Display "Idle to On" transition resets USB Hub

Hi all,

I have been pulling my hair lately with a issue I am experiencing with the Spectrum… :frowning:

I love the idea of using 1 cable from my laptop to the monitor and use the built in USB hub to keep USB to Eth and other stuff in the back… this way, all cables are behind the monitor and I end up with a clean desk environment :slight_smile:

The issue is that once the display exits from IDLE to ON, it resets the USB hub and my USB Ethernet card is disconnected/connected and all my SSH sessions are dropped (along with all other ongoing internet connections).

I have upgraded to v104 and enabled the option found in:
Input/Output > USB Hub > Always On but the issue still happens.
Here is a guide to reproduce the problem:

Test Environment:

  1. Eve Spectrum with v104 loaded
  2. LGgram 16Z90P i7-1165G7

Test methodology:

  1. Connect the laptop to the Spectrum using USB C cable
  2. In the OS options “Power Options”, set “Turn Off Display” to “1 min” and “Put Computer to sleep” to “Never

Expected result:

  1. After 1min the display is turned off (soft off)
  2. USB hub is still powered on
  3. A key is pressed
  4. Display turns on

Observed result:

  1. After 1min the display is turned off (the Xe driver stops sending display signal to the Spectrum)
  2. After 1min more, the Spectrum shows “No signal” and powers off (soft off?)
  3. The USB hub is still ON (good) and all SSH sessions are running fine, the display is off (good)
  4. Press a keyb button or move the mouse
  5. The monitor is turning on
  6. Immediately the sound of USB disconnect/connect is heard
  7. I can see that am not connected to internet for 1-2 seconds (USB Ethernet interface coming up?)
  8. All my SSH sessions and ongoing downloads drop
  9. Connected again

What is strange is that the USB hub is reset not when the display enters IDLE mode or soft-off but when it is turning on again.

Another way to reproduce the same behaviour (although software and hardware-wise this methodology might be different) is simple to use the physical button to power the Spectrum off (the hub will remain on), then power it on (the hub will reset).

Thanks,
ER

6 Likes

That part that you mention in your Observed results, step 6 - I don’t actually keep anything connected to the Hub as I’m on a desktop, but if I already have my headphones powered on and receiving a signal from my PC, I, too, hear that exact sound.

Let me test on my end, I’m on 104 also, but I suspect your behavior is exactly what I’m gonna see also.

3 Likes

I’ve noticed similar behaviour. My mouse and keyboard have RGB lighting, and I can see the lighting turn off and back on again when the Spectrum wakes from sleep.

Hopefully this and the USB auto-selection bug I reported will get fixed in a future update.

4 Likes

This explains the problem I’m having here USB Hub appears to disconnect drives when monitor sleeps - when my Mac wakes it must disable the hub for a second which is ejecting the USB SSD.

1 Like

I have also observed this issue, with the USB hub appearing to reset upon wake from standby. I also observe USB disconnect sounds very occasionally while idle. Should note I have the USB hub set to remain powered when in standby.

2 Likes

Hi, @ERodrigues!

We appreciate that you documented in detail the behavior of USB hub disconnection when waking up from standby. Also, it is nice to hear about the observation from fellow community members. This phenomenon on our list for investigation later down the line.

2 Likes

How much later down the line? It has forced me to stop using Spectrum’s USB altogether.

Unfortunately, we cannot estimate the time it takes due to the complexity of resources involved in investigating, implementing, and testing such an improvement.

While not wanting to commit you to an outcome, I’d be happy with “looking at it in the next week or month” vs “best guess is several months out because we have a number of other things to focus on first.”

1 Like

Sure. Several months is probably closer to the outcome.

2 Likes

Hi @Lore_Wonder ,

On one hand I know and understand that at this stage timelines are likely not be realistic, but wouldn’t a list of issues on which Eve R&D confirmed to be “working on” be helpful and provide some hope to the community?

At least the ones that could comply with certain requirements like:

  1. Easy to reproduce (in my view this is key :innocent:)
  2. Known to have impact on several End-Users
  3. Has significant negative impact on User Experience

Thanks,

@Lore_Wonder

It has been 2 months since we discussed this issue and I have also raised it via email. There are other threads in the forum describing similar issue where the root cause should be the same as mentioned here.

