[Solved] USB not working - Raspberry pi 2

[vitor]

Hi,
I purchased a rpi2 last month and used it only a couple times with the same setup everytime (same sdcard, same psu, same usb adapters plugged - wifi dongle and logitech wireless keyboard+mouse which always worked on both rpi 1 B and rpi 2 B until now.
A week ago I turned it on and somehow it isn't recognizing any usb devices I plug in.
Ever since it's been flashing red light once it boots. I know it's related to bad power supply, but I'm pretty sure power supply is good and I've tried others 2 psu which resulted in the same red flashing light on PWR led. I tried 4 different psu being three of them 5v/2a (oneplus one, nook hd and ipad) and one 5v/1A (apple).

Ethernet is working normally and I can SSH to rpi.
Also, I formatted sdcard changing from noobs with dual-boot to raspbian so I could have ssh always on.

When typing 'dmesg' I'm having this being flooded 4x/second:

[ 195.381316] usb 1-1-port2: over-current change

I'm not sure but I think lsusb is reporting correctly:

Code: Select all

pi@raspberrypi ~ $ lsusb
Bus 001 Device 002: ID 0424:9514 Standard Microsystems Corp.
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 003: ID 0424:ec00 Standard Microsystems Corp.
pi@raspberrypi ~ $ lsusb -t
/:  Bus 01.Port 1: Dev 1, Class=root_hub, Driver=dwc_otg/1p, 480M
    |__ Port 1: Dev 2, If 0, Class=hub, Driver=hub/5p, 480M
        |__ Port 1: Dev 3, If 0, Class=vend., Driver=smsc95xx, 480M

dmesg | grep usb

Code: Select all

pi@raspberrypi ~ $ dmesg | grep usb
[    1.279697] usbcore: registered new interface driver usbfs
[    1.285377] usbcore: registered new interface driver hub
[    1.290848] usbcore: registered new device driver usb
[    1.655843] usbcore: registered new interface driver smsc95xx
[    2.178820] dwc_otg bcm2708_usb: DWC OTG Controller
[    2.185306] dwc_otg bcm2708_usb: new USB bus registered, assigned bus number 1
[    2.194190] dwc_otg bcm2708_usb: irq 32, io mem 0x00000000
[    2.211805] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002
[    2.220212] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[    2.229032] usb usb1: Product: DWC OTG Controller
[    2.235276] usb usb1: Manufacturer: Linux 3.18.11-v7+ dwc_otg_hcd
[    2.242986] usb usb1: SerialNumber: bcm2708_usb
[    2.261635] usbcore: registered new interface driver usb-storage
[    2.373206] usbcore: registered new interface driver usbhid
[    2.381445] usbhid: USB HID core driver
[    2.657925] usb 1-1: new high-speed USB device number 2 using dwc_otg
[    2.868231] usb 1-1: New USB device found, idVendor=0424, idProduct=9514
[    2.876821] usb 1-1: New USB device strings: Mfr=0, Product=0, SerialNumber=0
[    3.178070] usb 1-1.1: new high-speed USB device number 3 using dwc_otg
[    3.298278] usb 1-1.1: New USB device found, idVendor=0424, idProduct=ec00
[    3.307281] usb 1-1.1: New USB device strings: Mfr=0, Product=0, SerialNumber=0
[    3.383319] smsc95xx 1-1.1:1.0 eth0: register 'smsc95xx' at usb-bcm2708_usb-1.1, smsc95xx USB 2.0 Ethernet, b8:27:eb:b6:2d:2a
[    3.399528] usb 1-1-port2: over-current change

Anything I should try?
Thank you in advance.

[FTrevorGowen]

...
Ever since it's been flashing red light once it boots. I know it's related to bad power supply, but I'm pretty sure power supply is good and I've tried others 2 psu which resulted in the same red flashing light on PWR led. I tried 4 different psu being three of them 5v/2a (oneplus one, nook hd and ipad) and one 5v/1A (apple).
...

