OpenWrt Forum Archive

Topic: wr703n hub problem

The content of this topic has been archived on 1 Apr 2018. There are no obvious gaps in this topic, but there may still be some posts missing at the end.

Hi,
I used to have a bootable openwrt uSD with this configuration:
wr703n <--> usb-hub <--> microSd (with openwrt)

One day it stopped to boot from uSD and the usb stick started blinking fast when I connected it to the hub. Also, booting stopped and I get this response from 'dmesg':

[ 2219.700000] hub 1-1:1.0: connect-debounce failed, port 1 disabled

I then detached the usb stick from hub of wr703n and connected it to my computer running Ubuntu. There I did an fdisk and then mkfs. All ok, I checked the uSD with no error.

Back to wr703n: the same thing happened.

I detached the hub (which is recognized as true high speed hub) and directly connected the usb stick to wr703n: now in openwrt I have it recognized with no problem.

I have no idea, how to further trace this error.

Any help please?

I just got an idea of changing the usb cable that connects my usb stick to the hub.

2 wires gave the same result.

I attached the 2rd one which is a webcam usb cable and finally the uSD is recognized with no problem!

Now I still need help to understand, how those cables could make the problem?

I use one of them to connect an avr programmer to computer when programming MCU, and the other is a mobile charging cable. When a mobile is connected with that cable to the computer and the mobile contains this same uSD, it is recognized by the computer with no problem.

Could anyone please give an idea?

Some chinese cables have really, really crappy, thin wire.  Often, it's actually not stranded copper, it's strands of aluminium, plated with a very thin layer of copper.

What this means:

For a cable that never worked, the resistance on the power and ground wires might be too high to provide the required current with 5 volts at the other end.  The phone may tolerate any voltage because the battery charger will try to pull only as much current as it can without the input supply sagging too much, and if it can't pull much, it can simply run from battery anyway.

For a cable that used to work but doesn't now, one of the data wires may have simply broken partially inside, under a strain relief. With fewer strands connecting one end on only one of the wires in the LVDS pair, the resistance increases in one wire without a corresponding increase in the other, and can be too unbalanced for the AR9331's known-to-be-buggy USB block to communicate with it properly.

Thank you very much for details.
I first could hardly believe that it is only the problem of cable. Now after having a whole day lost, but having my router booting up from memory card with 1 healthy cable I found in my room, I'm pretty sure it is the health of cable.

Also, LED of usb stick lights much brighter than when with those other cables.

Cheers OpenWrt!

The discussion might have continued from here.