless mode is operating properly.

View attachment 409096

Can you please run IORegistryExplorer followed by File —> Save As... and upload the saved file? This will tell us whether IGPU is in headless mode. It would also be good to update Lilu, WhateverGreen, and AppleALC (and run Kext Utility to rebuild kernel cache). All three were updated recently.