• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 39 subscribers
  • Views 720 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.

TIDA-010933: TIDA-010933 and TIDA-010938 BESS Charging

BitBricker
Prodigy 10 points
Part Number: TIDA-010933
Other Parts Discussed in Thread: TIDA-010938, LMG3522R030

Tool/software:

TI Team, 

You offer 2 very interesting ref designs, and for that I thank you! However, do have a few follow up questions:


1)TIDA-010933: I see that this is a great little design that offers both PV and 48v batt inputs. However, how is the charging of 48v done on this unit. I havent looked into the code too deeply yet, but do we have ability to select PV charging or AC charging on the fly? 

2)TIDA-010933: Lets say we are in simple setup of 48v batt input and 240v output, "exporting" to grid. Now, we want to reverse the flow on the fly and charge the battery. What are the flow of events here to reverse the flow and what type of timing we are looking at.

3)TIDA-010933: This unit is well powered however, what would it take to increase the capacity to lets say 3.2kw with battery as being supply. The main GaN on AC side are high current units, and only limiter I see is the UCC21540DW, which is limited to 6A.

4)TIDA-010933: T1 Bourns 181923 transformer. This is the part listed omn the BOM however doesnt seem to exist anywhere. 

5) TIDA-010938: Very interesting design with great power handling. One downside of it is that the BESS side is HVDC for the most part (80-320v it seems). Of ocurse we can use boost to bring up the 48v pack to the 300v, however what can be done for bidirectional operation on that case (ability to use this device to charge and discharge the batt.)

Hope its not too many questions. Thanks!

  • 0
    TI__Intellectual 2920 points

    Hello!

    Thank you for your warm feedback and your questions!

    1) To start charging of the 48V battery you need to apply negative current reference to the channel. In this case the current regulator will start to push the current from low voltage bus to the channel. If you have not enough power coming from the PV inputs the low voltage bus will start to discharge and this will cause CLLLC power flow from 400V DC link to LV side. 

    2) To do this you need to do the following: 1. To set negative current reference in the battery channels. 2. You need reconfigure the syncronous rectification control from secondary side control to side. Theing se design is uscondary side current to stop rectification when current reaches zero current. In case of chagring from AC you need to use primary side current to stop primary side rectification. The voltage loop in AC control circuit will change the current flow automatically to maintain 400V on DCLink. Please note: This mode should be possible, but I never tested this mode of operation in our lab.

    3) The bottle neck of the design is the LV side of CLLLC converter. Now it capable to transfer 1.6kW, and the RMS current in the pimary side is already high ~24Arms. with 3.2kW the current will be ~48A.

    So, if you want to achieve 3.2kW you need to redesign LV side to use 4x LMG3100 and use LMG3522R030 on the HV side. The AC side need to use 30mOhm GaNs too. 

    4) The transformer is custom part. Please contact Bourns for the additional info and samples.

    5)  I'm sure that the battery port is bidirectional by default. No need for any changes to charge and discharge the battery.

    I hope this helps!

    Best regards, Vsevolod.

  • 0 in reply to Vsevolod Elantsev4
    Prodigy 10 points

    Thank you for your reply. Yes I did notice the low side LMG2100 being on a weaker side. Both the LV CLLLC circuit as well as lv dc supply seem to use that fet. However, in our design based on this ref, we would not have any 4 channel solar input, just a single 48v battery input. With that being said, why couldnt we jump up to a fet like LMG3522R030 on both the lv dc input and the lv clllc? Or even better yet, go with 2 parallel fets for the clllc on the LV side, so 2 parallel fets instead of U7 and 2 parallel fets instead of u9. Wouldnt that be feasable. However, wouldnt that require changing the T1 transformer as well, or that one is safe to use?

    What about current limitation? With this setup, using C2000, can the current be limited in both directions on the fly? Or that was never implimented?