This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AWR1642BOOST: AWR1642 not working with Uniflash anymore

Part Number: AWR1642BOOST
Other Parts Discussed in Thread: AWR1642, UNIFLASH

I have AWR1642 Boost and from the forum it looks like I have es1.0.  I used this sensor about 6 months ago and didn't have any issues as long as I had it in administrator mode when running the browser.  That was using the mmWave demo and I originally flashed it about 1 year ago.  I'm not sure what changed.  I cannot get the cloud Uniflash to work with my device, standalone doesn't work either.  It has a "cannot connect to com6" issue and yes it's the correct com.  It's the XDS110 Uart com port which is com6.  What could be the issue?  Is it an overprotective security issue for com ports?  Is it an issue with the hardware?  

The crazy thing is it autodetects the device.  Why would it be able to do that and not open the com port?

  • Hi Justin,

    ES1.0 AWR1642 devices are not compatible with SDK 2.0.  You will have to revert to SDK 1.2.0.5.

    Regards,

    Justin

  • So the cloud version is no longer supported? Imagine I'm just trying to out of the box do the mmwave demo. Run it from the cloud and nothing happens. The SDK doesn't have anything to do with the ability to uniflash format the device from what I've seen. That would only apply if I already made the vital sign ccs build correct?
  • Hi Justin,

    Any software compiled with SDK 2.0 or later will not run on the ES1.0 device. When you run the OOB demo, ensure that you have flashed the device with a demo that was compiled with SDK 1.2.0.5. Then, when running the GUI Visualizer, ensure you select SDK 1.2.0.5 from the drop down menu.

    Since you last used the device 6 months ago, have you changed the software on the device with Uniflash? If you have not changed the software on the device, please update the software to the SDK 1.2.0.5 OOB Demo. Finally, ensure that the jumpers are configured properly. SOP0 should have a jumper. SOP1 and SOP2 should not have a jumper when trying to run the demo. When flashing, ensure that SOP2 has a jumper.

    Regards,
    Justin
  • Hi Justin,

    Did this solve your question?

    Regards,
    Justin
  • Not at all.  Here is the issue.  I'm not getting it through to you that the device won't format.  I'm just trying to format and it won't connect to format.  Explain how to format the device before talking about compiling the code.  If you can't connect it in a cloud environment to format it.  I have mmwave sdk 1.02.00.05.  so please don't respond with the identical message for the 3rd time.  It no longer formats.  I tried from a linux machine and it formats just fine.  I can't connect it to the device.  I get the error I mentioned before.

  • Hi Justin,

    Uniflash will fail to connect to the device for two likely reasons:

    1. The jumpers are not set properly on the EVM - ensure that SOP0 and SOP2 have a jumper when flashing, and SOP1 does not have a jumper. After the flash is complete, remove the jumper from SOP2
    2. Another program is connected to the com port. Common examples for me are Matlab having used the COM Port, or uniflash. If uniflash crashes while trying to flash a device, it can steal the COM Port.  In this case, it is usually easiest to restart the PC to connect to the COM Port.

    Regards,

    Justin

  • I don't think that is the case for a couple reasons.  

    1. Jumper is setup correctly.  

    2. I do it on the restart and have tried that repeatedly.  

    Thank you,

    ~Justin Kahl

  • Can you give me some pointers on how to confirm the port is occupied by some other program?
  • Hi Justin,

    We may need to get help from the Uniflash forum but wanted to confirm some things first.

    1. Are you clicking the "Load Images" button on the Program tab? (I see the "Format SFLASH" button shown in the pic you sent)

    2. Are you power cycling the EVM after setting the jumpers in flash mode before you click the "Load Images" button?

    Regards,
    John
  • 1. Not that time but I have with the same result. In that one picture, I just clicked format first to format it manually upfront to test.

    2. Yes.

    Thank you,
    ~Justin Kahl
  • And just to add. I restart the computer. Run everything in administrator mode. I tried both the desktop version of uniflash and the cloud. No effect.
  • Hi Justin,

    It looks like we may need some help from the Uniflash forum. I will change the title of the thread slightly to match the problem and then move it to the Uniflash forum to see if they have more suggestions that may help.

    Regards,
    John
  • Hello John,

    Did you do this already?

    Thank you,
    ~Justin Kahl
  • Justin,

    Yes, this is being looked at by the UniFlash team right now. I am contacting the mmWave team to see if they have any information on why the device is refusing connection on the COM port.

