Contribute
Register

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

Have you tried booting with Clover?

no not at this point, I’m still trying to pinpoint the root causes for these wakeups. Everything else has worked very well on OpenCore for me. I believe I just found the culprit of the one that is actually causing phantom wakeups

AppleACPIPlatformPower Wake reason: PEG1 PEGP PEG2 PEGP PXSX RP03 PXSX UPSB DSB0 NHI0 DSB1 DSB2 XHC5 DSB4 RP06 PXSX RP07 PXSX RP08 PXSX PXSX RP10 PXSX RP11 PXSX R

.... Find my Mac - do you know what would be making this happen?
 
@jb007 @ralphonz

Just moments ago I decided to update the Thunderbolt firmware on my main system. Up to now I've done all my tinkering on the test bench, but after @jb007's procedure documented here, I decided to see how easy it would be to (a) first read the chip three times in a row, and then (b) flash it with the modified firmware.

To my delight, the procedure worked marvelously.

Three perfect reads were done and I felt very comfortable to proceed with the firmware update. And that too was completely successful.

So now both of my Designare Z390 systems have Thunderbolt Bus and are operating at the full 40 Gbps link speed.

View attachment 458522
That's good news! On a technical level, all that is happening is:
  • you are removing the 3.3V current burden from the Raspberry Pi and using the motherboards 3.3V, it still works because the two devices are joined by a common ground (pin 4 on chip). No resistors or caps needed.
  • also, the motherboard is setting up the chip select lines most probably ( I really should put my logic probe or oscilloscope on the chip to 100% prove this) into a high impedance state so the RP can now toggle these pins to do its stuff.
Still, it would be good if we could find a way to say under a live Linux boot disk to program these chips. Something I will try and pursue.

Can you tell me if you have successfully got either the Titan or Alpine ridge working in conjunction with the onboard TB? Which is the best in your opinion as I have both cards including a second TR that is similar to your ASUS card you damaged i.e. it is dead, but firmware can be read and written okay.
 
no not at this point, I’m still trying to pinpoint the root causes for these wakeups. Everything else has worked very well on OpenCore for me. I believe I just found the culprit of the one that is actually causing phantom wakeups

AppleACPIPlatformPower Wake reason: PEG1 PEGP PEG2 PEGP PXSX RP03 PXSX UPSB DSB0 NHI0 DSB1 DSB2 XHC5 DSB4 RP06 PXSX RP07 PXSX RP08 PXSX PXSX RP10 PXSX RP11 PXSX R

.... Find my Mac - do you know what would be making this happen?
These are my Location Services Settings:
Screen Shot 2020-03-30 at 4.14.48 PM.png
Find my Mac has always been enabled on both of my systems. No sleep/wake issue.

I suggested Clover for the purposes of reducing/eliminating variations between your system and mine. I'm not suggesting to switch over to it, but simply to use it for testing.
 
That's good news! On a technical level, all that is happening is:
  • you are removing the 3.3V current burden from the Raspberry Pi and using the motherboards 3.3V, it still works because the two devices are joined by a common ground (pin 4 on chip). No resistors or caps needed.
  • also, the motherboard is setting up the chip select lines most probably ( I really should put my logic probe or oscilloscope on the chip to 100% prove this) into a high impedance state so the RP can now toggle these pins to do its stuff.
Still, it would be good if we could find a way to say under a live Linux boot disk to program these chips. Something I will try and pursue.

Can you tell me if you have successfully got either the Titan or Alpine ridge working in conjunction with the onboard TB? Which is the best in your opinion as I have both cards including a second TR that is similar to your ASUS card you damaged i.e. it is dead, but firmware can be read and written okay.
I have the GC-Titan Ridge (not the GC-Alpine Ridge) and I've found that it works reasonably well with the modified NVM 23. Some issues:
  • USB-C devices must be connected at boot. The USB controller does not seem to be powered on unless a device is pre-attached.
  • There might be some issues switching Thunderbolt devices between the two ports. One of the ports may work more reliably and consistently than the other.
But with the flashed firmware on the Designare itself, there are simply no discernible issues. It's just amazing how well the on-board flashed controller works. But we're still waiting to see if those dreaded Antelope Audio devices will work.
 
These are my Location Services Settings:
View attachment 458524
Find my Mac has always been enabled on both of my systems. No sleep/wake issue.

I suggested Clover for the purposes of reducing/eliminating variations between your system and mine. I'm not suggesting to switch over to it, but simply to use it for testing.