This bug prevents people from using the HUB with HDD/SSD safely as well as maintain connections when using USB Ethernet adapters, just to mention 2 showstopper impacts.

I am not looking to get a estimate when this is fixed, but at the same time, having a high-end monitor like this one and one of its perks being unusable is frustrating.

Regards,
Edson

1 Like

I’m in the same position. I’m now used to reopen SSH sessions and VPN every time my monitor goes to sleep, which is annoying. I cannot imagine using an USB drive in that configuration considering the risk of corruption of the file system.

Seems like the USB hub (or the USB-C controller?) is doing something for whatever reason when the display is powering up.

I hope this issue will be strongly considered in the EVE roadmap and corrected.

My environment:

  • Spectrum with 105 firmware
  • Legion 5 pro using USB-C only (DisplayPort + USB 2.0)
2 Likes

Sorry @ERodrigues but still no ETA on this but we should have an update soon.

2 Likes

We’re testing to see if this has been fixed in new test builds of what will eventually be fw v106. Will let you know more after testing in the next couple of days.

1 Like

As an update - thus far we’re being told that this is a limitation of the hardware in the monitor, and that it always occurs when a new video signal is detected. Specifically, the PCB puts certain parts of the hardware to sleep when the monitor enters standby, but when it wakes up, it sends a reboot to all functions, instead of only the ones that were put to sleep.

Apparently, there is nothing that can be done about this because it is a direct hardware issue - as in the hardware used in the ES07D03 was designed in this way. Nevertheless, I’m still pursuing this one.

1 Like

Thank you for your reply.

Based on experience, I find it hard to understand the hardware limitation information provided to you by R&D but I understand that this is what we get and that’s it.

I am struggling not to be just ‘another guy’ ranting on a forum, speculating about solution A or B when in reality, this does not get this issue solved anyway… :slightly_frowning_face:

Nevertheless, it is substantially frustrating to have a high-end monitor like this one and be unable to use its USB hub properly so I find myself in a sad situation that maybe I have no other choice but to start looking for other alternatives which will likely include selling my Eve Spectrum :sob:

I guess I am ready to throw in the towel.

5 Likes

Well that’s unfortunate and a fairly bit oversight as it renders USB drives connected through the hub useless :frowning:

1 Like

Correction. It may render a USB drive connected to the Hub useless in the event that you are actively using it and the monitor goes to sleep and you wake it up. And even then, it won’t necessarily cause an issue.

If it is passive, i.e. just connected but not actually being used, it becomes available upon the reset. And I’ve left USB devices connected to my PCS through power off and restarts for decades now and still have every single one, going back as far as my tiny little 32 MB drive that basically holds a few text files now. I’ve lost all of 2 drives, one through my own stupidity in connecting it to a JTAG in a router that was disassembled because it was supposedly not getting power (it most certainly was!).

As for oversight - 100% agreed there. We should have known about this sooner, rather than later, but it is not something our fw dev team thought to check for, until we brought it up. Which was only after customers reported it.

We’re trying to make sure that this will not be the case moving forward.

2 Likes

Hi @nkyadav ,

My 2-cents:

Rebooting a computer with USB storage devices connected to it may be different from having those same USB Storage devices connected to a USB hub which suffers from a sudden HW reset (like the one with Spectrum).

Let’s say you are copying files or simply have a document opened straight from a USB Storage connected to the spectrum USB hub, what happens when there is a event sent from the host OS to the video card to set the monitor on IDLE mode? In theory, when the monitor is going back to “on” mode, there will be a reset of the USB hub (HW reset?!) so in this event, I believe that the chances of data corruption are high.

If you look at the observed results from my test methodology, my SSH sessions have packets flowing to maintain session (keep-alive), I understand that this happens on a higher layer (software) but the HW layer (let’s simplify and call it USB layer) should know that there is activity therefore a power off/on should never happen… however, it appears that the USB HUB implementation is completely transparent to these events even when the Always-On option is set to Enabled.

I love the Spectrum, it is really hard to let go of such a piece of HW/SW engineering, I got the “glossy is now available” email and … and my-oh-my I wanted to get one… but, i think it would just be stupid of me :frowning:

I still can’t for the life of me, understand why this cannot be fixed… IMHO, EvE has many battles to endure and this is just one of those that R&D will postpone as much as possible.

ER.

5 Likes