When you tried the different power supplies were you (always) using the same uUSB cable to connect them to the Pi? (If the cable's faulty, changing the PSU won't help).
What happens if you try booting w/o anything connected to the USB ports? (Does the red PoWeR LED still flash. Can you SSH in via the network?)
Trev.

[vitor]

Thank you for your answer.
Actually I managed to fix the red flashing issue by changing to a 3rd cable. (28awg). I got some help on IRC channel.

The Ethernet port is working fine, so even if pwr led is flashing I can ssh to it.

Even now that pwr led isn't flashing but turned red all the time, I still got the "usb 1-1-port2: over-current change" message flooded in dmesg output. I'm having this issue with or without anything attached to usb port.

[FTrevorGowen]

Thank you for your answer.
Actually I managed to fix the red flashing issue by changing to a 3rd cable. (28awg). I got some help on IRC channel.
The Ethernet port is working fine, so even if pwr led is flashing I can ssh to it.
Even now that pwr led isn't flashing but turned red all the time, I still got the "usb 1-1-port2: over-current change" message flooded in dmesg output. I'm having this issue with or without anything attached to usb port.

O.K. - some good news then. I'm not sure about the "over-current change" message/issue. To rule out "SD card (file) corruption" due the previous "brownouts" can you write a clean, fresh, O.S. (Raspbian) image to another card and report back. Also, had you enabled the higher USB "maxpower" option in config.txt (so that the Pi could provide up to 1200mA** shared between the four USB ports as apposed to the default of only 600mA)?
Trev.
** When powered with a suitable PSU and good connecting cable

[vitor]

I only have 5v/2A power supplies from smartphones, so I'm not sure if I'm able to allow 1200mA max power. Also, I don't know how to do this. Should I try it?

I just writed raspbian to another sdcard as you suggested but it didn't fix anything. USB not working, "usb 1-1-port2: over-current change" dmesg output, red light normal, not flashing.

Should I try increasing usb port output to 1200mA even with a 2a only psu?

[FTrevorGowen]

I only have 5v/2A power supplies from smartphones, so I'm not sure if I'm able to allow 1200mA max power. Also, I don't know how to do this. Should I try it?
I just writed raspbian to another sdcard as you suggested but it didn't fix anything. USB not working, "usb 1-1-port2: over-current change" dmesg output, red light normal, not flashing.
Should I try increasing usb port output to 1200mA even with a 2a only psu?

Yes - setting up that option doesn't mean that 1200mA will be drawn through the USB ports (IIRC the B+ and P2B's polyfuse limits the total current available, including that for the Pi, as well as via the USB ports to 2.1A in any case) - any attached USB devices (assuming they're not faulty) will only draw the current they need**. Given that most wifi dongles can need 500mA+ (whilst scanning for example) and some wireless keyboard (dongles) may need 100mA+ then only allowing 600mA is a marginal condition.
Trev.
** The lsusb and usb-devices commands can provide some indication of such as shown in these examples:
http://www.cpmspectrepi.webspace.virgin ... erReq.html
http://www.cpmspectrepi.webspace.virgin ... rives.html
http://www.cpmspectrepi.webspace.virgin ... pters.html
http://www.cpmspectrepi.webspace.virgin ... Dlist.html

[vitor]

I only have 5v/2A power supplies from smartphones, so I'm not sure if I'm able to allow 1200mA max power. Also, I don't know how to do this. Should I try it?
I just writed raspbian to another sdcard as you suggested but it didn't fix anything. USB not working, "usb 1-1-port2: over-current change" dmesg output, red light normal, not flashing.
Should I try increasing usb port output to 1200mA even with a 2a only psu?

