Bricked eXplorer ?

Technical questions regarding the xTIMEcomposer, xSOFTip Explorer and Programming with XMOS.
psupine
Member++
Posts: 22
Joined: Wed May 16, 2018 7:28 am
Location: Australia

Bricked eXplorer ?

Postby psupine » Thu Nov 07, 2019 6:05 am

My XTAG3 has stopped talking to the xCORE-200 eXplorer board.

There is no obvious reason for this. Same laptop/Win7/USB ports, same cables. I hadn't touched it for a month or so (while building something else) and now it won't play. It was powered down for that time.

The XTAG3 is showing up in the device list, and the status LEDs indicate that target is not detected (D1 is red). The eXplorer has D6 and D8 lit green and the 5V and 3V3 check out.

Has anyone brought a board back to life from this state? [[sad]]

EDIT: It's worse than I thought. None of the international distributors hold stock and the factory lead time is 12 weeks, so I really do have to fix it.
User avatar
mon2
XCore Legend
Posts: 1765
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Thu Nov 07, 2019 10:15 am

Hi. I think the communication is not solid between the 2 tools. Faced a similar issue earlier this week and thought the xcore-200 mc audio kit was dead. The XTAG3 kit latches up or some related event. Really do not think the explorer kit is dead.

Try this,

1) remove xtag3 from explorer kit
2) remove power to the xtag3 by removing the USB cable
3) download and install a fresh copy of the tool chain from xmos website
4) then add the USB cable to xtag3 and confirm it is detected by device manager
5) keep xtag3 powered and now power up the explorer kit
6) mate the 2 boards together
7) check to see if the explorer kit is detected through the xtag3 tool

Post your results.

You may have some virus tool blocking the low level device driver as well that is causing this quirk. Do try all of your USB ports, different PC and shorter but fully loaded USB cables. One of these days will try to inspect the USB traffic between the tool with a bus analyzer. Worst case is to source a new xtag3 tool (<$20usd) but do not think it is this severe of a case (yet).
psupine
Member++
Posts: 22
Joined: Wed May 16, 2018 7:28 am
Location: Australia

Postby psupine » Thu Nov 07, 2019 11:06 am

Step 3) was the one that took all the time until now.

I also checked that the environment variable for the startup delay was still there.

The XTAG3 device is showing on the Windows list, but I've now been sitting looking at "Performing startup actions: (0%)" at the bottom right of the xTIME IDE for five minutes. The LEDs on the XTAG3 are unchanged. Just posting this before I reboot the laptop and try again.

EDIT:

Rebooting didn't alter things, but the reinstall seemed to create a whole load of project file synchronisation issues. I deleted all the XMOS examples and libraries I've accumulated into the IDE over the last year and just kept the code I'm working on at the moment. That got me past the "Performing startup actions:" stage.

At the end of all this, I still have that XTAG3 appears in the device list (for all the USB ports I tested) but it still can't see the eXplorer board. Looks like the next step is to order another XTAG3 before they also disappear from stock. Maybe it's a BGA problem? Here's hoping.
User avatar
mon2
XCore Legend
Posts: 1765
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Thu Nov 07, 2019 12:54 pm

Download the xtag3 hardware manual, review the schematic at the end of the document. Confirm the various voltage rails on the xtag3. Check also on the status of the polymer fuse at the xtag connector.
psupine
Member++
Posts: 22
Joined: Wed May 16, 2018 7:28 am
Location: Australia

Postby psupine » Thu Nov 07, 2019 10:38 pm

Hi mon2. (It's morning in Australia again).

That's exactly what I did last night. All the voltages are correct, and the fuse is intact. I also washed the boards down with IPA in case I had some shorted pins somewhere.

It looks like all the other test points are dynamic, and my scope is old and analogue, so unless something is on a fast retry I'm unlikely to see anything. But that's what I'll check now.

UPDATE:
The SPI_CLK on the serial ROM (TP28) is ticking away like a champion and the SPI_CS (TP29) is asserted. Both these go away when I press RESET, so you're right that the processor is not dead. I have never flashed the ROM so there is nothing on the QSPI data lines.
User avatar
mon2
XCore Legend
Posts: 1765
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Fri Nov 08, 2019 2:32 am

Excellent.

So:

a) XTAG3 with the USB cable appears fine under Device Manager? Device driver is mapped and not a yellow question mark? This was the reason to request for the full tool installation - also hoping some Antivir or similar tool does not steal away required drivers.

b) Assuming yes to above, then power up explorer kit and mate this tool - then enter Command Prompt shell under xtime tool shortcuts and run:

Code: Select all

xrun -l

What is the output?
psupine
Member++
Posts: 22
Joined: Wed May 16, 2018 7:28 am
Location: Australia

Postby psupine » Fri Nov 08, 2019 2:56 am

Capture.PNG
The XTAG-3 Driver thinks that it is working properly. The Driver version is 0.1.12.1
You do not have the required permissions to view the files attached to this post.
User avatar
mon2
XCore Legend
Posts: 1765
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Fri Nov 08, 2019 3:01 am

Ok, what does the Device Manager show with the XTAG3 tool connected? Does it appear as a XMOS device? I recall that libUSB is used for this hardware.

Also, be sure to review this thread:

viewtopic.php?t=5846

(you / we are not alone- Dingo ate my XTAG3?)
User avatar
mon2
XCore Legend
Posts: 1765
Joined: Thu Jun 10, 2010 11:43 am

Postby mon2 » Fri Nov 08, 2019 3:07 am

While I cannot vouch for this without testing but it appears there is a more current device driver release of libUSB from here:

https://sourceforge.net/projects/libusb ... s/1.2.6.0/

If I was there, besides trying to pet a (joey) kangaroo, would install the latest and greatest to see if this resolves your quirk. You can always right mouse click and remove & delete the driver package if not suitable. Worth a try.
psupine
Member++
Posts: 22
Joined: Wed May 16, 2018 7:28 am
Location: Australia

Postby psupine » Fri Nov 08, 2019 3:14 am

Yeah, it thinks it's an XMOS XTAG-3 with a picture like a small refrigerator. The driver itself is libusb0.sys and .dll

I found replacement stock at Farnell (subsequently element14) for both the XTAG and the eXplorer. They're both on order. If I can isolate the problem to the XTAG then I'll get another one and have two working sets.

I'll give the replacement driver a try, but I haven't changed anything in the software setup for a long time. It used to work ... and now it doesn't. I'm inclined to think that I killed something physical.

Who is online

Users browsing this forum: No registered users and 1 guest