Contribute
Register

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

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
Hi guys, sorry for disturbing you
I am a beginner, I would like to update Clover versions from 4 to 5.
I find the Mini Guide where there is written to run "Clover 5xxx .pkg installer"
My question is probably silly, but where can I find this ? I have to download ?
Thank you very much for your helping

By the way I have a problem with sound. When I am seeing a video (YouTube or by using VLC), the sound lost synchronization with the video by stoping few seconds. Do you know where the problem could come from ?

Have a good day :)
Clover packages can be downloaded from CloverHackyColor page on GitHub:
Screen Shot 2020-04-24 at 5.18.48 AM.png
 

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
@c0c0p0ps
Sometimes an IP address can get mistakenly blocked by the forum software. It's best to contact @adamsmasher when this happens. Send him your IP that is blocked and an explanation of what happened.
Thank you! I've just sent a private message to him with traceroute output. For me, access to the site is still down so I purchased 1-month of VPN service which is allowing me to connect.
 
Joined
Feb 7, 2012
Messages
284
Motherboard
Gigabyte Z390 Designare
CPU
i7-9700K
Graphics
Vega 56
Mac
  1. MacBook
  2. MacBook Pro
Mobile Phone
  1. iOS
Upon closer inspection it seems:
  • Only one DP-In port as you stated.
  • Two internal USB 2.0 headers: F_USB1 and F_USB2.
  • Support for 5V addressable RGB. Looks like two ARGB headers.
  • Though unseen, there's probably an Intel CNVi slot with Intel AX200/AX201 module (WiFi 6).

Wow. Imagine if Gigabyte have been watching this thread and have directly translated the requirement of 2 USB headers because of the splitters you advocated for Hackintoshers!
 
Joined
Jul 20, 2013
Messages
81
Motherboard
Gigabyte Z390 Designare
CPU
i9-9900K
Graphics
RX 590
Just to set expectations...
  • It can be tricky both to attach the SOIC clip and to read the chip. These are of course two separate actions as @SuDoDmz mentioned.
  • How to know if you've attached the clip properly?
    • Turn off power to the board completely (physically pull power plug from PSU or from the wall).
    • Connect SOIC clip to Winbond chip with Pin 8 (Vcc) connected.
    • If SOIC clip is properly attached, LEDs on the motherboard should light up.
    • Then gently disconnect Pin 8 (Vcc) from the Raspberry Pi or elsewhere.
  • Now follow the supplemental procedure for reading the chip (link in Spoiler #2 in Post #1).
I've been trying to flash my onboard Z390 Designare Thunderbolt firmware for the last few days, and failed (miserably).
In the process I ran into many problems (see my story below), and may have fried or corrupted my Winbond chip or motherboard.

I could really use some help and advice:
Do you think there is still a way to recover, or do I need to order a new motherboard?


1. Flashed the Winbond chip with patched firmware (Verifying flash .... VERIFIED).
For flashing I used a Raspberry Pi and followed the guides for wiring and flashing (see spoiler)
Programming equipment and wiring
On Tuesday, I received the following parts (all new) for programming the Winbond chip:
  • Raspberry Pi 4 4GB
  • SOIC8-clip, with flatcable and socket adapter.
  • Breadboard, 47uF capacitor and 2K Ohm resistor.
  • Jumper wires (M/M, M/F, F/F).
Installation and wiring
I installed and configured the Raspberry Pi with "Raspbian Buster with desktop and recommended software" and enabling SPI, following the Mini-Guide for Flashing SPI ROM Chips using Raspberry Pi 3B or 4.

Then attached SOIC8-clip wires to GPIO according to the guide: Possible Improvement for In-Circuit Read/Write tot SPI Flash ROM Chips.

