Mac is causing many unsafe shutdowns to OWC express 1m2

[zelaxx]

I've been having the same problem with my new Mac Mini M4 and OWC Express 1M2.

I had over a dozen kernel panics before I realised it was the 1M2 causing the problem:

/Library/Logs/DiagnosticReports/panic-etc...
> panic(cpu 4 caller 0xfffffe003d0656b8): busy timeout[0], (60s): 'pci1c5c,174a' (a,4020001) @IOService.cpp:5829

Since then, the partition doesn't seem to get recognised when I plug it into the TB4 port. The device shows up under Thunderbolt/USB4 in system information app, but it does not show the partitions, and nothing is mounted in macos.

When I instead plug it into one of the USB-C ports (non-USB4/non-TB4), the drive is recognised fine.

I wonder if the ASM2464 is broken for USB4 connectivity... Unfortunately I don't have another USB4 drive or USB4/TB4 computer to test the 1M2 with.

Does anyone else have the same problem? i.e. doesn't work when connected via TB4/USB4 to macos, but does with plain USB-C?

 

[herbert7265]

@zelaxx

I wonder if the ASM2464 is broken for USB4 connectivity... Unfortunately I don't have another USB4 drive or USB4/TB4 computer to test the 1M2 with.

Does anyone else have the same problem? i.e. doesn't work when connected via TB4/USB4 to macos, but does with plain USB-C?

Okay, my problem was totally different (slow data transfer speed of the 1M2), but maybe it's worth to check one or two things in your case.

Do you have a Windows computer available, or a friend who has a Windows computer?

If so, you may want to check the firmware version of your 1M2 controller, potentially update it, and maybe play around with the configuration to see if there is a change in the behavior of your 1M2 connected to your Mac Mini M4.

I simply attach to this post the little "manual" which I made concerning my problem (hopefully that is allowed). The manual contains the link to the 1M2 controller firmware version I used and the necessary explanations how to check and update the firmware and modify the configuration of the controller.

REMARK: There is now an even more actual firmware available (Asmedia ASM2464 NVME/USB 4.x Controller Firmware Version 250717_85_00_00), just go from the link in my manual to the newest firmware version and download that package. The process itself should still be the same.

The only little problem in all this: The app for the firmware updates and configuration settings is (still) Windows only! It does not work in a virtual machine on Mac!

Again, I may be wrong, but the 1M2 is a USB4 device, not TB4, and in my case it was the U4 Mode setting in the 1M2 controller that solved the problem (not recognized as a USB4 device when connected via CalDigit TS4 dock, only when directly connected to a Mac Book Pro M2 Max).

Any more questions just let me know.

Herbert

 

[Alpha Centauri]

I've got the same enclosure and SSD combination so will try it. I haven't had issues as it's being used as a DAS for TM, which after completion I eject immediately. Next time I'll let the MBP sleep whilst still mounted.

 

[TimC00ked]

I've been having the same problem with my new Mac Mini M4 and OWC Express 1M2.

I had over a dozen kernel panics before I realised it was the 1M2 causing the problem:

/Library/Logs/DiagnosticReports/panic-etc...
> panic(cpu 4 caller 0xfffffe003d0656b8): busy timeout[0], (60s): 'pci1c5c,174a' (a,4020001) @IOService.cpp:5829

Since then, the partition doesn't seem to get recognised when I plug it into the TB4 port. The device shows up under Thunderbolt/USB4 in system information app, but it does not show the partitions, and nothing is mounted in macos.

When I instead plug it into one of the USB-C ports (non-USB4/non-TB4), the drive is recognised fine.

I wonder if the ASM2464 is broken for USB4 connectivity... Unfortunately I don't have another USB4 drive or USB4/TB4 computer to test the 1M2 with.

Does anyone else have the same problem? i.e. doesn't work when connected via TB4/USB4 to macos, but does with plain USB-C?

This doesn't seem to be the same problem. I don't have panic reports and my enclosure is recognized when connected to TB4.