I’m getting another SSD in mail on Wednesday. I think I’m going to try testing on that with clover to see if I have same issues. Is it at all possible since we have same build to use your EFI (removed SSID / SERIAL / ETC) so that I can be certain our folders match with all settings to test properly? If not I completely understand
 
I’m getting another SSD in mail on Wednesday. I think I’m going to try testing on that with clover to see if I have same issues. Is it at all possible since we have same build to use your EFI (removed SSID / SERIAL / ETC) so that I can be certain our folders match with all settings to test properly? If not I completely understand
There's no need for another SSD because you can do this:
  • Grab a USB flash disk. Even the one with OpenCore on it.
  • Mount the EFI partition.
  • If OpenCore is installed, rename the EFI folder to EFI-OC temporarily.
  • Now download Clover 5107 (download the .pkg file) and run the Clover installer.
  • Point it to the USB flash disk.
  • Follow the instructions in the Catalina 10.15.4 Mini-Guide, Step 1, but omit bullet #3 (because we don't need to install macOS). Clover installer will create a new EFI folder (not a new EFI partition).
  • You can also copy your serial number and SmUUID from OpenCore config.plist into the Clover config.plist.
  • Then boot from the USB flash disk (F12 at BIOS splash screen).
The same EFI partition can be used to boot either Clover or OpenCore. To switch back to OpenCore, just rename the new EFI folder to EFI-Clover. And rename EFI-OC back to EFI.
 
I have the GC-Titan Ridge (not the GC-Alpine Ridge) and I've found that it works reasonably well with the modified NVM 23. Some issues:
  • USB-C devices must be connected at boot. The USB controller does not seem to be powered on unless a device is pre-attached.
  • There might be some issues switching Thunderbolt devices between the two ports. One of the ports may work more reliably and consistently than the other.
But with the flashed firmware on the Designare itself, there are simply no discernible issues. It's just amazing how well the on-board flashed controller works. But we're still waiting to see if those dreaded Antelope Audio devices will work.
Crossing fingers that someone with an Antelope Audio device will be able to test with the flashed Designare firmware. I have a feeling this is going to be a game-changer for many of the audio guys and gals out there. I'm sure there are users of this Designare that have or know others that might be able to borrow to test this combo out?
 
There's no need for another SSD because you can do this:
  • Grab a USB flash disk. Even the one with OpenCore on it.
  • Mount the EFI partition.
  • If OpenCore is installed, rename the EFI folder to EFI-OC temporarily.
  • Now download Clover 5107 (download the .pkg file) and run the Clover installer.
  • Point it to the USB flash disk.
  • Follow the instructions in the Catalina 10.15.4 Mini-Guide, Step 1, but omit bullet #3 (because we don't need to install macOS). Clover installer will create a new EFI folder (not a new EFI partition).
  • You can also copy your serial number and SmUUID from OpenCore config.plist into the Clover config.plist.
  • Then boot from the USB flash disk (F12 at BIOS splash screen).
The same EFI partition can be used to boot either Clover or OpenCore. To switch back to OpenCore, just rename the new EFI folder to EFI-Clover. And rename EFI-OC back to EFI.

I've read some places that by doing that, clover creates some junk files or data that has to or should be removed when going to OpenCore -- is that not the case?

also if I was to do this, do I have to follow the MSR 0x32 as well, or I can skip that since I already installed?

Also, will your setup allow me to have a second separate SSD for Windows, or will it cause issues?
 
I've read some places that by doing that, clover creates some junk files or data that has to or should be removed when going to OpenCore -- is that not the case?

also if I was to do this, do I have to follow the MSR 0x32 as well, or I can skip that since I already installed?

Also, will your setup allow me to have a second separate SSD for Windows, or will it cause issues?
If you follow the Catalina 10.15.4 guide (Step 1) that shows what Clover options to check on, then no extra files will be installed. Only the EFI folder in the EFI partition will be created.

MSR 0xE2 has to be unlocked for the 10.15.4 guide. My test bench is unlocked.

The setup I'm suggesting is on a USB flash disk and is temporary. It will have no impact on any future disk or OS configuration.

Having said this, I'm not suggesting that Clover will fix your sleep/wake issues. Instead, what I'm getting at is this:
  • You have a system whose sleep/wake functionality does not work.
  • I have a system whose sleep/wake functionality works remarkably well.
  • So if we consider my test bench to be a "reference design" or "baseline", then doesn't it make sense to compare your system against a known working baseline?
  • And then begin to systematically look for differences?
 
If you follow the Catalina 10.15.4 guide (Step 1) that shows what Clover options to check on, then no extra files will be installed. Only the EFI folder in the EFI partition will be created.

MSR 0xE2 has to be unlocked for the 10.15.4 guide. My test bench is unlocked.

The setup I'm suggesting is on a USB flash disk and is temporary. It will have no impact on any future disk or OS configuration.

Having said this, I'm not suggesting that Clover will fix your sleep/wake issues. Instead, what I'm getting at is this:
  • You have a system whose sleep/wake functionality does not work.
  • I have a system whose sleep/wake functionality works remarkably well.
  • So if we consider my test bench to be a "reference design" or "baseline", then doesn't it make sense to compare your system against a known working baseline?
  • And then begin to systematically look for differences?

Okay let me create a USB and ill come back when done :)
 
Back
Top