Below a photo of the wired setup:
3. RT_08261 - RPi and SOIC8 wired with breadboard -small.jpg


Photo of the SOIC8-clip attached to the Winbond chip
4. RT_08300 - SOIC8 clip on Winbond chip - small.JPG


Components and cables attached to motherboard
I disassembled my motherboard from my case, removed all components and disconnected all cables, except:
  1. ATX 24-pin cable (attached),
  2. CPU with Kraken X62 cooler (physically mounted, cables disconnected),
  3. M.2 SSD,
  4. 2x 16GB Memory Banks.

Firmware used
DESIGNARE-Z390-NVM33-Elias64Fr.bin from Repository of Patched Thunderbolt Firmware Files.

Guides read and used
  1. Mini-Guide for Flashing SPI ROM Chips using Raspberry Pi 3B or 4
  2. Possible Improvement for In-Circuit Read/Write to SPI Flash ROM Chips
  3. Reply from CaseySJ to me, setting expectations and giving pointers
  4. Supplemental Procedure for Connecting External SPI Flasher to Designare Z390
  5. Tips for Attaching SOIC8 Clip to Designare Z390 Winbond W25Q80DV
  6. Thunderbolt DROM Micro-Guide for Gigabyte Designare Z390 / Designare X299X

It took numerous attempts to get the SOIC8 clip attached properly. By properly, I mean that LEDs would be orange/red and not too bright, and that ROM reads were successful and checksums consistent.

Ultimately, I got consistent checksums and was able to flash the chip with patched firmware. See screenshot:
2. Screenshot 2020-04-21-222133 - FLASH success VERIFIED copy.png

Not sure if it means anything, but I noticed my Winbond chip does NOT have a blue dot, but a black dot.
1. Winbond-chip-black-dot.JPG

2. Mixed results: ThunderboltLocalNode populated, but SSP1 and SSP2 ports seem dead.
After reassembling my system and booting into macOS Catalina 10.15.4, the results were mixed:
  • In System Report > Thunderbolt, my Thunderbolt Bus and both Ports were listed instead of "No drivers loaded". However, several properties (UID, Firmware Version, ...) had a value of 0.
  • In Settings > Network, I did see a new Thunderbolt Bridge interface.
  • IOReg showed a populated ThunderboltLocalNode (only Domain UUID property) and ThunderboltPort@7 (several properties) under RP05, but USB ports SSP1 and SSP2 and Thunderbolt XHC5 were missing.
  • In Hackintool > USB, the Thunderbolt XHC5 controller was no longer listed, and SSP1 and SSP2 were no longer present in the list of USB ports.
  • I connected my Apple Thunderbolt Display (using an Apple TB3-to-TB2 USB-C adapter), but the screen stayed off. None of the display's internal interfaces (USB-ports, FaceTime camera, Network Port, Firewire port, Display Audio) were visible in System Report nor IOReg.
  • I connected my MacBook Pro /w touchbar (late 2016) with TB-cable to each of Designare's two TB3 ports. Nothing happened on either system.
  • I connected a USB3 Thumbdrive to both TB3 ports on the backplate of the Designare, and it would not mount.

It seems both TB3 ports are dead, both for USB and Thunderbolt connectivity.

3. Troubleshooting did not resolve anything.
I did my best to troubleshoot and analyse the issue, without results. Both TB3-ports remained dead, both for USB and Thunderbolt connectivity. Things I've tried include:
  • Checking all BIOS (F9b) settings according to post #1, loading optimized defaults and setup BIOS again to match the values in the guide.
  • Checking CLOVER/ACPI/PATCHED files, switching between SSDT-Z390-DESIGNARE-TB3HP-V4.aml, default SSDT-TBOLT3-RP05-PORT7-DESIGNARE-Z390.aml, and customized UID SSDT-TBOLT3-RP05-PORT7-DESIGNARE-Z390.aml
  • Cold booting with and without Apple Thunderbolt Display attached,
  • Complete power/PSU off and on again.
  • Checking MSR unlock,
  • Clearing NVRAM,
  • Checking Hackintool > Logs > System. There were only 5 lines related to IOThunderboltFamily (no errors), and no messages related to DROM.

