2018 Mac Mini No Sleep With Catalina And Dell U2415 USB Hub

If I plug in the line from the computer to the monitor's hub, the mini will not sleep. If I unplug that line, the mini will sleep. This is true whether anything is attached to the hub or not. Even more interesting, if I initiate sleep with the line attached, the computer fails to sleep, but if I then detach the cord -- without doing anything else at all -- the computer will go to sleep normally.

If, on the other hand, I plug in either of my other two hubs (one is USB3, one is USB2) the computer will sleep normally, whether there are devices attached to them or not..

If a computer does not sleep, and I actually do want it to sleep, I think that constitutes a problem. Of course, it would also be nice if it would sleep while attached to the hub on the Dell monitor (which in all other respects performs admirably), However, I do not currently have a method to make that work. I further suspect others may see this problem.

I could post a snapshot of my Energy settings (currently, "put hard drives to sleep when possible" is the only thing checked) but it would be irrelevant. During testing I did all 32 possible combinations, along with the two possible USB settings on the monitor. None of them had any effect. Without the hub attached, sleep is accomplished normally with any of the settings checked. It will also initiate sleep properly when the monitor goes to sleep (I also changed that setting to several values, including "never", also with no effect). But again, if I plug in the line from the computer to the hub, the computer will not go to sleep with any of the normal triggers.

You may wish, in light of this, to review your first sentence.

The Kill-a-watt is a device that measures the power consumption at the wall; it is not an app. It's hardware. When the mini sleeps, it consumes about .5 watts (there is some small fluctuation but that's a good average). When the mini is operational, it averages about 12 watts used, jumping all over the place depending on demand. Thus it is very simple to tell whether the computer is actually sleeping regardless the status of any other component.

The Kill-a-watt does not, however, present one with snapshots.

My bad for not explaining what it is; they're a common tool in testing. If all this is too confusing, feel free to leave the bandwidth open for anyone who sees the same problem (or even better, a solution).

Okay, think I found it. The problem appears to be when the hub is daisychained off a USB-C port, which in theory should be okay. I assumed the hub in the monitor is powered, but maybe not. And different controllers act differently; some are downright weird. The problem connection was a hub attached to the USB-C port, which looks to be supported by a GenesysLogic controller. Now the monitor hub is attached directly to one of the standard ports, which look to have a TI controller. At any rate, when run straight from a standard port, sleep works. Unfortunately, I don't have an adapter that would allow me to run the monitor hub from one of the USB-C ports, so I don't know if it's the daisy chain or the controller. Or, for that matter, the USB-C 3.1 hub.

Have the problem isolated to ports 7&8 on the monitor hub. It will require someone with a more thorough knowledge of USB topology to say "why."

Neither System Information nor lsusb -v were particularly useful on Mac. But I was able to make progress using the excellent uhubctl utility. The good news is the Dell U2415 hub is "smart" and can be controlled from the command line.

To make a long and tedious story much shorter, the hub will operate all usable ports and sleep properly when ports 7 and 8 are turned off. If you have such a monitor (or other similar Dell) you might be wondering where those ports are, since there are only five accessible. I wonder the same thing.

Internally, there are two hubs visible with, as you might expect, eight ports. Port 4 on the primary hub controls access to the secondary hub, also perfectly normal. The first two ports on the secondary hub control the two non-adjacent sockets. The other two ports are a complete mystery, but if they are on and have power, the Mac Mini 2018 will not go to sleep. Turn them off and everything is great.

For a truly odd side note, if you activate all of the ports on the primary hub (or any of them) without activating a port on the secondary hub, the system (computer and monitor) uses about 2 watts when sleeping. If you then turn on port 5, or 6, or both ... the system will sleep using about 1.25 watts.

Wrong again. That only worked for one iteration, then back to the old behavior. Now I'm really confused.