• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 82 subscribers
  • Views 361 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TMS320F28035: HRPWM for EPWMxA's falling edge and EPWMxB's rising edge

Jack Tan80
Mastermind 7100 points
Part Number: TMS320F28035
Other Parts Discussed in Thread: TMS320F280025C, TMS320F280049C

Dear C2000 expert,

Currently I plan to use F28035 to control interleaved buck with synchrouse recifier, and switch frequency is 400Khz. 

Please see below picture about my expectation about PWMxA( control upper switch of Buck ) and PWMxB( control lower switch of Buck ), this requires to use high resolution for EPWMxA's falling edge and EPWMxB's rising edge. Can you please help me know if F28035 could realize it? If so, can you share some suggestion of how to configure ePWM? Thanks...

  • 0
    TI__Mastermind 42195 points

    On TMS320F28035 the HR capabilities are only availability on the ePWMxA channel, for this device please use two different PWM modules for you high and low side switches. This adds some complexity in synchronizing, but should work.

    Alternatively devices with a type 4 PWM module have HR on both the ePWMxA and B channels. See TMS320F280049C and TMS320F280025C.


    Regards,
    Cody  

  • 0 in reply to Cody Watkins
    Mastermind 7100 points

    Hi Cody,

    Thanks for your suggestion. I know that F280049 and F280025 have some improvement on EPWM and other peripherals, but it looks that those devices are currently out of stock from ti order page, F28035 can be ordered directly from ti.

    Another question, if PCMC(peak current mode control) method are used, dose the HRPWM still work under PCMC?

    Regards,

    Jack

  • 0 in reply to Jack Tan80
    TI__Mastermind 42195 points

    Jack,

    F280049 and F280025 are showing to have some availability to me... that being said it looks like some packages are out of stock. 

    I check with our back end people and they said on average you should expect a 6 week lead time on any order <~1ku and 20-26 weeks on larger orders.

    For prototyping EVMs of all devices are in stock and ready to prototype. For your type of application I really do think the newer PWM features will be helpful.

    Nevertheless, I don't see why PCMC couldn't be used with HRPWM, but if you're using PCMC aren't you going to set the duty to 100% and then allow the CMPSS to turn off the PWM output when the current hits your configured trip point?

    Regards,
    Cody

  • 0 in reply to Cody Watkins
    Mastermind 7100 points

    Hi Cody, 

    Thanks for your suggestion and the ordering info of the new C2000 products, I will discuss it with my team.

    For this application, I will choose PCMC method to control the interlaved BUCK. In PCMC,the max duty cycle will be limited to 99% via COMPA/B register value if the peak current do not hit the trip point in one switch cycle. 

    As my understanding, the clock of CMPSS is based on SYSCLK. So the CMPSS has resolution limitation to reponse PCMC event, what my understanding is that CMPSS response resolution is 10ns in F280049, isn't it?  If my understanding is correct, how PCMC could achieve 150ps resolution for EPWM?

    I'd appreciate if you could provide some suggestion on configurations to achieve 150ps resolution for EPWM under PCMC.

    Regards,

    Jack

  • +1 in reply to Jack Tan80
    TI__Mastermind 42195 points

    Jack,

    For PCMC you need to trip the PWM with the CMPSS. This delay is published in the datasheet.

    If you want the fastest possible path you will need to forgo filtering in the CMPSS module and use the ASYNCH path, please see the TRM block diagram for more info on this.

    Finally, there is simply no way to get the a 150pS resolution, the device will turn off as soon as it can detect the event, which alone takes longer than 150pS.

    You still of course have the High-Resolution Dead-band between the A and B channels of the PWM, but as for responding to the CMPSS trip event the action is simply done as quickly as possible.

    Regards,
    Cody