4. Reset CMOS caused macOS to hang at boot.
As nothing had fixed the issues, I opted to clear BIOS using a CMOS reset.
After clearing CMOS, I loaded Optimized Defaults and set all BIOS settings according to post #1. Then checked MSR 0xE2 Register unlock status (which had reverted to 0x1, so set it to 0x0 again).

Then at Clover Boot I reset NVRAM and attempted to boot into macOS Catalina 10.15.4.
Booting halts shortly after showing the Apple logo, without showing the usual progress bar.
I did a hard reboot, and enabled verbose mode in Clover Boot Menu. Verbose screen shows this:
5. boot-verbose-stuck-after-CMOS-reset.jpg

5. Attempts to re-flash the Winbond chip keep failing.
The mixed results and boot-issues led me to believe that the chip was corrupted or not flashed correctly.

So after completely disassembling my system again, over the last two days I did countless attempts to re-flash the patched firmware to the Winbond chip. Unsuccessfully.

I tried many ways to get succesful reads and writes to the Winbond. Nothing worked. Flashing keeps failing.
  • Disconnecting, reconnecting, and wiggling SOIC8 clip.
  • Reading and Writing ROM with and without power from PSU to motherboard
  • Reading and Writing ROM with and without VCC (pin8) connected
  • Removing RAM and M.2 SSD from my motherboard, leaving only CPU on the motherboard.
  • Removing the breadboard with resistor and capacitor, and rewiring GPIO to SOIC8 clip according to the Guide for flashing with Raspberry Pi.
  • Rebooting Raspberry Pi.
  • Using the flashrom -f parameter to force reading and writing ROM.

6. Some observations and error messages during read and write.
Looking back at my notes and logs, the things below stand out:

On Wednesday
I did many attempts to read ROM and get at least 3 identical checksums. I found that:
  • ROM reads only work with PSU powered OFF.
  • ROM reads fail with PSU powered ON (No EEPROM/flash device found).
  • ROM writes fail with PSU powered OFF (No EEPROM/flash device found)
  • ROM writes fail with PSU powered ON (Erase failures).
I rewired GPIO/SOIC8 to the Raspberry Pi configuration (without breadboard etc) and repeated. Same problem.

At some point, the flashrom output states my Winbond chip is in an "unknown state" after a write operation.
Code:
pi@RBPi4B-RT:~/Desktop $ sudo flashrom -p linux_spi:dev=/dev/spidev0.0,spispeed=2048 -r Designare-8.bin
flashrom  on Linux 4.19.97-v7l+ (armv7l)
flashrom is free software, get the source code at https://flashrom.org

Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns).
Found Winbond flash chip "W25Q80.V" (1024 kB, SPI) on linux_spi.
Reading flash... done.

pi@RBPi4B-RT:~/Desktop $ shasum Designare*
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-1.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-2.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-3.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-4.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-5.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-6.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-7.bin
cceb66ecb01c7da3ff9e0de4031082d1af33f57c  Designare-8.bin

pi@RBPi4B-RT:~/Desktop $ sudo flashrom -p linux_spi:dev=/dev/spidev0.0,spispeed=1024 -w DESIGNARE-Z390-NVM33-Elias64Fr.bin -f
flashrom  on Linux 4.19.97-v7l+ (armv7l)
flashrom is free software, get the source code at https://flashrom.org

