We've got some devices out in the field and have suddenly started noticing connection problems over the J9 connector. One clue is that the D2 "USB Power Good" doesn't always illuminate and that seems to correspond with a failure to be able to communicate with the board over ADB or using a network dongle.
Any suggestions as to what we can check to what is going wrong?
Could this be a software problem, or hardware?
Thanks.
@tom Yes sir.
Voxl 2 mini J9 USB port <=> Android Phone attempting to use accessory mode to directly communicate without additional external hardware for networking or similar.
I've used that port as the host for USB storage previously, ADB, etc. but this use is resisting our efforts.
@Moderator for others who find the question and need an answer.
I believe that @mkriesel managed to resolve the issue when he figured out that driver name needed to be changed depending on which connector he wanted to connect the camera to:
com.qti.sensormodule.imx678_0.bin for J6
com.qti.sensormodule.imx678_2.bin for J7
This doesn't seem to be mentioned in the documentation, or at least we couldn't find it mentioned.
Cheers,
Brendon.
We've been trying to get a Samsung phone to connect to the VOXL 2 Mini using J9 and have so far been failing to get the connection setup.
We are using libusb on the VOXL and the same test program that we developed on a Ubuntu Laptop. We can see the USB connection in syslog, but the messages are identical if our Accessory Mode program is running, or if it isn't running.
Questions:
Is there any hardware reason that would prevent accessory mode from working?
Do we need any special software or configuration to make this work?
https://source.android.com/docs/core/interaction/accessories/protocol
Thanks,
Brendon.
@Moderator Can you take a look at my new post.
I can see that the mount command in syslog. No errors, but the mounted file system doesn't show up.
Thanks,
Brendon.
Here is the output from Syslog:
Mar 2 12:59:59 m0104 kernel: [ 120.224638] usbpd usbpd0: Type-C Sink connected
Mar 2 12:59:59 m0104 kernel: [ 120.225257] qpnp-pdphy c440000.qcom,spmi:qcom,pm8150b@2:qcom,usb-pdphy@1700: Linked as a consumer to regulator.90
Mar 2 13:00:00 m0104 kernel: [ 120.300716] msm-dwc3 a600000.ssusb: DWC3 exited from low power mode
Mar 2 13:00:00 m0104 kernel: [ 120.303802] usb usb3: New USB device found, idVendor=1d6b, idProduct=0002, bcdDevice= 4.19
Mar 2 13:00:00 m0104 kernel: [ 120.303816] usb usb3: New USB device strings: Mfr=3, Product=2, SerialNumber=1
Mar 2 13:00:00 m0104 kernel: [ 120.303826] usb usb3: Product: xHCI Host Controller
Mar 2 13:00:00 m0104 kernel: [ 120.303834] usb usb3: Manufacturer: Linux 4.19.125 xhci-hcd
Mar 2 13:00:00 m0104 kernel: [ 120.303843] usb usb3: SerialNumber: xhci-hcd.1.auto
Mar 2 13:00:00 m0104 kernel: [ 120.310048] usb usb4: We don't know the algorithms for LPM for this host, disabling LPM.
Mar 2 13:00:00 m0104 kernel: [ 120.310519] usb usb4: New USB device found, idVendor=1d6b, idProduct=0003, bcdDevice= 4.19
Mar 2 13:00:00 m0104 kernel: [ 120.310531] usb usb4: New USB device strings: Mfr=3, Product=2, SerialNumber=1
Mar 2 13:00:00 m0104 kernel: [ 120.310540] usb usb4: Product: xHCI Host Controller
Mar 2 13:00:00 m0104 kernel: [ 120.310549] usb usb4: Manufacturer: Linux 4.19.125 xhci-hcd
Mar 2 13:00:00 m0104 kernel: [ 120.310557] usb usb4: SerialNumber: xhci-hcd.1.auto
Mar 2 13:00:00 m0104 kernel: [ 120.311521] hub 4-0:1.0: USB hub found
Mar 2 13:00:00 m0104 kernel: [ 120.311673] hub 4-0:1.0: 1 port detected
Mar 2 13:00:00 m0104 kernel: [ 120.628570] usb 4-1: new SuperSpeed Gen 1 USB device number 2 using xhci-hcd
Mar 2 13:00:00 m0104 kernel: [ 120.646650] usb 4-1: New USB device found, idVendor=105d, idProduct=0009, bcdDevice= 1.00
Mar 2 13:00:00 m0104 kernel: [ 120.646664] usb 4-1: New USB device strings: Mfr=1, Product=3, SerialNumber=2
Mar 2 13:00:00 m0104 kernel: [ 120.646674] usb 4-1: Product: U432TNJGR-XN000-D
Mar 2 13:00:00 m0104 kernel: [ 120.646683] usb 4-1: Manufacturer: Delkin Devices
Mar 2 13:00:00 m0104 kernel: [ 120.646692] usb 4-1: SerialNumber: 29910500000000000065
