I am using intel 3dcamera with USB 3.0 port built in TX2.
However, there is a problem that the usb is disconnected intermittently.
(I have already posted about it several times before.)
So, I did various tests, and among them, the usb was not disconnected when connected using the powered USB 3.0 hub as below.
before: (USB 3.0 port <-> 3dcamera) (NG)
after: (USB 3.0 port <-> powered 3.0 HUB <-> 3dcamera) (GOOD)
In my opinion, the powered 3.0 HUB seems to amplify the signal strength.
Is there a tuning point to amplify the signal strength of USB 3.0?
You need do USB signal measurement when tring to tune it.
Is this related to USB contaction? Does your device move in use?
Are you try other USB3.0 device, Like disk for data transmission?
Basically, eye-diagram measurement was completed and passed for the four tuning values (HS_SLEW, HS_CURR_LEVEL, HS_SQUELCH_LEVEL, HS_DIN_DLY_SEL).
However, with the determined 4 tuning values, as mentioned in the thread, there is an intermittent usb disconnect issue in the intel 3dcamera.
So I asked for a value to control the drive strength, and in my opinion, HS_CURR_LEVEL seems to have an effect, so I tested it by changing it to minimum (0x3f), maximum (0x00), medium (0x1f), but still disconnected.
For reference, it was originally attached to a moving device, and is now reproduced in a fixed state.
In addition, the 3D camera was not directly mounted on the TX2 board, but it was not reproduced when the powered USB 3.0 hub was installed on the TX2 board and then the 3d camera was mounted on the hub.
HUB is a device that serves as a representative signal repeater.
So I suspected that the drive-strength wasn’t enough.
p.s: There is no external HDD we have, so we cannot test the mess data transmission.
The register(HS_SLEW, HS_CURR_LEVEL) is for USB2.0 tuning. Not helpful for USB3.0.
If the eye-diagram result is pass then this should be not related to signal quality.
Would you provide the failure log?
The USB hub may reduce the failure rate even it is not related signal quality.