Yes - setting up that option doesn't mean that 1200mA will be drawn through the USB ports (IIRC the B+ and P2B's polyfuse limits the total current available, including that for the Pi, as well as via the USB ports to 2.1A in any case) - any attached USB devices (assuming they're not faulty) will only draw the current they need**. Given that most wifi dongles can need 500mA+ (whilst scanning for example) and some wireless keyboard (dongles) may need 100mA+ then only allowing 600mA is a marginal condition.
Trev.
** The lsusb and usb-devices commands can provide some indication of such as shown in these examples:
http://www.cpmspectrepi.webspace.virgin ... erReq.html
http://www.cpmspectrepi.webspace.virgin ... rives.html
http://www.cpmspectrepi.webspace.virgin ... pters.html
http://www.cpmspectrepi.webspace.virgin ... Dlist.html

Thanks! I'll try to find how can I make my rpi output more amps to usb ports and verify how much amps my usb devices need.


Also, searching a bit more I found this thread and follow what he did on 2nd post.
So I also tested the voltage (as described in viewtopic.php?p=628505#p628505) and can confirm the same results, except for PP27:
(always against PP3=GND)
PP1: 5.11V
PP2: 5.11V
PP7: 5.05V
PP8: 3.30V
PP9:1.80V
PP10:1.80V
PP27: 0.5V-0.15V (going from 0.5V to 0.15V).

As described in the original thread (2nd thread linked), PP27 is attached to VBUS (USB 5V power)
Maybe this is why I'm having this issue. Anything shorting maybe?

[vitor]

I know almost nothing about electrical things, but I was trying to understand why my rpi2 was having this usb-port over current issue.
So, I got a multimeter and found out that 5V USB (PP27) is being shorted to GND. Tested both PP3 and GPIO pin 06. I did this test using 200Ω as scale in multimeter while my rpi was turned off and it displayed 02,1.

According to @mahjongg 's post, PP27 is supposed to be 5v USB and PP3 to be GND as well as GPIO 06.

Is it possible to track and fix it? Or is it damaged and there's no way to be fixed? There is probably a voltage regulator shutted off while 5v and GND are shorted.
I saw that there isn't any rpi2 schematics but it's very similar to B+, but I didn't manage to find any usb schematics.
Ethernet port is working fine.

I'm not willing to send it to warranty since I purchased it in a different country so I could get it cheaper. Sending it to replacement would cost a new one price equivalent.

Appreciate any advice!

[Burngate]

Is it possible to track and fix it? Or is it damaged and there's no way to be fixed? There is probably a voltage regulator shutted off while 5v and GND are shorted.
I saw that there isn't any rpi2 schematics but it's very similar to B+, but I didn't manage to find any usb schematics.
Ethernet port is working fine.

There's a partial diagram in viewtopic.php?f=29&t=88789 which may help.

Between the main 5v rail and the USB is a AP2553 chip that's reporting a fault to the hub chip, as well as preventing the short-circuit from pulling the main 5v down.
It's possible that the USB sockets are damaged; if so the damage might be visible and repairable.

[drgeoff]

Have you looked into the USB sockets to see if any of the contacts inside have come loose and got pushed against the metalwork?

If you are really seeing 2.1 ohms between USB 5 volt pins and ground that is really low. I suspect it will be even lower if you subtract the reading you get on your meter when the probes are joined together. That should give 0.00 but rarely does with budget class meters.

[vitor]

There's a partial diagram in viewtopic.php?f=29&t=88789 which may help.

Between the main 5v rail and the USB is a AP2553 chip that's reporting a fault to the hub chip, as well as preventing the short-circuit from pulling the main 5v down.
It's possible that the USB sockets are damaged; if so the damage might be visible and repairable.

If you are really seeing 2.1 ohms between USB 5 volt pins and ground that is really low. I suspect it will be even lower if you subtract the reading you get on your meter when the probes are joined together. That should give 0.00 but rarely does with budget class meters.

Thank you both Burngate and drgeoff for your suggestions. At least there's still hope.
I think all USB ports are ok, unfortunately. It'd be a lot easier to fix it if they were shorting.



Maybe I should verify AP2553 but I don't know how to. I found a bit hard to understand those schematics, but I'll try to take it to a guy who knows it better than me and see if he can found where it's shorting.
So, AP2553 is referred in rpi board as U13, right?


