• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 3824 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.

UCC28951-Q1: EV charger station Design

Carlos Ramos21
Prodigy 240 points
Part Number: UCC28951-Q1
Other Parts Discussed in Thread: TLV316-Q1, LM4132-Q1, LM4132, UCC28950, LM5051

Hello,

My team is in charge of designing an EV charger station and we are looking at products available to create such system. I have a few questions in regards of design complexity, performance and available data to design this charger. I stumbled upon “DC/DC Converters for Automotive Applications” series and I’ve found that the UCC28951-Q1 would be perfect for our application.

I have a series of questions in regards of the picture below:

 

  1. The only external hardware to perform CC-CV is two TLV316-Q1, one LM4132-Q1, and a current sensor, is this right? No need for a compensator network/control loop? Any caveats?

  2. Why are the diodes opposing the TLV output? Isn’t that supposed to be ‘ORing’ those outputs?

  3. In the case of needing to change the CC-CV reference, do I change the LM4132 output to; for example, accommodate other types of battery voltages and charging currents?

This is for the case of just one charger/power stage. In the case I want to create a charging station with multiple stages giving me a multiplicative factor of n for n staged chargers, how can I achieve this with the UCC28951-Q1? Do I need a special control scheme to achieve this or the UCC28951-Q1 will regulate them in order to achieve CC-CV?

 

In terms of the design:

  • How does the design tool accommodate with this charger design?

  • Can I use the same tool for the charger?

  • How about the SPICE example?

  • What type of transformer model did you use to design the magnetics in this case? What types of assumptions do I have to make?

  • Same with the current sensor transformer

  • How do I turn off the SR scheme to accommodate a full bridge rectifier at the output?

  • What would be better for the charger, two output “center tapped” or a single output transformer?

  • How would the design change if I want to boost the voltage? I know the PSFB is a buck derived converter but, how do I convert it into a boost? Is it just playing with the transformer turns ratio or is it something else?

  • Is there a mathematical model or an open loop model of this scheme?

  • Any other tips and tricks using this device? Books etc.?

 

Thanks for the help and very respectfully,

 

Carlos Ramos, Electrical Engineer

Resonant Aero

  • 0
    Prodigy 240 points

    Forgot to add the aforementioned photo.

  • 0
    TI__Mastermind 31580 points
    Hi Carlos,

    I have asked one of our applications engineers to respond to your post, you should see a reply later today.

    Regards

    Peter
  • 0 in reply to Peter Meaney
    TI__Mastermind 25570 points
    Hi Carlos



    Q:The only external hardware to perform CC-CV is two TLV316-Q1, one LM4132-Q1, and a current sensor, is this right? No need for a compensator network/control loop? Any caveats?
    A: The slide was intended to illustrate the high level concept not the details. You will need compensation networks - have a look at our PSDS paper at www.ti.com/.../slup348.pdf of course that paper uses different op amps and so on, but the concept is the samel

    Q: Why are the diodes opposing the TLV output? Isn’t that supposed to be ‘ORing’ those outputs? -
    A: The diodes do provide an 'oring' function. The lower of the TLV outputs 'wins' and gets to set the demand signal at the EA+ input.

    Q: In the case of needing to change the CC-CV reference, do I change the LM4132 output to; for example, accommodate other types of battery voltages and charging currents?
    A: Yes, it's as simple as that - IN PRINCIPLE ! - you must double check that the scaling factors for Vo and Io are correct.

    You can parallel multiple power stages - the simplest is to take the EA+ signal generated by the error amplifiers and daisy chain it across to multiple power stages. Each power stage receives the same error (demand) signal and they should all output the same curren

    Q: How does the design tool accommodate with this charger design?
    A: The EXCEL design tool was developed for CV operation only. It will work for the CV portion of the design but not for the CI part. It won't design the external voltage error amplifier either. From the power train point of view, the worst case condition is maxVout and Max Iout.

    Q: Can I use the same tool for the charger?
    A: only for the CV portion

    Q How about the SPICE example?
    A: There is a SPICE model of the part but not - so far as I know - of a full battery charger.

    Q: What type of transformer model did you use to design the magnetics in this case? What types of assumptions do I have to make?
    A: The transformer was designed by an external contractor - you can see the typical specifications we gave to them in the datasheet for the UCC28950 - in the reference design portion at the end of the DS.

