• State Resolved
  • Locked Locked
  • Replies 13 replies
  • Subscribers 32 subscribers
  • Views 1144 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.

TM4C1294NCPDT: CDC mode woks on certain Tera Term, but not PuTTY

David Chance
Genius 4775 points
Part Number: TM4C1294NCPDT
Other Parts Discussed in Thread: EK-TM4C1294XL

Run into something very strange.

I use USB/CDC mode so that I can use serial terminal to talk to the micro.

If I set Optimazation level to 3, and the Speed vs size trade-off to 2, the program works fine in both PuTTY and Tera Term 

If I set Optimazation level to 3, and the Speed vs size trade-off to 3, the program only works fine in Tera Term, but  PuTTY can't connect to the device at all

Any pointer to solve this issue? Thanks!

  • 0
    Guru 117855 points

    Greetings,

    Since one of your terminal Apps performed under both, 'Speed vs. Size' tests - I'd bet that you've hit a, 'PuTTY or PC issue.'    Is another PC available - which may shed some 'PC specific' light?   

    In such cases it proves always worthwhile to,

    • Monitor & ideally capture/store your Signals - under both Speed/Size settings - looking for differences.
    • Minimize (most or all) other PC activity - while conducting fresh tests.     As always - your objective is to give your transfer 'Every chance to succeed.'
    • In some cases it proves best to 'Reset the PC' - prior to launching the terminal App.

  • 0 in reply to cb1_mobile
    Genius 4775 points

    Tried and confirmed the same result on multiple PCs. 

  • 0 in reply to David Chance
    Guru 117855 points

    And - you've:

    • Monitored & ideally captured/stored your MCU's Signals - for eased comparison?
    • Minimized (most or all) other PC activity - while conducting such tests?

    It proves likely that (something) changes during the shift from, "Speed/Size 2 to Speed/Size 3."     You are the one 'closest to the action' - thus best able to 'identify, document & present here' - the significance of that change.'

    It should be noted that yours proves a, 'Far more PuTTY/PC Issue' than an MCU one - you may consider, 'Requesting aid there as well...'     (if 'skilled' - they may well ask for similar detail...)

  • 0 in reply to cb1_mobile
    Genius 4775 points

    I captured two wireshark sessions, didn't see anything obviously wrong. but I don't know that much about USB/CDC protocol

    I don't think PC plays any role in this case. 

    ps. I merely want to see if anyone have a quick pointer, and I am sure PuTTY and TI will point finger at each other, so let's close the book unless TI FAE has more to share.

  • 0 in reply to David Chance
    TI__Guru 72500 points

    Hi David,

    Doing the wireshark sessions is the correct approach. Since they are not showing any error and since this issue occurs as the optimization is set to faster execution, I suspect that it is a data overflow problem that PuTTY is not handling well. Did you include the optional USB buffers in your implementation?

    I am going to ask Ralph to look at this thread as he is way better at USB than I am.

  • 0 in reply to Bob Crosby
    Guru 117855 points

    Bob Crosby said:
    I suspect that it is a data overflow problem that PuTTY is not handling well.

    Cannot this 'suspicion' be 'quickly/easily' tested - by substituting a, 'FAR SMALLER' transfer - in place of the 'LARGER ONE' - currently failing?

  • 0 in reply to Bob Crosby
    Genius 4775 points

    What do you mean by "optional USB buffers"? It doesn't ring a bell to me at this point

  • 0 in reply to David Chance
    TI__Guru 72500 points

    I assumed you were using the CDC Device Class Driver from the TivaWare usblib. The document is in the TivaWare directory:

    C:\ti\TivaWare_C_Series-2.1.4.178\docs\SW-TM4C-USBL-UG-2.1.4.178.pdf

    Here is the block diagram:

    An example of using the buffers in in the file
    C:\ti\TivaWare_C_Series-2.1.4.178\examples\boards\ek-tm4c1294xl\usb_dev_cserial\usb_structs.c

  • 0 in reply to David Chance
    TI__Guru*** 135355 points

    Hello David,

    Just so I am testing apples to apples, are you using a single CDC port or the TivaWare composite example with two CDC ports?

  • 0 in reply to David Chance
    TI__Guru*** 135355 points

    Hello David,

    I was not able to re-create your observation with my project. I used the Optimization settings you outlined in your post:

    With those settings, TeraTerm, PuTTy, and Termite as well all worked identically with my example.

    I am attaching the example project I used for your reference (this will be released in the next TivaWare): 1018.usb_dev_cdcserial.zip

    You'll need to reconfigure the optimizations for that project as I just used the existing zip file I already had made in the past to keep the distribution consistent on E2E.

  • 0 in reply to Ralph Jacobi
    Genius 4775 points

    Thanks, it could be the optimization of my added codes, I will see if I can extract it.

    Anyway, this is already beyond the scope of example program, so I will stop here

    Thank you all for your inputs!

  • 0 in reply to Ralph Jacobi
    Guru 117855 points

    As vendor's Bob noted earlier - the 'Transfer Size' may play a role here.

    Should the transfer size which has 'worked' prove much smaller than that used by the poster (which failed) - the 'solution' may not be as 'broad & effective' as desired.

    Testing w/poster's (previous failing) data would prove more 'convincing' - is that not true?