When I try to test PP3 (GND) against GPIO pin 6 (GND), I get 01,0, just as PP7 (5v after fuse) against GPIO 2 (5V), and as when I join both probes together.
This is the multimeter I used. It's indeed very cheap.


Thank you again.

[rpdom]

That multimeter should provide a reasonably accurate value for general measurements. I've got one the same as a spare for when I just want to do a quick test and can't be bothered to dig out my decent meter, or when I need to measure two things at the same time.

You could always check it against a known resistance value.

[vitor]

Since I don't know much about electrical, I took it to a guy who understands it a lot more than I do. I showed him the USB schematics and he founds out that C97 was damaged. He replaced it then PP27 (USB 5V) isn't being shorted to GND anymore. He told me if I could find a replacement to ap2553 he could replace it and it'll probably be fixed. He thinks the usb regulator is damaged because I tried to use a powered usb hub to see if usb ports would work.

So, do anyone knows where I can find an ap2553 or equivalent to be purchased? I found in amazon but it was a lot expensive, like $15 shipped.

[Burngate]

I don't know a better place to get an ap2553, but (being something of a cowboy) if it were mine, I'd just remove it and wire the 5v straight to the output.
Having converted your nice modern Pi 2B into an antiquated B v.2, you would lose the protection afforded by that chip, but plenty of us thought they were ok like that

[vitor]

I don't know a better place to get an ap2553, but (being something of a cowboy) if it were mine, I'd just remove it and wire the 5v straight to the output.
Having converted your nice modern Pi 2B into an antiquated B v.2, you would lose the protection afforded by that chip, but plenty of us thought they were ok like that

lol! I liked that! I asked it but he advised me to not bypass this chip. Who knows if there's an issue somewhere else. I had a rpi model B and have never experienced this issue, fortunately.

I found out digikey is selling AP2553W6 for $1 + shipping. I think it's very reasonable. I might be getting it in mid may and will report if all goes fine!

Thank you all very much for helping!

[vitor]

Just a small update.
C97 is out of the board. The guy who helped me removed it, placed it again to see that the short between PP27 and GND is over, then removed again until I get AP2553 replacement.

Since Burngate told I could bypass it, I tried to use a powered hub (actually, I don't have one, but a USB-Y splitter, so I can power it by another port). I think AP2553 was never damaged, but its switch was OFF. I was still experiencing usb 1-1-port2: over-current change.

After plugged the USB-Y with a usb mouse I got this:

[ 133.401276] usb 1-1-port2: over-current change
[ 133.657310] usb 1-1-port2: over-current change
[ 133.913340] usb 1-1-port2: over-current change
[ 134.169379] usb 1-1-port2: over-current change
[ 136.448537] usb 1-1.2: new low-speed USB device number 4 using dwc_otg
[ 136.553401] usb 1-1.2: New USB device found, idVendor=093a, idProduct=2510
[ 136.553417] usb 1-1.2: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[ 136.553427] usb 1-1.2: Product: USB OPTICAL MOUSE
[ 136.553437] usb 1-1.2: Manufacturer: PIXART
[ 136.559584] input: PIXART USB OPTICAL MOUSE as /devices/platform/bcm2708_usb/usb1/1-1/1-1.2/1-1.2:1.0/0003:093A:2510.0001/input/input0
[ 136.559981] hid-generic 0003:093A:2510.0001: input,hidraw0: USB HID v1.11 Mouse [PIXART USB OPTICAL MOUSE] on usb-bcm2708_usb-1.2/input0
[ 149.019507] usb 1-1.2: USB disconnect, device number 4

Now usb 1-1-port2: over-current change has disappeared! It doesn't show anymore and my rpi USB ports works the same way it used to without having to replace AP2553.

THANK YOU VERY MUCH!!


--EDIT
Actually I replied too fast. Once I removed the USB-Y cable, if the Pi is plugged in, the port will remain working, but if I reboot it, dmesg error will come back and USB port won't work unless I plug USB-Y cable again.
At least it's working and I think replacing AP2553 might work just as it used to since the beggining.

