• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 83 subscribers
  • Views 375 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.

Original question:

TIDM-02002: DC-DC / inverter stage solutions for lv battery PCS system

TIDM-02002: BCM mode low battery resistance considerations

Edvard Oblocinski
Prodigy 75 points
Part Number: TIDM-02002
Other Parts Discussed in Thread: TIDA-010054

Greetings,

I have been trying to calculate the values of a resonant tank for CLLLC DAB based on TIDM-02002 and simulate it with limited success. 

I have used the provided excel spreadsheet (slightly modified) to calculate the values of resonant tank and got following results:

While I was trying to settle down on the values of the resonant tank, I used provided CLLLC_tankSelection matlab script to calculate the gain and this is where I run into an issue.
The system I am designing would be charging and discharging LiFePO4 16S1P battery modules and certainly more of such modules in parallel i.e. 16S2P, 16S3P etc.:
The main concern I have at the moment is that such battery module would have a dc resistance in the range of miliOhms as far as I understand. 

Question 1. I calculated that RL @ max power (at the voltage where I need max power during charging) is 0.936Ohm, and if I work backwards and use battery dc resistance I get that max power for 0.2Ohm battery resistance would be at 24V. Would it be possible to move this max power point somewhere closer to battery charging voltage? would that yield max charging power? or am I misunderstanding something.

Question 2. For the battery resistance that I assume to be in miliOhms range the Gain graph that I get when running provided matlab script is very shallow (Which would mean that DC-link would have to have very little ripple for the DAB to function in its intended switching mode, if I understand the theory correctly. See figure attached below) What would be the best way to correct it? Again this could hinge on my misunderstanding of RL as I assume it to be a battery pack's dc resistance.




  • 0
    TI__Expert 4580 points

    Edvard,

    We actually have a presentation going through how we design the CLLLC stage. You can follow the slides below (from page. 29) for CLLLC design process. Please go through the slides and let me know if you still have question.

    https://www.ti.com/seclit/ml/slup399/slup399.pdf

    For what RL means to a resonant converter like CLLLC we are using here, please refer to slide 6 of the below presentation:

    https://www.ti.com/seclit/ml/slup377/slup377.pdf

    Re in the slide should be the RL in your calculation. 

    Regards,

    Sheng-Yang Yu

  • 0 in reply to Sheng-yang Yu1
    Prodigy 75 points

    Thank you for your swift response, I have reviewed the slides that you have provided and I have a few things that need clarification.

    So if I understand it correctly if we have a battery that we want to charge at 53V with 3kW then the current going into the battery would be ~56.6A. As far as I understand to the supply the battery would look like it is a 0.936Ohm load and then we can calculate Re which would be ~37.19 Ohm. Please correct me if I am wrong.

    Also cosindering TIDM-02002 seems like Re = RL' which is equal to RL * (NCLLLC * NCLLLC)  


    I have updated the calculations considering this and I still get quite none linear gain curve (especially at desired power level):.

    So if I understand it correctly, because this gain curve is so none linear, we'd have to have very stable DC-link to deliver max power to battery at a desired voltage. So if we consider 53V output voltage, and 7:1 turns ratio DC link would have to be at 371V, which considered 5% ripple would equate to 352-390V. We basically need to be able to sustain 0.136 - 0.151 gain which seems impossible considering the chosen parameters. Is there a way to meet these goals? 

  • 0 in reply to Edvard Oblocinski
    TI__Expert 4580 points

    The Re calculation looks right to me. And the gain curve plot is also making sense. Please note to ensure the resonant converter to have soft-switching, we need to have operating point at the gain curve where it has negative slope. When the load is heavy, the gain curve peak will converge to series resonant frequency. That is, we have to make sure heavy load operation frequency needs to be higher than the series resonant frequency to ensure soft-switching operation. Therefore, at 53V 3kW, you need to have higher input voltage. For example, you could set CLLLC input voltage to be 380V to 420V. That way, the required gains are 0.126 to 0.139, which allows you to operate at frequency a little bit higher than series resonant frequency.

  • 0 in reply to Sheng-yang Yu1
    Prodigy 75 points

    Hey, 
    Thank you for your response. Are there any downsides to having such a narrow operating frequency range?


    TIDA-02002 refers to non monotonic gain curve below resonant frequency as risk, due to possible loss of ZVS and loss of control. What are the implications of such a none linear gain curve, can the loss of control be avoided in our case? Maybe albeit less efficient, single-phase Dual Active Bridge (similar to TIDA-010054) is a better fit topology for our application? 



  • 0 in reply to Edvard Oblocinski
    TI__Expert 4580 points

    At mid to heavy load, gain curve is steep, and it is easy for CLLLC to cover wide output range.

    While at light load, gain curve above series resonant frequency is flat, you might need additional control scheme to achieve lower gain.

    DAB is another option but I think resonant converter is still feasible for this application.

    Just need to be careful of not to have the operation point away from resonant frequency too much. The efficiency will drop a lot because hard-switching on the rectifier.