 

[zelaxx]

@zelaxx

Okay, my problem was totally different (slow data transfer speed of the 1M2), but maybe it's worth to check one or two things in your case.

Do you have a Windows computer available, or a friend who has a Windows computer?

If so, you may want to check the firmware version of your 1M2 controller, potentially update it, and maybe play around with the configuration to see if there is a change in the behavior of your 1M2 connected to your Mac Mini M4.

I simply attach to this post the little "manual" which I made concerning my problem (hopefully that is allowed). The manual contains the link to the 1M2 controller firmware version I used and the necessary explanations how to check and update the firmware and modify the configuration of the controller.

REMARK: There is now an even more actual firmware available (Asmedia ASM2464 NVME/USB 4.x Controller Firmware Version 250717_85_00_00), just go from the link in my manual to the newest firmware version and download that package. The process itself should still be the same.

The only little problem in all this: The app for the firmware updates and configuration settings is (still) Windows only! It does not work in a virtual machine on Mac!

Again, I may be wrong, but the 1M2 is a USB4 device, not TB4, and in my case it was the U4 Mode setting in the 1M2 controller that solved the problem (not recognized as a USB4 device when connected via CalDigit TS4 dock, only when directly connected to a Mac Book Pro M2 Max).

Any more questions just let me know.

Herbert

Thanks I took a look.

Do you have the stock ini that OWC uses? If I flash the firmware it will overwrite the serial number (easily fixed), vendor and device IDs, etc.

Edit: found it in the PDF. Ultimately updating the firmware and changing the USB4 mode didnt help. Will have to find another USB4 computer to test this on.

 

[herbert7265]

Edit: found it in the PDF. Ultimately updating the firmware and changing the USB4 mode didnt help. Will have to find another USB4 computer to test this on.

Sorry to read that this potential solution didn't solve your problem, despite it was a long shot from the beginning.

Nevertheless it is quite interesting reading about different problems of the 1M2 in combination with different computers and equipment. While a really nice and fast enclosure when it works it seems to have its quirks, something I didn't face for example with an Acasis enclosure, despite it being a little slower.

Herbert

 

[DaveEcc]

Not sure if this is a separate long-standing Mac issue, or part of the same thing... Does this occur on every shutdown, or only sometimes. If it's only sometimes, perhaps what I've noticed is the cause.

I have my photos on an external T7 Shield. Occasionally, after processing some photos I'll AirDrop them to my wife via the share button in Finder. If I do this, I can't eject the drive. AirDrop keeps an open handle to something on the disk, and the OS isn't smart enough to relinquish that handle on an eject request. I can choose to Force Eject it, and it's fine.

 

[rumorguy]

OWC sells complete external drives, I have an OWC Envoy with an M1 iMac. Works great. OWC has occasional sales.

 

[OWC_TAL]

I've been having the same problem with my new Mac Mini M4 and OWC Express 1M2.

I had over a dozen kernel panics before I realised it was the 1M2 causing the problem:

/Library/Logs/DiagnosticReports/panic-etc...
> panic(cpu 4 caller 0xfffffe003d0656b8): busy timeout[0], (60s): 'pci1c5c,174a' (a,4020001) @IOService.cpp:5829

Since then, the partition doesn't seem to get recognised when I plug it into the TB4 port. The device shows up under Thunderbolt/USB4 in system information app, but it does not show the partitions, and nothing is mounted in macos.

When I instead plug it into one of the USB-C ports (non-USB4/non-TB4), the drive is recognised fine.

I wonder if the ASM2464 is broken for USB4 connectivity... Unfortunately I don't have another USB4 drive or USB4/TB4 computer to test the 1M2 with.

Does anyone else have the same problem? i.e. doesn't work when connected via TB4/USB4 to macos, but does with plain USB-C?

Sounds like you have a faulty unit. Please contact our support so we can make it right!

 

[DeepSix]

It's fine, I've had Mac minis running for nearly a decade on 24/7 with the fan running. I stopped shutting them off when I started using them as a media server, it's no big deal.