    Thanks,
    Ricky
  • Hello Ricky,

    I don't know if this piece of information helps to diagnose but I am adding a picture for using the 3rd method for uniflash.  1st method being autodetect.  2nd method being manual select device with new configuration.  3rd method being use ccmxl file to load existing configuration.  I get a slightly different error in the 3rd method.  It says "can't initialize target CPU: The driver property "endianness" was not defined."  So maybe this is a uniflash issue with bit size/orders defined since some previous update?

    Here is the method screen I mentioned earlier in case it wasn't clear.

  • Oh and one more thing! I also just tried to uninstall the drivers for xds110 manually and reinstall the mmwave sdk 1.02.00.05 afterwards. Not sure how to uninstall a sdk package and reinstall it so that is the best way I know how. The "xds110 Class Application/User UART" driver version is 1.7 driver date 4/5/2016 which I'm not sure is correct either. Both times the same driver version comes up for the com. So if there is a device manager check to make sure that the 2.0 mmwave sdk version software i tried to uninstall and delete folders for when i reinstalled the 1.02.00.05 (because AWR1642 esb1.0), perhaps I could get help checking that 2.0 doesn't still have some floating files or something that is making it difficult.

    Thank you,
    ~Justin Kahl
  • Hello Justin,

    I appreciate all the trials you have done with Uniflash and AWR1642.

    As you have mentioned that few months back it was working but not now, could you try these steps to confirm if the device is still alive and responding in SOP-4/5 mode?

    As per your last snapshot, UniFlash is able to connect over COM port but the device doesn't respond for Break Signal generated by Uniflash.

    • Set Device in SOP-5 mode (follow jumper settings as mentioned in the user guide) and do the physical power cycle.
    • Open TeraTerm and connect 'Application/UART' COM port @115200 baud rate.
    • Log the screen output to a file. Menu->File->Log. Select 'Binary' option checkbox in the pop-up window and provide file location.
    • Send a Break signal to the device from TeraTerm. Select 'Menu-> Control-> Send Break' option.
    • Device should respond with few bytes.
    • Open logged file in binary mode, it should have this byte sequence. [0004 cc00 cc0d 0a]

    If the device is not responding during the above step then there is some issue with the device/EVM.

    Regards,

    Jitendra

  • Hello Jitendra,

    I would do that and could if I knew what teraterm is.  What is this program?  I can't find it anywhere in the user manual.  Googling seems to show it as a terminal emulator?  Are you asking me to install a shell emulator or do you mean to have me use CMD.  This is windows machine if that wasn't clear.  

    I have been using sop 5 mode because I'm simply trying to use the format before installing the vital signs lab onto the device with CCS after.  I am working with Uniflash because that has worked in the past on this PC. 

    The short story is... it should already be flashed and I can't test the mmWave demo.  I've tried that and it didn't work.  What is the root cause is beyond me to determine.  

    The previous mmWave Demo software that was on the device shouldn't be on there anymore as a successful flash occured when I tried to shift to a linux PC.  I had some linux issues and it seemed it didn't fully install correctly like it should have.  I saw other forums that had the same issue and decided to just go back to the windows machine and that this would be the quickest route to a solution because linux environment setup could take a very long time if I don't understand the issue.  But that said, the format option in the terminal version of linux did work. 

    If this is outside of the user manual, could you include pictures of what you mean.  It seems that you are using TeraTerm but I would like confirmation of what you mean by that.  

    Thank you,

    ~Justin Kahl

  • Hello,

    I looked into the teraterm.  I think I know what you mean.  We may have slightly different versions as the BPS rate was an option I had to set.  The result though doesn't just interpret the binary.  It for some reason opens notepad to view.  So do you have a viewer?  I tried it in notepad++ but I don't see binary in the form you just posted.  What is the encoding?  UTF-8?  If so, here is my result.  

    Seems like that is not what you expected but I'm not sure what binary output I should see.  I'm also not sure if it's formated if this should be the case.

    Thank you,

    ~Justin Kahl

  • I'm not sure looking at this binary data what it means.

    If you haven't captured this data in baud rate of 115200 then captured data will be totally different.

    Notepad++ provides an converter plugin which you can add and use to convert captured binary to Hex. This is snapshot of my captured file which is converted to '2004CC20CC' in Hex.

    Regards,

    Jitendra

  • Not quite the same but kinda close?

    Is the difference maybe using ESB 1.0 vs. ESB 2.0?