[vitor]

Just a small update.
Now my Pi is fully fixed. USB ports are working just like when I received it. Got the AP2553 for $4 shipped from digikey. So, after replacing both C97 (which shorted and damaged U13) and U13 for working ones my pi's usb ports are working again without and usb powered hub or Y cable.

[dettofatto]

I removed C97 from the board and the problem went away and all 4 USB ports now work even without C97 in the circuit!!!!!!!!!!
Thank you!
https://app.box.com/s/1rqa21xknc0px86sm0nm5o52dh3ddyio

[schmorrison]

I removed C97 from the board and the problem went away and all 4 USB ports now work even without C97 in the circuit!!!!!!!!!!
Thank you!
https://app.box.com/s/1rqa21xknc0px86sm0nm5o52dh3ddyio

I can confirm this 'fixes' my RPi 2 as well.

I was having the same problems with all manner of power supply and USB cables. Regardless of how the Pi2 was started, headless, or with attached peripherals, the red light would blink. If I managed to attach a network device or monitor+keyboard without a "Brownout", the dmesg output would be flooded with the "Over current change" message.

Removing C97 didn't work until I switched to the power supply and cable I thought were the best but hadn't worked correctly while C97 was still connected(1.5A and 3ft. 28/24AWG). Kept issuing "shutdown -r" until I was satisfied the dmesg output was stable. Have been running the unit for +24 hours without a "brownout" with ethernet and a USB wifi device connected at startup.

Thanks for the help,

Schmorrison

[jcmygod]

Hi, just wanted to contribute to this post as I just got around this problem by bypassing the AP2553 chip as suggested by Burngate.

I have a 3 weeks old Pi 3 model B v 1.2, it has a wireless USB and a couple of RTL-SDR dongles connected to the USB ports. A couple of nights ago returning from work I noticed none of the USB were working on the Pi. Did a little Google and found this site to be most useful.

/var/log/messages had this error flooding the log: usb 1-1-port2: over-current change

Per suggestions from other posts, tried removing C97 didn't help, it is likely the AP2553 chip had failed but I didn't want to wait to get a replacement so I removed the chip and bypassed the circuit by connecting pin 1 (in) to pin 6 (out) on the pads. USB now working, I might still get a replacement chip and reinstall it later.

The spec of the chip is here, Pi 3 uses the SOT25 package:
http://www.diodes.com/_files/datasheets/AP255x.pdf

[ali_chen]

Bypassing the upside power switch might be not a totally good idea. The logs indicate that the system has received the OVERCURRENT event. The LAN9512 combo device (hub + LAN) has a special (patented) way to sense the port OCS event using a single pin (PRTCTLx), to save on pin count. This pin controls enabling of AP2552 switch (EN), and simultaneously senses the FAULT signal from the switch. I have not found the USB part of Pi3 schematics, but I guess that the Pi design follows the Microchip recommended hookup where EN and FAULT pins are connected to the same PRTCTL pin. This connectivity has some flakey behavior, and was not tested on all switches. I would suggest to disconnect (lift up or cut) the pin3 of AP2552 instead of bypassing of the entire switch.

[nzppak]

im also suffering this issue which kind of tradesmen would best deal with soldering and removing these chips?

[FUHGETTABOUTIT]

I tried this and I still get the same error. Can anyone else verify this workaround?

[jbeale]

The same thing happened to my RPi 3. Removing C97 did not help. Removing U13 pin 3, and shorting U13.1 to U13.6 fixed the problem. See also:
viewtopic.php?f=28&t=162103

[ahmedawad]

I took out U13 (BK5tE) chip and the Capacitor C97 completely out of my board and shorted pin 1 and pin 6 of the U13. The Pi became stable once again and the power red led stopped blinking. Thanks for @jbeale, his thread and discussion was helpful.

I would suggest, first use an Avometer and while the Pi is Off, try using the buzzer and check the 2 pins of C97. If sound comes out then the problem is with the capacitor, take it out and try again. If the problem still exists, then take the U13 chip out and make the short. This what happened on Pi3 and it did work.