Using clock_gettime for delay loops (clk_id: 1, resolution: 1ns).
Found Winbond flash chip "W25Q80.V" (1024 kB, SPI) on linux_spi.
Reading old flash chip contents... done.
Erasing and writing flash chip... FAILED at 0x00019002! Expected=0xff, Found=0xef, failed byte count from 0x00019000-0x00019fff: 0xffe
ERASE FAILED!
Reading current flash chip contents... done. Looking for another erase function.
FAILED at 0x00019002! Expected=0xff, Found=0xef, failed byte count from 0x00018000-0x0001ffff: 0xffe
ERASE FAILED!
Reading current flash chip contents... done. Looking for another erase function.
FAILED at 0x00019002! Expected=0xff, Found=0xef, failed byte count from 0x00010000-0x0001ffff: 0xffe
ERASE FAILED!
Reading current flash chip contents... done. Looking for another erase function.
FAILED at 0x00019002! Expected=0xff, Found=0xef, failed byte count from 0x00000000-0x000fffff: 0x3ff8
ERASE FAILED!
Reading current flash chip contents... done. Looking for another erase function.
FAILED at 0x00019002! Expected=0xff, Found=0xef, failed byte count from 0x00000000-0x000fffff: 0x3ff8
ERASE FAILED!
Reading current flash chip contents... done. Looking for another erase function.
No usable erase functions left.
FAILED!
Uh oh. Erase/write failed. Checking if anything has changed.
Reading current flash chip contents... done.
Apparently at least some data has changed.
Your flash chip is in an unknown state.

On Thursday
I continued my pursuit to re-flash the chip. Wiring still according to RPi guide (without breadboard, capacitor, resistor). Rebooted and reconnected everything.

With VCC (pin 8) disconnected and PSU disconnected, and many attempts to clip the SOIC8 correctly, I was getting consistent read checksums.

When trying to write, I found that:
  • ROM writes with PSU power ON consistently result in: "No EEPROM/flash device found"
  • ROM writes with PSU power OFF consistently result in: "Verifying flash... FAILED, your flash chip is in an unknown state."
  • Note that the behavior of writing with PSU power ON/OFF is opposite of behavior on Wednesday.
Code:
Found Winbond flash chip "W25Q80.V" (1024 kB, SPI) on linux_spi.
Reading old flash chip contents... done.
Erasing and writing flash chip... Erase/write done.
Verifying flash... FAILED at 0x00000ffc! Expected=0x11, Found=0xff, failed byte count from 0x00000000-0x000fffff: 0xf76c5
Your flash chip is in an unknown state.


When I inspected the contents of Thursdays' dumped ROM files with Hex Fiend, I discovered that all values from begin till end are FFFFFFFF.
6. Hex Fiend - FFFFFFFF.png

Something feels terribly wrong here ... Any help on how to proceed is very much appreciated!
 
Last edited:

trs96

Moderator
Joined
Jul 31, 2012
Messages
18,907
Motherboard
GA-Z97X-UD3H-BK
CPU
i5-4690K
Graphics
HD4600 / RX 570
Mac
  1. MacBook Pro
  2. Mac mini
Mobile Phone
  1. Android
Thank you! I've just sent a private message to him with traceroute output. For me, access to the site is still down so I purchased 1-month of VPN service which is allowing me to connect.
Yes, this even happens to moderators too. Hope we can resolve it soon.
 

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
@CaseySJ Assuming the motherboards come out Q1 next year with the appropriate processors, will you upgrade at that point but still maintain your Z390 system with the latest applicable MacOS releases?

I'm only asking as if you're not, I'm going to have to curb back spending at Christmas so I can upgrade with you :p. My plan was to keep this system for at least a 3 or so years, but don't mind upgrading earlier if it means you are (and won't be supporting the Z390 further).
Every two years or so Intel releases a new CPU socket that prevents existing motherboards from using new CPUs. This is not entirely a bad thing because progress marches on and support for legacy parts cannot continue for long periods of time. Because Z390 boards will likely cease production in a couple of months, new users wishing to duplicate this build will probably look to the Z490 successor. So this build guide will adapt accordingly, but support for Z390 Designare will continue. So no worries there.

