USB downs on plugging devices / Hotplug and "cold" plug

[ThoLag]

Hi!

 

I've got a problem getting USB devices to work with Armbian 22.02.1 (Tinkerboard_bullseye_current_5.15.25) - fresh install, bare image without any modifications.

 

While a simple usb keyboard would more or less work, plugging some devices seems to crash the built-in USB hub.

This happens in "cold"-plug as in "hot"-plug, whereas there are slight differences.

 

Booting up without any USB device attached:

 

lsusb
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 002: ID 05e3:0608 Genesys Logic, Inc. Hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 003 Device 002: ID 0bda:49f6 Realtek Semiconductor Corp. USB Audio
Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

 

After hot-plugging in any USB device the Genesis USB Hub has disappeared:

 

lsusb
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 003 Device 002: ID 0bda:49f6 Realtek Semiconductor Corp. USB Audio
Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

 

dmesg has these two entries:

 

[  131.740593] usb 1-1: USB disconnect, device number 2
[  134.624629] usb usb1-port1: connect-debounce failed

 

So far what I tried to find out a little more:

 

• USB Card Readers and Datasticks work if "cold" plugged (=before power on), but crash on hotplug.
• USB Card Readers crash if hotplugged or "cold" plugged, removed and re-attached.
• USB stays crashed if the system reboots without a power off in between. Always!
• Huawei Surfsticks that do a modeswitch (e.g. USB_MODESWITCH etc. to get from  storage- to network-mode) crash the system. Maybe modeswitch triggers the same problem as hotplug?
• I tried Tinkerboard R2.0 and S R2.0 - and serveral of each from my collection. Same problem with every board.

 

Does anyone have more information on or suggestions how to analyse these issues?
I read hotplug was in issue in 2018, does this remain?

 

[Werner]

Providing logs with 

armbianmonitor -u

 helps with troubleshooting and significantly raises chances that issue gets addressed.

[jock]

I guess that many of these issues may come from a bad/unsuitable power supply and microusb connector that is not able to deliver proper current to USB when there is higher demand.

 

Are you feeding the tinkerboard with a microusb connector and a generic USB power adapter or via GPIO pins?

I have had no particular issues with a raspberry pi microusb power adapter rated at 5.1V/2A, but didn't really try demanding devices to USB ports

[ThoLag]

vor 19 Stunden schrieb jock:

Are you feeding the tinkerboard with a microusb connector and a generic USB power adapter or via GPIO pins?

I have had no particular issues with a raspberry pi microusb power adapter rated at 5.1V/2A, but didn't really try demanding devices to USB ports

 

Thank you, I am actually using the original Raspberry PI power adapter that delivers 5.1V/2A via MicroUSB.

I considered power issues, but the problem was so consistent and once the USB devices work they never failed in the further days of uptime where I experienced power problems to be more randomly.
At that point, Raspberry has the benefit of showing undervoltage via the rainbow-overlay.

But I will definately give it a try via GPIO and give you a feedback if it makes any difference.

[jock]

45 minutes ago, ThoLag said:

 

Thank you, I am actually using the original Raspberry PI power adapter that delivers 5.1V/2A via MicroUSB.

I considered power issues, but the problem was so consistent and once the USB devices work they never failed in the further days of uptime where I experienced power problems to be more randomly.
At that point, Raspberry has the benefit of showing undervoltage via the rainbow-overlay.

But I will definately give it a try via GPIO and give you a feedback if it makes any difference.

 

Yep! Usually there is a big dropout when the device is attached and there anything may happen; once the device is settled up the absorbed current usually gets stable and there are no more subtle changes. Of cource this is not true for all loads, but the most common cases are like this.

 

Anyway, if you have the chance to feed the board via solid and thick GPIO wires it would surely make difference in troubleshooting!