Fixing a USB 3.0 Hub

Posted by crimsonpisang on   0 comments   0 trackback

I've recently ordered a generic USB 3.0 hub from Ebay for under USD 5. As it turns out, I received a faulty unit. Whenever I've plugged the hub into my PC, Windows reported a power surge on that port. Seems like a short circuit somewhere in the hub.

DSC_0021.jpg

Upon opening up and inspecting for physical damage, I found nothing in particular so I've started probing for short. Turns out the short occurs in the cable itself. After desoldering the bad cable, next I needed to test if the hub was still functional. The hub controller was a GL3520 by Genesys Logic.

GL3520 block diagram

The block diagram shows the controller with both USB 2.0 and 3.0 hub section. Firstly, I've verified that the USB2.0 part is working by only connecting the 4 wires (Vcc, D+, D-, GND) and plugging into my PC. Windows detects the USB drive connected to it. Next, I need to test out the USB 3.0 part and in order to do that, I need a USB 3.0 type A cable/connector.
It's pretty interesting to see the insides of a good quality cable. The overall was surrounded by a foil and braided shield and each individual differential data pairs are further shielded. This was to prevent cross-talking which could affect the signal integrity, especially since USB 3.0 bit rate is 5 gbps.

DSC_0025.jpg

DSC_0024.jpg

Initially, the USB 3.0 hub managed to appear briefly on Windows. So, I thought it was just loose connection or not enough power. After some futile fiddling and a couple cups of coffee, then it suddenly hits me. Since it was going at 5 gbps, it might have to do with signal integrity. Adding more twist to the TX and RX pairs does seems to do the trick.
I used a neat software called USBTreeView by Uwe Sieber to verify the USB 3.0 device was working at super speed mode.

DSC_0028.jpg

USBtreeview.jpg


Post comment

Only the blog author may view the comment.

Trackback

trackbackURL:http://hwjunkie.blog138.fc2.com/tb.php/19-4bcb649f