• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 31 subscribers
  • Views 3162 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.

TM4C1290NCPDT: What is USB 2.0 High Speed maximum expected transfer rate using Communication Device Class (CDC)? Comparison with Full Speed and CDC. ULPI Interface.

Dan Patten83
Prodigy 60 points
Part Number: TM4C1290NCPDT
Other Parts Discussed in Thread: TUSB1310A

Hello. 

We have developed a few USB connected applications using the TM4C1290. Our first tests have used the built-in USB 2.0, Full Speed and CDC. We are transferring an 8KByte block and toggling a GPIO pin at start and finish of the transfer to time how long the transfer took.

Full Speed Configuration: Built-in USB Phy; Full Speed (12Mbit/Second) connection; TM4C1290 running at 120MHz; CDC Communication Device Class

We are able to achieve 8KBytes transfer in 8.18mS. 1/(8.18mS/8000) = 978KBytes per second or 7.82Mbits per second.  Pretty pleased with this result in that maximum would be 12Mbits and there is a lot of USB overhead.

When we add an external USB Phy connected to the ULPI interface.  We are using a Microchip USB ULPI Phy (USB3300).  Using an oscilloscope we can see bit rates of 480Mbits, however, our 8KByte transfer rate stays almost identical.  

3 Questions:

(1) Curious if anyone else can confirm these findings? 

(2) Is it possible that the CDC class Windows driver has a speed limitation?  

(3) I suppose that we might need to figure out how to use the Bulk Transfer mode?

Thanks

Dan

  • 0
    Guru 117855 points

    May I note two 'borderline comments' - intended to 'support' your, 'Need for Speed!'    (Perhaps even beyond your (wildest) 'NEED/SPEED' DREAMS!)

    • To my group - 'Transfer implies (both) a transmission AND a successful reception.'      Thus - if  'hand-shaking' and/or other (communication impacting) methods are 'in play' - such 'transfers' may be impacted by the receiving side!     (and the receiving side may 'vary' in its response - depending upon its usage & load demand.)     Has this factor been considered?
    • Your note of another's 'USB ULPI Phy' brought to mind a recent post by (another) here - who sought (both) 'Speed & Robustness INCREASE' - via the technique known as, '8b/10b.'

    This vendor had produced an IC 'TUSB1310A' - which (appears) to enable the use of  'USB3.0!'      As you have 'proved open' to the use of an 'added IC' (surely in the quest for speed)  I believed my arrival  (reasonable) - to insure that you have this (hopefully) new - powerful 'Competitive Awareness!'      As 'Daenerys'  (surely) would advise  'COMMAND'  - (prior to 'Down Boy!')  ...  HAVE NO  SMALL DREAMS!'     

    First - 'others have long 'made the case' for this '8b/10b'  (vastly improving) method:    (developed in/around 1983/84)

    Note that Robustness is improved - as well as Speed - due to the unique 'DC Balance Achieved' - insuring the optimal performance of 'Differential Signalling!' 

    Your mention of 'ULPI'  -  with the 'LPI'  (LOW PIN INTERFACE)  (i.e. FEW pins) ... sucked me in.

    As you originally were 'casting for ideas to improve performance w/in USB2.x'  - would not  INSTEAD -  your Spectacular ELEVATION TO USB3.0  (AND 5 Gbps) - prove FAR SUPERIOR?   (and win for you - that (needed) and now most deserved - promotion?)     And power you, your firm - LONG into the future?

    Here's the link to the other post - poster '12 Squared' was the first here (this forum) to make this 8b/10b 'connection.'

    https://e2e.ti.com/support/microcontrollers/tiva_arm/f/908/t/719154

  • 0 in reply to cb1_mobile
    Prodigy 60 points

    Thank you for the information.  It would be interesting to look at possibility of using USB 3.0 attached to the ULPI.  FYI, the TI Website states the TUSB1310A as NRND.  I am really interested in speed/throughput experience of other users of the Tiva microcontroller and combination with CDC Windows connection.  Thanks. Dan

  • 0 in reply to Dan Patten83
    Guru 117855 points

    Thank you - indeed my (initial post - to poster '12 Squared' (linked for you)  DID well NOTE the 'NRND' status of 'TUSB1310A.'

    However - if  'Speed is your goal' - hundreds of those (TUSB) devices remain in 'disty stock' - and may be available (at (very) nice discount) should you (gently) note the device's 'death sentence.'   (as we have!)

    The 'experience' of (others, here) - has, 'Yet to arrive' - and most certainly - cannot 'come close' to the 'extraordinary Speed and Robustness Increase' - offered up by the new USB3.0 standard - and proposed (by my group) in your behalf...

    Do not (all other) 'user experiences' - represent 'band-aids' at best?    You may note the 'always increasing - drive toward higher performance' - which is the (likely) cause - of  TUSB1310A's demise...    

    Is it not true - that (this vendor) - follows Daenerys' command - to, "Suffer NO small dreams?"     Should not you  - especially due to the dearth of (experience-based) replies - follow her lead - as well?