Mar 2 13:00:00 m0104 kernel: [ 120.649173] usb-storage 4-1:1.0: USB Mass Storage device detected
Mar 2 13:00:00 m0104 kernel: [ 120.650176] scsi host1: usb-storage 4-1:1.0
Mar 2 13:00:01 m0104 kernel: [ 121.665628] scsi 1:0:0:0: Direct-Access DDINC USB DISK 3.1 1809 PQ: 0 ANSI: 6
Mar 2 13:00:01 m0104 usbmount[2351]: loaded usbmount configurations
Mar 2 13:00:01 m0104 usbmount[2351]: creating /var/run/usbmount directory
Mar 2 13:00:01 m0104 usbmount[2351]: trying to acquire lock /var/run/usbmount/.mount.lock
Mar 2 13:00:01 m0104 usbmount[2351]: acquired lock /var/run/usbmount/.mount.lock
Mar 2 13:00:01 m0104 usbmount[2351]: /dev/sdg contains filesystem type ext4
Mar 2 13:00:01 m0104 usbmount[2351]: mountpoint /Storage is available for /dev/sdg
Mar 2 13:00:01 m0104 usbmount[2351]: executing command: mount -text4 -osync,noexec,nodev,noatime,nodiratime /dev/sdg /Storage
Mar 2 13:00:01 m0104 kernel: [ 121.883601] EXT4-fs (sdg): recovery complete
Mar 2 13:00:01 m0104 kernel: [ 121.883626] EXT4-fs (sdg): mounted filesystem with ordered data mode. Opts: (null)
Mar 2 13:00:01 m0104 usbmount[2351]: executing command: run-parts /etc/usbmount/mount.d
Mar 2 13:00:01 m0104 usbmount[2351]: usbmount execution finished
But after this completes, the drive is not mounted on /Storage is not mounted and /dev/sdg doesn't appear under df
Thanks for any insight.
@Eric-Katzfey Fantastic. We had an engineer pull this down on Saturday and so far it seems to work. I'll give you an update if we run into issues or are able to do the flight tests using this UART instead of the workaround.
Thanks very much guys, it looks like you saved us a lot of rework.
@Eric-Katzfey We are all setup to do a few high profile test flights which absolutely need a 1 Mbaud UART available, and ideally a second lower speed one, but we can work around that for now. The final version needed for test flights in October needs both.
A separate effort we are working on needs just a single UART and would benefit from a second. That is lined up for delivery by mid August.
So anything you'd be able to share that can get us closer to our goals would be very helpful and faster than the parallel workarounds we are exploring. @Chad-Sweet has been briefed on our use cases.
@Eric-Katzfey I don't need PX4 running on this board at all, so it would be fine to change the configuration.
Can you share how how we can change the configuration to get that working?
Thanks!
I just wanted to confirm that the Voxl 2 mini doesn't have any UARTs available on the Linux side of the OS?
We need to programmatically access two UARTs, with one needing to be a high speed UART. We are very space constrained, which is why we are using the Voxl 2 mini, so really need to get access to the available UARTs without add on boards.
In the event that the lpsi ones are the only ones that can work, what is the timeline for them to be accessible from Linux?
Thanks guys.
@Chad-Sweet , the documentation for the Voxl 2 mini is quite a bit more difficult to negotiate. For example: Voxl 2 guide vs voxl 2 mini guide There are quite a few similar examples where finding the information requires a lot more work.
@Moderator I'll take a look at that code. Can you tell me which UARTs currently are available as tty devices?
I may be able to rewiring things to use different ports while I wait.
Thanks.
I'm trying to communicate with some external devices using two of the UARTs in J19, but can't find any documentation on how the physical UARTs are mapped to the tty devices.
In particular, I need to use both GNSS and ESC serial ports from C and/or as a login console.
Thanks.
@Moderator Thank you, that is fantastic news, and makes my life a lot easier.
Good morning. Looking in the documentation I see mention on using the J9 connector as a secondary HOST port.
My plan is to use one of the UARTs as a tty to access the device using a shell to debug our software and/or to toggle between HOST mode and the default mode.
What I can't find is a tutorial or specific steps I need to take to be able to change the port into HOST mode and back to peripheral mode. In particular:
Any examples would be fantastic.
Thanks,
Brendon.