Are you saying you leave your Minis on 24/7 around the clock and never shut them down or put to sleep? But the fan always spins at the same speed if left on 24/7. Won't it wear out the lifespan of the fan?

And sleep mode is messed up and doesn't work properly. My LED keyboard and my 1M2 drive keeps waking up the Mini anyways every 5 mins so I stopped using sleep mode altogether. There is no fix for this. Sleep on Mac mini is unusable. But its great on a MacBook.

 

[DeepSix]

The usb/thunderbolt power software stack in macos is somewhat finicky and unreliable. Your best bet would be to keep the mini from going to sleep. Energy consumption difference is minimal. The alternative would be a shutdown of the mini, which would be more sustainable anyways of you let your mac sleep during nighttime.

There is no real solution to this at the moment, as Apple needs to improve their drivers.

I have the 1M2 as well and have the same issue as OP. Powering off the M4 mini doesn't fix this problem. My 1M2 always disconnects when the M4 mini is powered down and I have to unplug and plug it back in for the M4 mini to read it again. Extremely annoying.

From my research it's a power issue with the TB4 ports on the back of the M4 mini. Im going to pick up a TB4 hub and connect the 1M2 through there to see if this fixes the issue. The hub takes in 65W so this should fix the power distribution issue if that's the case.

 

[!!!]

Are you saying you leave your Minis on 24/7 around the clock and never shut them down or put to sleep? But the fan always spins at the same speed if left on 24/7. Won't it wear out the lifespan of the fan?

I guess so, but it hasn't happened so far. The fan is spinning at its base RPM 99% of the time anyways. The lifespan of the fans is probably longer than the useful lifespan of the mini anyways.

And sleep mode is messed up and doesn't work properly. My LED keyboard and my 1M2 drive keeps waking up the Mini anyways every 5 mins so I stopped using sleep mode altogether. There is no fix for this. Sleep on Mac mini is unusable. But its great on a MacBook.

Yep. And it's not always great on MacBooks, my M1 consumes battery like crazy unless I disable Wi-Fi before closing the lid.

 

[EugW]

I don't have this problem with my M4 Mac mini with Hagibis & Qwiizlab ASM2464PD USB 4 SSD enclosures. Drives are Samsung 990 Pro 4 TB and Kioxia XG8 4 TB. They would shut down at sleep and wake properly. So it sounds like an OWC 1M2 problem, or else a cable problem. Have you tried a new cable?*

However, now in Tahoe I have to prevent from Mac mini from going to sleep anyway, because my Huawei MateView monitor no longer wakes from sleep. The monitor can go to sleep, but in Tahoe I can't let the Mac go to sleep too because then the monitor won't wake up. These are the pleasures of using 3rd party equipment. The frustrating thing is that monitor wake up worked just fine with my M1 and my M4 previously before Tahoe. It seems that Tahoe is the issue for me. What OS are you guys running?

*Note I did have a disconnect issue with my Plugable Thunderbolt 4 hub. After about a year, it would randomly disconnect. It turns out it was the cable which failed after a year, and it was the original Plugable cable. What made it confusing is that it wasn't a catastrophic fail, but a fail where the cable became borderline, so sometimes it would work, and sometimes it wouldn't. After some troubleshooting confirmed it was the cable, I made a warranty claim for a new cable. However, Plugable didn't quite believe my troubleshooting and instead of sending me the new cable I asked for, they sent me a whole new hub complete with all the accessories and of course cable. So now I have two fully functioning Plugable Thunderbolt 4 hubs.

 

[DeepSix]

I guess so, but it hasn't happened so far. The fan is spinning at its base RPM 99% of the time anyways. The lifespan of the fans is probably longer than the useful lifespan of the mini anyways.

Yep. And it's not always great on MacBooks, my M1 consumes battery like crazy unless I disable Wi-Fi before closing the lid.

I wonder how many hours the mac mini fans is rated for. Are they easily purchased and replaced? I've heard of people running them 24/7/365 as a server with no issues.