    Q: Same with the current sensor transformer
    A: We purchased one off the shelf. Again, you can see the specifications we used in the UCC28950 DS.

    Q: How do I turn off the SR scheme to accommodate a full bridge rectifier at the output?
    A: two possible approaches to do this. 1/ simply leave the OUTE and OUTF pins open circuit
    2/ Set the DCM threshold high enough to turn the OUTE and OUTF outputs off
    3/ it is possible to do full wave rectification with SRs and the UCC28950 - but obviously you would need extra drivers and of course the MOSFETs themselves.

    Q What would be better for the charger, two output “center tapped” or a single output transformer?
    A: Difficult to know - it depends on the output voltage level. Centre tapped is easier for low voltages, single winding better for high voltages. There is also the possibility to use a current doubler. Please let me know your intended outptu votlage range and the current levels.

    Q : How would the design change if I want to boost the voltage? I know the PSFB is a buck derived converter but, how do I convert it into a boost? Is it just playing with the transformer turns ratio or is it something else?
    A: it's inherently a buck converter so the way to deal with this is to play with the transformer turns ratio. Don't forget that the transformer turns ratio could get you an output voltage range higher and lowerr than the input - eg, 400V in, 350 to 450Vout
    or 400Vin, 750Vout

    Q Is there a mathematical model or an open loop model of this scheme?
    A: not so far as I know -

    Q: Any other tips and tricks using this device? Books etc.?
    A: There's lots of material available about the PSFB. The design review link above is a good stareting point.
    TI design seminars are available at www.ti.com/.../login.shtml and this is probably one of the best repositories of PSU design information around - PSFB articles are at www.ti.com/.../login.shtml`and you should look at the tab called 'Full Bridge Power Stage'

    Let me know if you need any further information
    Regards
    Colin
  • 0 in reply to Colin Gillmor
    Prodigy 240 points
    Colin,

    Thank you so much for this prompt response. I'll check the material you sent me and soon I'll follow up with you or any application engineer available.

    Respectfully,
    Carlos
  • 0 in reply to Carlos Ramos21
    Prodigy 240 points

    Colin,

    I have a few questions in regards the PSFB converter design workflow:

    Note: Our design is a high voltage output (200-500V probably, could be a little bit less), high power EV charger

    1. Is the excel file still good to calculate and size my Full Bridge Power Stage, even if the file does not work for CC-CV?
    1a. This paper "Design review of a 2-kW parallelable power-supply module" contains all the sizing formulas? If so, what would I need to change to the formulas from center tap to full wave rectifier output (needed for high voltage, right?)? In general, if I have all the formulas for sizing, I should be able to design the power stage with no problems, right?

    2. If no open loop is available for this converter, how do I design a compensation network for this system? Any tips, advice or suggestions?

    Respectfully,

    Carlos

  • 0 in reply to Carlos Ramos21
    TI__Mastermind 25570 points

    Hi Carlos

    The Excel file is good for the CV part of the operation - this includes out to the corner where the current limit comes in.

    The CI characteristic will pull down the votlage over some range as it regulates - you will know the minimum voltage you need to regulate to

    I'd suggest that you run the Excel calculator twice - once for the max Vo condition and once for the min Vo condition. Keeping the fixed parameters (transformer turns ratio for example) the same for both calcuations.

    As regards compensation - I'm afraid I can't really offer more than you will find in the articles at  http://www.ti.com/ww/en/power-training/login.shtml?DCMP=psdslibrary&HQS=tlead-power-psdslibrary-apec2015-pwrhouse-20150312-lp-en   The articles under the 'Feedback Loop Compensation' will be most relevant - especially the one at https://www.ti.com/seclit/ml/slup340/slup340.pdf 

    Regards
    Colin

  • 0 in reply to Colin Gillmor
    Prodigy 240 points

    Hallo Colin,

    how can the diodes in this case ORing?

    Isn't ORing diodes logic suppose to construct like this?

    I still don't get how the lowest error "wins" the gate logic.

    Could you give an example of the diodes in this application? Is this standard zener diode or is it some kind of Low Side OR-ing FET Controller (LM5050/LM5051)?

    Regards

    Natthapol