I'm not sure whether I'll personally upgrade to Z490 -- at this time the value proposition for me is just not there. Have you seen the suggested pricing on Z490? The Vision D might be US$393 versus US$269, which is a 46% increase in price. But most of all, the performance of my two Z390 systems is already more than I need.
Besides, these builds are way cheaper than Mac Pros and even iMacs, and come with more headroom for gaming etc (no thermal limitations, can upgrade SSD whenever etc) so we've saved loads of money in the long-run anyway. A Comparable iMac with i9/16GB RAM/1TB NVMe SSD and a Vega 48 (WHAT!?) comes to a whopping £3600 (granted there's a 5K screen too), which is astounding. I've always only bought MacBook Pros as they're not too bad when it comes to buying and then selling on after a few years, plus would never dream of hacking a laptop as too much effort. But I just can't justify Apple's desktop pricing. The mini is a thermal frying pan when it comes to heat, and I've read/heard setting up eGPUs with the mini and then bootcamping etc is almost as much of a pain as an actual Hackintosh!
Agreed, especially on the laptop side. I have not considered building a laptop Hackintosh simply because Apple has so many excellent choices at different price points. My 2012 MacBook Air is still going strong, but the new rumored 14" MacBook Pro is calling my name!
 

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
@CaseySJ
...
I have gone back to the first patched NVM 50 on my TR card. My devices are recognized again, and I'm seeing the TB bus again, but basically my 4K monitor is not recognized. I want to clarify that I do not have any TB monitors, but I am passing 2 DP channels out of my Radeon VII into my GC-Titan Ridge and then going into a CalDigit TS3 Plus dock. The dock has a built in DP which is where my 4K monitor is plugged in. Originally I could plug my HD monitor into the TB3 port on the dock using a usb-c to HDMI adapter. So far none of the patched firmware I have tried allows the CalDigit DP to be recognized.
I see -- I can try the same later today with my Belkin Dock Pro.
 

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
Hey CaseySJ.

#1 – I’ve changed my SMBIOS lately from iMac 19,1 to test iMac Pro 1,1 and MacPro 7,1 and now back to iMacPro 1,1 and I’m now having a strange issue where my Messages won’t deliver and new ones don’t show up. Everything else iCloud related works fine. I’ve tried logging out and back into icloud but same thing. I’m using native NVRAM; maybe that’s got something to do with it. Any idea what might be causing the non deliveries and inability to get new text updates?

#2 – Also, earlier I thought I saw that you said there was something possibly wrong with the SSDT for AMD acceleration I sent you. What did see anything weird in it?
I looked at the SSDT again yesterday and it's perfectly OKAY.
EDIT: #1 – I figured it out. Would still like your thoughts on #2 though when you get a chance. I also am testing the Radeonboost.kext way for GPU injection and it seems to be a winner as well with no configuration needed in config.plist. Nice speed improvement.
For issue #1, was it due to missing en0 in System Information --> Ethernet?
 

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
...
Now I remembered another question: Why is it that important to only flash the chip with the blue dot? Why not green? Why not both?
...
On GC-Titan Ridge, blue dot chip contains Thunderbolt firmware, but green dot chip contains Texas Instruments Power Delivery (PD) firmware. PD is necessary to manage and deliver power to connected devices (i.e. if you want to charge a laptop computer through Thunderbolt port).
 

CaseySJ

Moderator
Joined
Nov 11, 2018
Messages
13,463
Motherboard
Gigabyte Z490 Vision D
CPU
i5-10400
Graphics
RX 580
Mac
  1. MacBook Air
  2. MacBook Pro
  3. Mac Pro
Classic Mac
  1. Quadra
Mobile Phone
  1. iOS
@faithie999 @Elias64Fr Firstly, thanks for putting together the custom firmware for the NUC8i3BEH. And major thanks for @CaseySJ too! Can one of you tell me where the ROM chip is on the NUC8i3BEH? I have the same model and I'm planning on flashing it with the custom firmware @CaseySJ has attached in post #16395. Attaching the mainboard images below. Help is appreciated! Thanks :)

View attachment 464547
View attachment 464549
First look for any chip with 8 pins (four on one side, four on the other). Then look closely at the label on the chip. Does it say:
  • Winbond / W25Q80DV...
  • MXIC / MX25L8006E...
The chip should be in close proximity to the Thunderbolt controller (Intel JHL 6xxx or 7xxx). It could even be behind the Thunderbolt chip (i.e. on the other side of the board).
 
Top