 

[Brian33]

I've heard of people running them 24/7/365 as a server with no issues.

My 2012 Intel I7 Mac Mini has been running as a headless server for over 11 years, 24/7 for nearly all that time. It's still going strong. Most of the time the fan was low speed, but it had significant usage as a Minecraft server where the fan would run at max speed (>5000 rpm, IIRC) for hours and hours. Very pleased with that Mini's durability. I replaced the thermal paste and the coin cell battery once. No other maintenance.

I expect it would be easy enough to replace the fan with a 3rd-party fan, like a Noctua. If it fails soon I guess I'll find out!

 

[Soondae]

I had the same issue with notifications of unsafe disconnect, then a friend talked me into formatting the NVME SSD in my hub to APFS (it was originally FAT+) the notifications went away, I have not had any issues since converting to APFS.

 

[EugW]

I had the same issue with notifications of unsafe disconnect, then a friend talked me into formatting the NVME SSD in my hub to APFS (it was originally FAT+) the notifications went away, I have not had any issues since converting to APFS.

Ah. My Hagibis and Qwiizlab are both APFS and are working fine.

 

[DeepSix]

After doing a lot of research on external drives ejecting from M4 Mini's. I am 99.99% confident that it's a power distribution problem with the Mini not being able to supply enough consistent power to the drives. I read a couple reviews online that said after getting a powered thunderbolt hub, it stopped the issue.

I really didn't want to spend the money but I have no choice if I want to stop my OWC 1M2 from ejecting. So I ordered a Kensington Thunderbolt 4 hub that comes with a 1500W power supply. That will be more than enough power for the OWC 1M2. I did consider getting the OWC TB4 hub for $30 less but some have said that when more than one drive is connected the speeds get cut in half. That's almost a deal breaker for me.

Plus the Kensington has a bit better reviews on amazon and an extra year of warranty (2 years vs 3 years). And Kensington has great customer support. Can't say the same for OWC. The hub will be here tomorrow and will report back with my findings. If I doesn't fix the issue then I will send it back and get the OWC hub. But I'm confident the Kensington will stop the ejections on the 1M2.

 

[DeepSix]

My 2012 Intel I7 Mac Mini has been running as a headless server for over 11 years, 24/7 for nearly all that time. It's still going strong. Most of the time the fan was low speed, but it had significant usage as a Minecraft server where the fan would run at max speed (>5000 rpm, IIRC) for hours and hours. Very pleased with that Mini's durability. I replaced the thermal paste and the coin cell battery once. No other maintenance.

I expect it would be easy enough to replace the fan with a 3rd-party fan, like a Noctua. If it fails soon I guess I'll find out!

From my findings on the M4 Mini too is the thing is simple to take apart. This was important to me, I watched a tear down video prior to buying one. Because I plan on cracking mine open once a year to blow out the dust.

With that being said, I did manage to fix my M4 mini being woken up by my mechanical keyboard and the OWC1M2. Im not sure how I fixed it but I did yesterday. So happy that now I can use sleep mode on the Mini. I was never a fan of powering it on and off every time. Puts lots of thermal cycles on the Minis components.

 

[EugW]

After doing a lot of research on external drives ejecting from M4 Mini's. I am 99.99% confident that it's a power distribution problem with the Mini not being able to supply enough consistent power to the drives. I read a couple reviews online that said after getting a powered thunderbolt hub, it stopped the issue.

I really didn't want to spend the money but I have no choice if I want to stop my OWC 1M2 from ejecting. So I ordered a Kensington Thunderbolt 4 hub that comes with a 1500W power supply. That will be more than enough power for the OWC 1M2. I did consider getting the OWC TB4 hub for $30 less but some have said that when more than one drive is connected the speeds get cut in half. That's almost a deal breaker for me.

Plus the Kensington has a bit better reviews on amazon and an extra year of warranty (2 years vs 3 years). And Kensington has great customer support. Can't say the same for OWC. The hub will be here tomorrow and will report back with my findings. If I doesn't fix the issue then I will send it back and get the OWC hub. But I'm confident the Kensington will stop the ejections on the 1M2.

