- Joined
- Aug 27, 2019
- Messages
- 41
- Motherboard
- Gigabyte Z390 Designare
- CPU
- I9-9900K
- Graphics
- Vega Frontier
USB "Disk Not Ejected Properly" during sleep, Gigabyte Z390 Designare with memory above 2666 MHz.
This problem has been reported under other threads including CaseySJ's amazing Golden Build, however, I decided to make a separate, concise thread to try and solve this issue for the purpose of expediency and not getting lost in other support discussion.
The obvious fix is to set your computer to never sleep if you want to clock your memory above 2666 MHz, but that turns your computer into a massive energy hog if you are not constantly using it and need sleep functionality for quick wakes to have it ready to go when you need your system and frequently use USB drives.
If you click your memory to 2666 MHz and below, this problems is a non issue. The disks do not eject during sleep and are simply still there after wake/login and aren't visibly remounted. They just show up in the desktop as they would normally.
*If this issue is important to you, please follow this thread.
*If you have found a solution, please let us know!
*Also, if you are convinced this is a motherboard issue, please contact Gigabyte Support and/or reply to a thread created on their forum, specific to this issue, in order to influence their decision in creating a fix in a BIOS update if that is the culprit. A lot of people buy their boards to do these builds after all, so minimal support could be warranted.
Motherboard: Gigabyte z390 Designare
(This happens on several systems; however, this thread is mainly considering this, and other Gigabyte motherboards (it has been reported to be fixed on at least one ASUS board - see below under "SUSPICIONS").
Memory: Kingston HyperX Predator Black 3600 MHz
(Memory brand does not matter. This happens to all memory when exceeding 2666 MHz (you can clock down to this speed and the problem goes away)).
Operating System: Mojave 10.14.6
(This has also been reported on Catalina and at least a couple other Mojave versions. Also, it has been reported on setting Clover system profiles in Clover Configurator to iMac Pro 1,1 or iMac 19,1)
MY SUSPICIONS:
***The problem has reportedly been fixed on the Asus Asus Z390-A motherboard through an UEFI BIOS update, leading me to believe suspicion C is the culprit (or suspicion B). Link on the report here: https://www.tonymacx86.com/threads/...derbolt-3-i7-9700k-amd-rx-580.267551/page-763
A. When memory is set to exceed 2666 MHz, macOS will automatically and inproperly eject USB drives during sleep and fail to remount them without a re-plug of the USB drive.
B. When memory is set to exceed 2666 MHz, macOS will automatically and improperly eject USB drives and the motherboard will not recognize the USB drive without a re-plug of the USB drive.
C. When memory is set to exceed 2666 MHz, the motherboard will automatically and improperly eject USB drives and the motherboard will not recognize the USB drive without a re-plug of the USB drive.
OTHER CLUES AND INFORMATION:
Terminal:
During the eject/remount failure, Terminal cannot see any ejected USB disk without a re-plug of the USB drive (at which point the USB drive mounts normally).
Available apps and scripts do not work to solve the issue:
There are several scripts and apps designed to fix this (or a similar issue) on regular Macs, however, they only half work in this case. USB disks will be ejected (protecting data), but will fail to remount.
Perhaps during sleep, the motherboard (or OS) does something that still makes it impossible for terminal or the apps/scripts to see an therefore remount the drives.
***HOWEVER: The Apps/Scripts/Terminal function normally if memory is set to 2666 MHz and below in the motherboards UEFI BIOS. This renders the usefulness of these solutions redundant, but may offer a clue as to what is happening.
The scripts and apps that have been tested are: Jettison, SleepWatcher, and SafeSleepUSBv1.2
Other things setting memory in UEFI to 2666 MHz and below fixes:
At this memory setting, it is possible to remove "darkwake" setting completely form Clover/Clover configurator. The computer will sleep normally, without issue and wake/sleep as it should. It has also been reported to wake up more quickly. I didn't notice a noticeable wakeup difference from leaving darkwake at darkwake=8, but exceeding 2666 MHz in UEFI BIOS, darkwake must be enabled to have the machine wake with one single keystroke. Otherwise, typical issues include a partial wake to a black screen and then having to strike the keyboard at least once or several times in order to wake the screen.
VERY IMPORTANT - SMBIOS MEMORY SETTINGS VS. GIGABYTE'S UEFI BIOS:
It seems UEFI bios ultimately supercedes SMBIOS.
Manual entry of memeroy in SMBIOS only has a cosmetic effect on macOS, and does not fix the problem nor does make a difference in performance.
If you set the memory frofikes in SMBIOS, About This Mac and System Report will show the Memory information that was placed in SMBIOS.
HOWEVER, if you run Geekbench (in my case V 4), Geekbench will give you scores based on what is set up in UEFI bios, and not what About This Mac or System Report show.
This suggests that even though macOS is showing one thing, the computer is still running/performing based on UEFI settings and not SMBIOS settings, making the SMBIOS settings purely cosmetic to About This Mac and System Report. USB Disk Ejected Improperly during sleep still occurs.
Other notes and clues:
During sleep, if a USB disk is inserted, it will wake the computer, but the drive will not mount.
Berfore commenting on this thread:
Before suggesting common fixes and asking "what has been done," especially if you do not have this problem, don't have this or Gigabyte motherboard, and more importantly, have not found a solution to this problem, please consult CaseySJ's awesome Golden Build here for general information on settings etc (this problem also occurs on other builds not following this guide BTW!):
www.tonymacx86.com
This problem has been reported under other threads including CaseySJ's amazing Golden Build, however, I decided to make a separate, concise thread to try and solve this issue for the purpose of expediency and not getting lost in other support discussion.
