Contribute
Register

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

tholstro said:
Some developments... Without the PowerColor VGA - AXRX 580 8GBD5-3DH/OC plugged in I don't get the phantom display. Tried with the LG UF 5k along side several other HDMI monitors I have and all behaves as it should.

Seems like that confirms the phantom display problem is caused by the video card. Do you agree?? I still have 3 weeks to return the card...
Click to expand...
Ok, finally a ray of hope! Several comments:
  • What happens if you plug that card into PCIe Slot 1 instead of Slot 2? What is currently plugged into Slot 1?
  • Have you tried the Board ID fixes per the screenshot in my previous post?
  • Does enabling CSM in BIOS make any difference?
If none of the above makes any difference, then it's time to consider a replacement RX 580. It's good to choose one of Apple's recommended cards such as the Sapphire Pulse (not Nitro). My MSI Gaming X has been flawless as well.

Still no luck with iMessages / FaceTime. Tried a brand new account and when prompted during the first login for iCloud it errors. After logging in with that new account I can then go to System Preferences, iCloud and get logged in just fine. Then when I launch iMessage I surprisingly get another authentication dialog, enter credentials and then the activation failure...
Click to expand...
I noticed that in RtVariables --> MLB you have copied the Serial Number instead of the Board Serial Number. So it may be worth trying this:
  1. Log out of iCloud again.
  2. Copy Board Serial Number from SMBIOS to RtVariables: MLB.
  3. Ensure that SMBIOS: SmUUID has a value (there's a "Generate New" button just under it in case the field is empty).
  4. Ensure that the SMBIOS: Trust checkbox is checked-ON.
  5. After making changes, reboot.
  6. Log in to iCloud.
  7. Then try logging in to Messages and FaceTime.
 
CaseySJ said:
Ok, finally a ray of hope! Several comments:
  • What happens if you plug that card into PCIe Slot 1 instead of Slot 2? What is currently plugged into Slot 1?
  • Have you tried the Board ID fixes per the screenshot in my previous post?
  • Does enabling CSM in BIOS make any difference?
If none of the above makes any difference, then it's time to consider a replacement RX 580. It's good to choose one of Apple's recommended cards such as the Sapphire Pulse (not Nitro). My MSI Gaming X has been flawless as well.


I noticed that in RtVariables --> MLB you have copied the Serial Number instead of the Board Serial Number. So it may be worth trying this:
  1. Log out of iCloud again.
  2. Copy Board Serial Number from SMBIOS to RtVariables: MLB.
  3. Ensure that SMBIOS: SmUUID has a value (there's a "Generate New" button just under it in case the field is empty).
  4. Ensure that the SMBIOS: Trust checkbox is checked-ON.
  5. After making changes, reboot.
  6. Log in to iCloud.
  7. Then try logging in to Messages and FaceTime.
Click to expand...
OK... I believe we're done!

I didn't have anything in slot 1. Put the RX 580 in there and the phantom monitor issue came back. Totally missed your Board ID recommendations! Haven't tried them but had already come to the realization that I'm better off with a video card recommended by Apple. And with an unsupported card I might be just one OS release from more complications!

After my last post I decided to dig in and follow the Idiots Guide to iMessage. While the list of tasks is long it was well worth it. I was also experiencing strangeness with a handful of App Store apps. Some apps worked and others error'd with the "Application is Damaged" message. After following @P1LGRIM's guide iMessage, FaceTime and the problematic App Store apps all started working.

Saturday well spent thanks to you Casey. Add me to the long list of people you've dug from that "Help me" hole!

Tom
 
Last edited:
CaseySJ said:
Yes you can certainly install High Sierra on this system, but I believe you may need to select a Kaby Lake platform ID (0x59120000) and device-id (0x5912). Hopefully someone running High Sierra can chime in.

@avtor, can you share some tips and guidelines for installing High Sierra?
Click to expand...

Absolutely,

@kageman , and @PicLock ,

I'm not sure if you are are starting from scratch with your build or you are moving your previous install to the new hardware. My case was the latter - I moved my install from failed MB. Also I'm running internal graphics, but technically from the installation stand point it does not matter.

What I did was:

1. I haver installed and configured clover on the USB (just that, nothing else) as High Sierra clover done by Unibeast will not work. You will be missing missing OsxAptioFix2Drv-free2000.efi at least and you have to have it there to start your system. Also I wanted to have something like generic boot loader just in case. So I was using this guide for clover-

@CaseySJ , please let me know if I can share a link to other website.

or you can use install clover and configure it based on the preinstallation steps as well. In fact I would recommend to compare your USB Clover config to Casey's one before the first boot.

2. Boot into Clover and here you have you High Sierra right after first boot
3. Follow the post installations steps to the t, EXCEPT for one point - Macmini8,1 obviously will not work on High Sierra as it simply did not exist back then, so I was on 14,2 and I left it that way.

