• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 82 subscribers
  • Views 1425 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

USB debugger disconnecting running motorware labs...

Richard Chapman
Intellectual 640 points

Other Parts Discussed in Thread: MOTORWARE

I have started having erratic debugger problems when running motorware labs.

Typically - running lab2c or lab2d - when I set the "run Identify" flag - the identification starts and continues to run correctly - but windows will "ping" that the USB has disconnected - and the watch window shows errors.

Having just reviewed this forum - it looks like there may have been issues with recent compiler releases.

I am running CCS6 and the "C2800 compiler tools" report as 6.2.8.

Can anyone shed any light on this problem - and if it is compiler related - how do I back out of recent compiler updates? Can I rather change optimization options or otherwise resolve the problem?

Thanks.

Richard.

compiler is 

  • 0
    TI__Guru** 119680 points

    Your compiler is fine and up to date. 

    What you describe is usually from system electrical noise, typically from a high voltage power supply. Are you using the HV KIT or your own inverter?

    Is your PC isolated from the other power supplies (if you have a laptop try running from battery only). 

  • 0
    Intellectual 640 points

    Thanks Chris

    I am using my own hardware based on the HVkit - and currently have a 40 volt supply. I will be running at higher voltages in the future.

    I am using the blackhawk? isolator with the XDS100V2 debugger. I think I have seen the problem when the laptop is on battery - but I am not 100% sure. I will check.

    I'll investigate noise... but I thought the Isolator should make things pretty noise free. it is difficult to run the labs when the debugger isn't reliable...

    Richard.

  • 0 in reply to Richard Chapman
    TI__Guru** 119680 points

    the isolation for the JTAG keeps high currents / voltages isolated, it doesn't really stop any spurious noise that is being sent out by other equipment. as mentioned, a common culprit in a high power lab are "cheap" power supplies / power conversion equipment.  I've seen this many times.

    that's where I would start.

  • 0 in reply to Richard Chapman
    Intellectual 640 points

    If anyone else sees this problem - I can recommend you try a different USB cable. I was using the TI supplied USB cable that came with the XDS100 V2  emulator - and had continual (noise?) problems with USB disconnection. I tried another identical cable with the same result. I then tried another USB cable that included a common mode ferrite on the cable - and all the problems went away.

    Richard.

  • 0 in reply to Richard Chapman
    TI__Guru** 119680 points

    Richard, 

    Oh man, I should have mentioned that first. I also use a ferrite USB cable. Let me see if we can start shipping our higher current and high voltage kits with this style of cable. 

  • 0 in reply to ChrisClearman
    Intellectual 640 points

    Good idea Chris...

    I think the "bad" usb cables I had came with the emulator rather than the HVkit. I can't recall whether a cable was shipped with the HVkit - and if so what it looked like. If a cable came with the HVkit - I have lost track of which one ti was - and probably haven't tried it in a noisy environment.

    Maybe a note in the labs document suggesting an "upmarket" usb cable might help put people on the right track...

    Richard.

  • 0 in reply to Richard Chapman
    Intellectual 640 points

    Hi Chris...

    Not sure if you will see this posting but...

    Do you know where I can get one of those superior USB cables with the ferrite? I think maybe they ship with the HVkit - but I need another one.

    I've searched the TI store without luck.

    Thanks Chris

    Richard.