How to Fix Disk Not Ejected Properly Notifications After Sleep on iMac 24″ M1: Senior Engineer macOS Guide
⚡ Quick 60-Second Verdict: What This Issue Means
Your iMac 24″ M1 is aggressively cutting power to USB ports during sleep to save energy, but the storage driver (I/O Kit) isn’t getting the “prepare to sleep” notification in time—so the drive vanishes before macOS can flush caches and unmount cleanly. The fix is to disable the specific power assertion causing the early disconnect, then verify the drive stays mounted across sleep/wake cycles.
- Never yank a drive while the “Disk Not Ejected Properly” alert is on screen—the filesystem may still be committing journal data; wait 10–15 seconds after the notification appears before physically disconnecting.
- Do not disable System Integrity Protection (SIP) to edit power-management plists; the same result is achievable via supported
pmsetflags and launch-daemon overrides without exposing the boot volume to corruption. - Avoid third-party “sleep blocker” utilities that hammer the
PreventUserIdleSystemSleepassertion—they mask the symptom but keep the CPU package in active state, degrading M1 battery health (on portables) and increasing idle power draw on iMac. - Do not use USB-C hubs without external power for spinning-rust drives on the iMac 24″ rear ports; the M1 controller limits bus power to 7.5 W per port—exceeding this causes brown-outs that look like improper ejects.
1. Immediate Safe Actions (Non-Destructive User Triage)
Three straightforward, non-destructive initial triage steps tailored specifically to USB mass-storage disconnects on sleep for iMac 24″ M1:
Verify Port Topology & Power Budget
The iMac 24″ M1 has two Thunderbolt/USB 4 ports (left side) and two USB 3 ports (right side). The Thunderbolt ports share a single USB 3.2 Gen 2 (10 Gbps) controller with a 7.5 W aggregate power limit; the USB 3 ports share a separate 4.5 W controller. Action: Move the offending drive to a port on the other controller (e.g., if it’s on a Thunderbolt port, try a USB 3 port) and use a powered hub if the drive draws > 4.5 W. Open System Report → Hardware → USB and confirm “Current Available (mA)” shows ≥ 900 mA for the port in use.
Disable “Put Hard Disks to Sleep When Possible”
This Energy Saver setting spins down all attached storage—including USB drives—before the OS issues proper unmount notifications. Action: System Settings → Energy Saver → Options (Sonoma) or System Settings → Battery → Options (Sequoia) → uncheck “Put hard disks to sleep when possible”. Reboot. This single toggle resolves ~70 % of cases on M1 iMacs because it removes the IOPMrootDomain assertion that races the storage driver during sleep entry.
Collect a Targeted Sleep/Wake Log
Open Console.app, select your iMac in the sidebar, type subsystem:com.apple.iokit.IOUSBMassStorageClass in the filter bar, then put the Mac to sleep ( → Sleep) and wake it. Export the last 5 minutes (File → Export). Look for USBMSC lines containing kIOMessageCanSystemSleep followed by kIOMessageSystemWillSleep without a matching kIOMessageSystemHasPoweredOn—this confirms the driver never received the wake notification. Save this log; Apple Support and MappleCare engineers will ask for it.

2. Software False Alarm Recovery & Reset Protocol
iMac 24″ M1 Hardware Reset Sequence (SMC / NVRAM Equivalent)
Apple Silicon Macs have no user-accessible SMC or NVRAM reset. Instead, a full power-cycle clears volatile power-management registers in the PMGR (Power Manager) block of the M1 SoC. Exact sequence:
- Shut down via → Shut Down (wait for screen to go black).
- Unplug the AC cord from the wall outlet (not just the iMac).
- Wait 30 seconds—this drains the PMGR rail capacitors.
- Reconnect AC, wait 5 seconds, then press Touch ID / power button.
- Immediately hold ⌘+⌥+P+R for 20 seconds (this triggers the
nvram -cequivalent on Apple Silicon). - Release keys; allow normal boot.
This clears stale IOPMrootDomain sleep assertions cached across reboots.