I did not have any major issue, those that I had mostly due to something was overlooked. It is very important to complete and follow Post Installation steps TO THE T.

If I were installing High Sierra from scratch I would have gone pretty much the same route - would not be using Unibeast in this specific case, but rather run manually High Sierra's createinstallmedia app and would install Clover on the USB in the end.

Let me know if that makes sense.
 
CaseySJ said:
Buy now or wait: This is always a very subjective decision!

We know that Intel's 10nm process has been delayed repeatedly, but at CES in January Intel stated that its Ice Lake processors are expected at the end of 2019. But I suspect there will be significant product shortages as there were for the 9th Gen, perhaps more so. There are several exciting technologies in Ice Lake such as of course lower TDP, but also up to 10 cores, improved IGPU, and built-in deep learning extensions. Thunderbolt 3 is also expected to be built into the chipset, along with WiFi 6.

Having said that, one of the most significant advantages of building a Hackintosh is modularity. Yes, we have that magical modularity right now. This means that if I cave in to the temptation of Ice Lake, then:
  • I can buy a new motherboard for perhaps $250.
  • I can buy a new Ice Lake processor for perhaps $400.
  • I am done.
Modularity allows me to:
  • Keep my graphics card
  • Keep my PC case
  • Keep my memory
  • Keep my NVMe drives
  • Keep my SATA drives
  • Keep my All-in-One Liquid Cooler
  • Keep my power supply
  • Keep my Thunderbolt, USB, and other peripherals
So for a total extra cost of $650 I can swap-in to Ice Lake. Moreover, I can even sell the old CPU and motherboard to offset some costs.

Modularity, in other words, changes the thought process for "buy now or wait". It allows us to do both.
Click to expand...


Modularity is also what Apple lacks today :)
 
avtor said:
Absolutely,

@kageman , and @PicLock ,

I'm not sure if you are are starting from scratch with your build or you are moving your previous install to the new hardware. My case was the latter - I moved my install from failed MB. Also I'm running internal graphics, but technically from the installation stand point it does not matter.

What I did was:

1. I haver installed and configured clover on the USB (just that, nothing else) as High Sierra clover done by Unibeast will not work. You will be missing missing OsxAptioFix2Drv-free2000.efi at least and you have to have it there to start your system. Also I wanted to have something like generic boot loader just in case. So I was using this guide for clover-

@CaseySJ , please let me know if I can share a link to other website.

or you can use install clover and configure it based on the preinstallation steps as well. In fact I would recommend to compare your USB Clover config to Casey's one before the first boot.

2. Boot into Clover and here you have you High Sierra right after first boot
3. Follow the post installations steps to the t, EXCEPT for one point - Macmini8,1 obviously will not work on High Sierra as it simply did not exist back then, so I was on 14,2 and I left it that way.

I did not have any major issue, those that I had mostly due to something was overlooked. It is very important to complete and follow Post Installation steps TO THE T.

If I were installing High Sierra from scratch I would have gone pretty much the same route - would not be using Unibeast in this specific case, but rather run manually High Sierra's createinstallmedia app and would install Clover on the USB in the end.

Let me know if that makes sense.
Click to expand...

​
Reckond I have to read up a bit more on installing with Clover as I only know how to install with UniBeast. I will be installing High Sierra from scratch. I might just sell my Nvidia GTX 1080 Ti and go with AMD instead for Mojave. Thanks Avtor.
 
Hi @CaseySJ. Thanks for this great build guide and thread.

Do you know if anyone has gotten an eGPU working with your build?

I have a Z370 Gaming 7 with a Titan Ridge (also have a Alpine Ridge), XFX VEGA 64 8GB LQ as the discrete with UHD 630 as headless for QuickSync, and a Sapphire Rx 580 Nitro+ 8GB in a Razor Core X eGPU. The Vega is driving 2 x 4K@60Hz displays, and the RX 580 a HD@60Hz display. I'm using the 18,3 SMBIOS. Should I be trying a different one? Everything works fine under Arch Linux and Windows 10.

I have a OCW TB3 10GBe adapter that I can hot plug on either ports on the TR. The eGPU's RX 580 is recognized, but the drivers fail to load. I did have it once display on the eGPU, but since then nothing. Sometimes the eGPU display looks like it it is going to work, but doesn't.
If I remove the eGPU, macOS tells me that I didn't' eject the eGPU properly. When I conect the eGPU and at the same time constantly hitting/refreshing by keying CMD + R whilst in the PCI section of the System Information tool, I can see that the RX 580 connects the drivers then disconnects.

I guess to save me time, it would be good to know if anyone has got a eGPU working with your build, and if so any pointers would be great.
 
mcljot said:
Hi Casey, in the spirit of giving back to your superb build guide, I wrote up a short procedure to run the NZXT liquidctl script automatically at login. Maybe it's a bit too specific and edge-case but please feel free to add it to your guide if you think it's useful.

Configuring the liquidctl script to run on startup (for all users). Copy the liquidctl.sh script to a shared directory and create a short .plist file in /Library/LaunchAgents:

  1. Copy the liquidctl.sh script to /Users/Shared.
  2. Create a new liquidctl.plist file in /Library/LaunchAgents.
  3. Edit the liquidctl.plist file with admin privileges and paste in the example below. Either use vi or a text editor with admin privileges: sudo vi /Library/LaunchAgents/liquidctl.plist. Remove the #Explanation: sections from the example below:
<?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> <plist version="1.0"> <dict> <key>Label</key> <string>liquidctl.sh</string> #Explanation: This is a verbose label for your login script. Name it something unique and identifiable. I just named mine the same as the script itself <key>ProgramArguments</key> <array> <string>/Users/Shared/liquidctl.sh</string> #Explanation: This is the absolute path to the liquidctl.sh script </array> <key>RunAtLoad</key> <true/> </dict> </plist>

4. Change ownership of the liquidctl.plist file so that root owns it:
sudo chown root /Library/LaunchAgents/liquidctl.plist
5. Check that you've done everything correctly by attempting to load the .plist manually:
sudo launchctl load /Library/LaunchAgents/liquidctl.plist
If this command returns anything other than a blank line, go back and check your steps.

Done! Now, even if another operating system has messed with your NZXT Kraken, every time you boot back into Mac OS your Kraken will configure itself automatically no matter what settings might try to persist from other operating system instances.

Caveat: the liquidctl.sh script unloads IOUSBHostHIDDevice. This is necessary to avoid [Errno 13] Access denied (insufficient permissions) that you might encounter when running liquidctl commands. However, this can result in temporary disabling/resetting of your USB devices. My Logitech wireless receiver sometimes loses connection with my wireless mouse or keyboard after running the script. Simply unplug and plug back in the unifying receiver. I don't know what happens with other USB devices, but I suspect unplugging and plugging them back in will work to bring them back online.
Click to expand...

@mcljot Running "sudo launchctl load /Library/LaunchAgents/liquidctl.plist" doesn't return an error but doesn't appear to be working either. Running from terminal getting the "(kernel) Can't remove kext com.apple.driver.usb.IOUSBHostHIDDevice; services failed to terminate - 0xdc008018. Failed to unload com.apple.driver.usb.IOUSBHostHIDDevice - (libkern/kext) kext is in use or retained (cannot unload)." followed by Access Denied for all the liquidctl commands.

Ramp up the CPU and temps rise but radiator fans don't spin up.

Do you know the cause or solution for this?

Thanks!

Tom
 
jb007 said:
Hi @CaseySJ. Thanks for this great build guide and thread.

Do you know if anyone has gotten an eGPU working with your build?

I have a Z370 Gaming 7 with a Titan Ridge (also have a Alpine Ridge), XFX VEGA 64 8GB LQ as the discrete with UHD 630 as headless for QuickSync, and a Sapphire Rx 580 Nitro+ 8GB in a Razor Core X eGPU. The Vega is driving 2 x 4K@60Hz displays, and the RX 580 a HD@60Hz display. I'm using the 18,3 SMBIOS. Should I be trying a different one? Everything works fine under Arch Linux and Windows 10.

I have a OCW TB3 10GBe adapter that I can hot plug on either ports on the TR. The eGPU's RX 580 is recognized, but the drivers fail to load. I did have it once display on the eGPU, but since then nothing. Sometimes the eGPU display looks like it it is going to work, but doesn't.
If I remove the eGPU, macOS tells me that I didn't' eject the eGPU properly. When I conect the eGPU and at the same time constantly hitting/refreshing by keying CMD + R whilst in the PCI section of the System Information tool, I can see that the RX 580 connects the drivers then disconnects.

I guess to save me time, it would be good to know if anyone has got a eGPU working with your build, and if so any pointers would be great.
Click to expand...
One of the many benefits of having PCIe slots is that we don't need an eGPU! :) Local installation of a graphics card reduces latency and maximizes speed because it can take advantage of 8 or 16 dedicated lanes. An eGPU is limited to 4 (shared) lanes.

Nevertheless, if anyone has tried or is using an eGPU with a Hackintosh, please let us know...
 
New LG 5K2K widescreen display w/HDR @CaseySJ is the TH3 input on this monitor of any use with the Designare TH3 ? I believe that UHD 630 can't output 5K/60 or am I incorrect ?
www.lg.com

LG 34BK95U-W: 34'' Class 21:9 UltraWide® 5K2K Nano IPS LED Monitor with HDR 600 (34'' Diagonal) | LG USA Business

Get information on the LG 34'' Class 21:9 UltraWide® 5K2K Nano IPS LED Monitor with HDR 600 (34'' Diagonal) . Find pictures, reviews, and tech specs for the LG 34BK95U-W
www.lg.com www.lg.com

394787
 
Last edited:
You must log in or register to reply here.
Copyright © 2010 - 2024 tonymacx86 LLC
Back
Top