Running two drives, with one of them being a USB 4 enclosure + Samsung 990 Pro directly connected to the Mac mini, I didn't have those power issues you mention.

However, I currently run both off two separate Thunderbolt 4 hubs. The reason I run them through Thunderbolt hubs is because when directly connected to the Mac, they are seen as USB 4 drives, whereas when I run them through a Thunderbolt hub they are seen as Thunderbolt 3 drives. Why do I care? Because idle power draw is lower in Thunderbolt 3 mode than in USB 4 mode.

Why two Thunderbolt 4 hubs? Just because I happen to have two. (One was a warranty replacement for the other one, but it turns out the other one wasn't defective. It just had a defective cable.) Also, I figured it'd be good to have each on its own separate port for bandwidth reasons.

 

[DeepSix]

Running two drives, with one of them being a USB 4 enclosure + Samsung 990 Pro directly connected to the Mac mini, I didn't have those power issues you mention.

However, I currently run both off two separate Thunderbolt 4 hubs. The reason I run them through Thunderbolt hubs is because when directly connected to the Mac, they are seen as USB 4 drives, whereas when I run them through a Thunderbolt hub they are seen as Thunderbolt 3 drives. Why do I care? Because idle power draw is lower in Thunderbolt 3 mode than in USB 4 mode.

Why two Thunderbolt 4 hubs? Just because I happen to have two. (One was a warranty replacement for the other one, but it turns out the other one wasn't defective. It just had a defective cable.) Also, I figured it'd be good to have each on its own separate port for bandwidth reasons.

You have yours through TB hubs so it's being given enough power. I've yet to find a person online saying that their TB4 drive still randomly disconnects when running through a TB hub.

I read a lot of threads online and pretty much everyone who is having or had connection issues with their SSD, they're running it directly to the M4 Mini. Also I think the OWC 1M2 is drawing a lot of power, so that would explain why extra power from a hub fixes the issue.

 

[EugW]

You have yours through TB hubs so it's being given enough power. I've yet to find a person online saying that their TB4 drive still randomly disconnects when running through a TB hub.

I read a lot of threads online and pretty much everyone who is having or had connection issues with their SSD, they're running it directly to the M4 Mini. Also I think the OWC 1M2 is drawing a lot of power, so that would explain why extra power from a hub fixes the issue.

As mentioned, I was running the Samsung 990 Pro directly connected to the Mac mini initially, in a Qwiizlab USB4/TB enclosure. It was connected like this for about 6 weeks before I decided to run it through the TB 4 hub instead. See below.

This is connected directly to Mac mini. The drive is detected as USB 4. No random disconnects.



This is connected through a Thunderbolt 4 hub. The drive is detected as TB 3 and idles at lower power this way.

 

[DeepSix]

Read up on ACTIVE TB cables, I discovered this the other day. I saw a post where someone fixed their OWC 1M2 from disconnecting by getting an ACTIVE TB cable. It must be 6.6ft. All TB cables less than 6.6ft are passive, that includes the one that comes with the 1M2. I ordered one, it's worth a shot.

 

[EugW]

Did you guys measure your power draw?

Read up on ACTIVE TB cables, I discovered this the other day. I saw a post where someone fixed their OWC 1M2 from disconnecting by getting an ACTIVE TB cable. It must be 6.6ft. All TB cables less than 6.6ft are passive, that includes the one that comes with the 1M2. I ordered one, it's worth a shot.

? The original OWC 1M2 cable is 1 foot long. A good quality 1 foot cable should not be a problem for signal integrity. Are people with the disconnects using third party cables? Cuz if so, there are a lot of sub-par cables out there.

 

[Soondae]

Ah. My Hagibis and Qwiizlab are both APFS and are working fine.

Yes mine is the QwizzLab UH61-M4 with a Kingston NVM3 1TB installed