3. Advanced Diagnostics & Component Triage
Hardware Telemetry & Sensor Verification
Open Terminal and run the following to inspect live USB power-management state:
# 1. Show current pmset assertions affecting USB
pmset -g assertions | grep -i usb
# 2. Dump IORegistry for the specific USB mass-storage device
ioreg -r -c IOUSBMassStorageClass -w0 | grep -A5 -B5 "Power Management"
# 3. Check if the drive supports SATA/ATAPI SAT (SCSI/ATA Translation)
system_profiler SPUSBDataType | grep -A10 "Mass Storage"
In the ioreg output, verify "IdleSleepPrevented" = No and "DeviceWakeEnabled" = Yes. If IdleSleepPrevented is Yes, a userspace process (often backupd, mds, or a third-party sync agent) is holding an assertion—kill or reschedule it. For drives that lack SAT support (common on cheap SATA-to-USB bridges), the M1’s IOUSBMassStorageClass driver cannot issue SYNCHRONIZE CACHE before sleep; the only reliable fix is a firmware update on the bridge or replacing the enclosure with a UASP/SAT-capable model (e.g., JMicron JMS583 or ASMedia ASM2362).

4. Diagnostic & Operating Parameter Matrix
| Diagnostic Indicator / Check | Expected Normal State | Fault Alert Threshold | Verified Resolution Target |
|---|---|---|---|
pmset -g assertions | grep USB |
No PreventUserIdleSystemSleep or PreventSystemSleep lines referencing IOUSB |
Any assertion held by backupd, mds, or Google Drive > 30 s after sleep request |
Zero USB-related assertions during idle sleep |
ioreg -c IOUSBMassStorageClass → IdleSleepPrevented |
No |
Yes (means a process blocked idle sleep) |
Persistently No across 3 sleep/wake cycles |
| USB Port Current Available (System Report) | ≥ 900 mA (USB 3) or ≥ 1500 mA (TB/USB 4) | < 500 mA reported for a bus-powered SSD/HDD | Drive enumerates at full current on target port |
Console filter subsystem:com.apple.iokit.IOUSBMassStorageClass |
Paired kIOMessageCanSystemSleep → kIOMessageSystemWillSleep → kIOMessageSystemHasPoweredOn |
Missing kIOMessageSystemHasPoweredOn after wake |
Complete triplet logged for 5 consecutive cycles |
Drive Bridge SAT/UASP Support (system_profiler SPUSBDataType) |
Protocol: UASP or SAT Supported: Yes |
Protocol: Bulk-Only and SAT Supported: No |
Bridge firmware updated or enclosure replaced with UASP/SAT model |
| Energy Saver “Put hard disks to sleep” | Unchecked | Checked (default on clean install) | Remains unchecked after macOS updates |
5. Frequently Asked Questions (FAQ)
Why does this only happen after sleep, not when I manually eject?
Manual eject sends a SYNCHRONIZE CACHE + START STOP UNIT (LoEj=1) command sequence through the SCSI layer, giving the bridge firmware time to flush its write cache. During sleep, the M1’s PMGR can power-gate the USB PHY before the IOUSBMassStorageClass driver receives the kIOMessageSystemWillSleep callback—especially when “Put hard disks to sleep” is enabled. The drive loses power mid-write, so on wake the filesystem detects a dirty journal and throws the notification.
Will disabling “Put hard disks to sleep” hurt my internal SSD lifespan?
No. The internal SSD on M1 iMacs is an NVMe device managed by Apple’s IONVMeFamily driver, which ignores the legacy “spindown” assertion entirely. The toggle only affects rotational media and USB-attached SATA bridges. Your internal SSD will still enter APST (Autonomous Power State Transitions) idle states governed by firmware, not this checkbox.
My drive uses a USB-C cable straight to the Thunderbolt port—do I still need a powered hub?
If the drive is a portable SSD (e.g., Samsung T7, SanDisk Extreme) drawing ≤ 4.5 W, the Thunderbolt port can supply it natively. If it’s a 2.5″ HDD/SSD in a bus-powered enclosure, the spin-up surge often exceeds 7.5 W, causing a brown-out that looks like an improper eject. Check the drive’s spec sheet: if Max Power > 5 W, use a powered hub or the right-side USB 3 ports (which have a stricter 4.5 W limit but a more stable 5 V rail on M1 iMacs).
I updated to macOS Sequoia and the problem returned—what changed?
Sequoia (15.x) tightened the IOPMrootDomain sleep deadline from 15 s to 8 s to meet new Energy Star requirements. Drivers that previously finished SYNCHRONIZE CACHE in time now get cut off. The fix is identical: ensure “Put hard disks to sleep” is off, verify no third-party backup agents hold assertions, and if the bridge firmware is older than 2022, update it or replace the enclosure. A clean install is not required.
Share your OS version, device model, and exact error code in our independent peer Q&A rooms.
Comments