The obvious fix is to set your computer to never sleep if you want to clock your memory above 2666 MHz, but that turns your computer into a massive energy hog if you are not constantly using it and need sleep functionality for quick wakes to have it ready to go when you need your system and frequently use USB drives.
If you click your memory to 2666 MHz and below, this problems is a non issue. The disks do not eject during sleep and are simply still there after wake/login and aren't visibly remounted. They just show up in the desktop as they would normally.
*If this issue is important to you, please follow this thread.
*If you have found a solution, please let us know!
*Also, if you are convinced this is a motherboard issue, please contact Gigabyte Support and/or reply to a thread created on their forum, specific to this issue, in order to influence their decision in creating a fix in a BIOS update if that is the culprit. A lot of people buy their boards to do these builds after all, so minimal support could be warranted.
Motherboard: Gigabyte z390 Designare
(This happens on several systems; however, this thread is mainly considering this, and other Gigabyte motherboards (it has been reported to be fixed on at least one ASUS board - see below under "SUSPICIONS").
Memory: Kingston HyperX Predator Black 3600 MHz
(Memory brand does not matter. This happens to all memory when exceeding 2666 MHz (you can clock down to this speed and the problem goes away)).
Operating System: Mojave 10.14.6
(This has also been reported on Catalina and at least a couple other Mojave versions. Also, it has been reported on setting Clover system profiles in Clover Configurator to iMac Pro 1,1 or iMac 19,1)
MY SUSPICIONS:
***The problem has reportedly been fixed on the Asus Asus Z390-A motherboard through an UEFI BIOS update, leading me to believe suspicion C is the culprit (or suspicion B). Link on the report here: https://www.tonymacx86.com/threads/...derbolt-3-i7-9700k-amd-rx-580.267551/page-763
A. When memory is set to exceed 2666 MHz, macOS will automatically and inproperly eject USB drives during sleep and fail to remount them without a re-plug of the USB drive.
B. When memory is set to exceed 2666 MHz, macOS will automatically and improperly eject USB drives and the motherboard will not recognize the USB drive without a re-plug of the USB drive.
C. When memory is set to exceed 2666 MHz, the motherboard will automatically and improperly eject USB drives and the motherboard will not recognize the USB drive without a re-plug of the USB drive.
OTHER CLUES AND INFORMATION:
Terminal:
During the eject/remount failure, Terminal cannot see any ejected USB disk without a re-plug of the USB drive (at which point the USB drive mounts normally).
Available apps and scripts do not work to solve the issue:
There are several scripts and apps designed to fix this (or a similar issue) on regular Macs, however, they only half work in this case. USB disks will be ejected (protecting data), but will fail to remount.
Perhaps during sleep, the motherboard (or OS) does something that still makes it impossible for terminal or the apps/scripts to see an therefore remount the drives.
***HOWEVER: The Apps/Scripts/Terminal function normally if memory is set to 2666 MHz and below in the motherboards UEFI BIOS. This renders the usefulness of these solutions redundant, but may offer a clue as to what is happening.
The scripts and apps that have been tested are: Jettison, SleepWatcher, and SafeSleepUSBv1.2
Other things setting memory in UEFI to 2666 MHz and below fixes:
At this memory setting, it is possible to remove "darkwake" setting completely form Clover/Clover configurator. The computer will sleep normally, without issue and wake/sleep as it should. It has also been reported to wake up more quickly. I didn't notice a noticeable wakeup difference from leaving darkwake at darkwake=8, but exceeding 2666 MHz in UEFI BIOS, darkwake must be enabled to have the machine wake with one single keystroke. Otherwise, typical issues include a partial wake to a black screen and then having to strike the keyboard at least once or several times in order to wake the screen.
VERY IMPORTANT - SMBIOS MEMORY SETTINGS VS. GIGABYTE'S UEFI BIOS:
It seems UEFI bios ultimately supercedes SMBIOS.
Manual entry of memeroy in SMBIOS only has a cosmetic effect on macOS, and does not fix the problem nor does make a difference in performance.
If you set the memory frofikes in SMBIOS, About This Mac and System Report will show the Memory information that was placed in SMBIOS.
HOWEVER, if you run Geekbench (in my case V 4), Geekbench will give you scores based on what is set up in UEFI bios, and not what About This Mac or System Report show.
This suggests that even though macOS is showing one thing, the computer is still running/performing based on UEFI settings and not SMBIOS settings, making the SMBIOS settings purely cosmetic to About This Mac and System Report. USB Disk Ejected Improperly during sleep still occurs.
Other notes and clues:
During sleep, if a USB disk is inserted, it will wake the computer, but the drive will not mount.
Berfore commenting on this thread:
Before suggesting common fixes and asking "what has been done," especially if you do not have this problem, don't have this or Gigabyte motherboard, and more importantly, have not found a solution to this problem, please consult CaseySJ's awesome Golden Build here for general information on settings etc (this problem also occurs on other builds not following this guide BTW!):
[SUCCESS] Gigabyte Designare Z390 (Thunderbolt 3) + i7-9700K + AMD RX 580
CaseySJ's Coffee Lake in Phanteks Evolv X ATX Case: Gigabyte Designare Z390 - i7-9700K - UHD 630 - AMD RX 580 (Please do not quote this build guide in its entirely. Post a link instead.) (See this post for how to refer to a Guide or Mini-Guide.) URGENT: Please avoid Clover 5123 and newer for...
Last edited:
