• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 63 subscribers
  • Views 715 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.

TPS57112q1 (=TPS54218): how to accelerate the controller Vramp amplitude

Bishnu Ban
Intellectual 500 points
Other Parts Discussed in Thread: TPS40195, TPS54478, TPS54218, TINA-TI, TPS54227

Hi. !

input =5V, Vout=1.2V, Ioutmin=300mA, Ioutmax=2A, controller TPS57112q1, allowed output voltage fluctuation 5%, used capacitors: ceramic e.g. output capacitor 44uF/5mohm, selected switching frequency 1666kHz (to save the size of L and Cout)

 compensation Network type II: take power stage as a voltage driven current source and forget the double LC-pole. you have just a pole created by Cout and Rload (Vout/Ioutmax) and a zero created by Cout and ESR of output capacitor. Gmps, gmea mentioned on the data sheet are enough to design a compensation network of type two.

pole is at 1.95kHz and Zero at 866kHz.

I have taken crossover frequency recommended by T I (66kHz) refer the page 20,

-I put just a zero with pole of plant together and pole of network with zero of plant.

But load regulation (change of load current from 300mA to 2A ) is very bad: output voltage goes down to 1V (more than 20% ).

Now I have to accelerate the controller: I could shift cross over frequency to e.g 300kHz.

it looks better but not enough.

I have to use compensation network type III.

Questions 1: why I have to change the model of the plant from single pole and zero (explained above ) to double complex pole generated by LC for type III compensation?

Question 2: DC gain of simple model is Gmps*Rload why I have to use Dc gain of  complex type to Vin/Vramp (not more Gmps*RLoad) ?

Question 3: I could not find the Vramp on the data sheet of TPS57112q1  for example it is mentioned clearly on the data sheet of TPS40195 (Vramp =1V)

I hope many can help me.

regards

Bishnu

  • 0
    TI__Guru**** 191445 points

    The simplified model for current mode control is as you describe.  But i real life there is slope compensation that starts at 10% duty cycle.  So for 5 V input and 1.2 V output you may have a consederable amount of slope comp.  That will affect teh dc gain as well as the output filter pole and zero characteristics. TPS40195 is voltage mode control, so it will have a specification for the PWM ramp.  Current mode control does not have that functional block.

    You can usually get a design to work with type 2 compensation as long as you do not crossover too high.  You could possibly use webench to generate your design.  Personally, I would use teh TPS54218 average spice model to design my compensation network.  You can use the procedure outlined in the TPS54478 datasheet.

    All that said, I f your design is stable, you should not get that poor of load regulation.  When you get your compensation figured out, post back and let us know if your regulation issue goes away.

  • 0 in reply to JohnTucker
    Intellectual 500 points

    Thanks John!

    it would be very helpful if I could select the type of compensation network type e.g. III.

    I have tried to surpass the design by changing cross over frequency to 300kHz (recommendation by TI-Data sheet is about 65kHz).

    I have changed the Rcom to 36k from about 82k and reduced Com accordingly.

    I looks better, the output voltage drops by 80mV (load current from 300mA to 2A) which is still more than requirement of load. allowed drop is maximum 60mV.

  • 0 in reply to Bishnu Ban
    Intellectual 500 points

    Hi john!

    can you forward (better the file here) this SLVM279 TINA-spice model/circuit! I could not download it

    by the way the TPS54478 is not available on Designsoft/tina!

    regards

    Bishnu

  • 0 in reply to Bishnu Ban
    TI__Guru**** 191445 points

    I would not try to cross over that high.  Usually around 100 kHz is maximum, but I like to stay around 30 - 70 kHz.

  • 0 in reply to Bishnu Ban
    TI__Guru**** 191445 points

    Actually, you do not want to use the model for TPS54478, you want to use the TPS54218 model and the instructions from the TPS54478 datasheet to design the compensation.  I posted the TPS54218 model here:

    1565.slvm507.zip

  • 0 in reply to JohnTucker
    Intellectual 500 points

    Hi john!

    I have calculated the compensation network (type 2A) again and simulated it with Tina-TI (designsoft). There were some mistakes on calculation and some facts I have written on past were not correct. I came to know that type 2A compensation with 300kHz cross over frequency is enough.

    I have collected the information supplied by TI (data sheet and seminar). As per Information onTI-seminar( TI-days) it is ok to use 300kHz cross over frequency (refer the attached file and here is also the schematic file from TINA-TI).

    it is clear to me why I have to select higher fco for better load regulation. but it is not clear me why there is so huge difference between two recommendations from same supplier (I mean TI)

    3755.TPS57112q1-1v2-new.TSC6675.Calculation of fco per TI.pdf

  • 0 in reply to Bishnu Ban
    TI__Guru**** 191445 points

    I don't think you can get any consensus on compensation design and crossover frequency.  If you put 10 engineers in a room you will get 11 different answers.  While both of the sources you cite are valid up to a point, I do not use either of them when I design compensation networks.

    For example 1, the intent is to put the crossover frequency at the lower of the geometric mean of Fpmod and Fsw/2 or the geometric mean of Fpmod and the output cap ESR zero.  That is all good but it assumes ideal current mode control with no slope compensation.  When you add slope compensation, the equations become very complicated.  But for this simple approach, the chosen crossover frequency is relatively low, so it results in a conservative design approach that can work for most circuits.

    Example 2 only concerns with keeping the crossover well away from Fsw/2.  Crossing over at or above Fsw/2 will violate the Nyquist criteria and is strictly prohibited.  Choosing Fco between Fsw/4 and Fsw/10 is fine in theory, but the assumption here is an ideal error amplifier with infinite (or very large) GBW.  For the TPS57112q1, I think you will find that there is not enough gain in the error amplifier to realize a 300 kHz crossover.

    My preferred approach is outlined in the TPS54478 datasheet.  I model the power stage gain and phase characteristics.  I look for the frequency where the power stage gain falls to -120 degrees.  That is the maximum Fco where 60 degrees of PM is possible with type 2 compensation.  Depending on the slope of the phase curve, I may go a little lower or higher (higher if I have the option to add Cff to boost phase).  Once I select the frequency I look at the gain of the power stage at that frequency.  By definition, the compensated error amplifier gain must be equal and opposite so the net gain is 0 dB.  I place the compensation zero one decade below Fco and teh pole one decade above Fco.  Usually Fco falls in the 30 to 80 kHz range.

    Let me know if this makes sense to you.

  • 0 in reply to JohnTucker
    Intellectual 500 points

    Thx john!

    I understood your approach very well.

    -GBW is not mentioned on Data sheet of TPS57112

    Unfortunately  the design with 70kHz fco has bad load regulation (refer the attached files)

    6320.70k.pdf

    0211.300k.pdf

    As per specification of the load, change of Vout should not be more than 5% if load current changes from light load(300mA) to full load(2A).

    Perhaps I have to select other buck controller for redesign.

  • 0 in reply to Bishnu Ban
    TI__Guru**** 191445 points

    Ok, the two files you show are not strictly "load regulation" but rather "load step transient response".  They are two separate specifications.  Load regulation concerns change in DC output voltage relative to dc output current.  You can see a graph of that performance in the datasheet and the tolerance is very tight.  It is an intrinsic property of the controller.  The transient response is the change in output voltage relative to a fast acting rise or fall in load currents.  While a fast acting loop will improve performance, It is generally a property of the output filter components. It is possible to improve performance by increasing the output capacitance and decreasing the output inductor value.  Also you need to be sure about the load current slew rates.  Fast slew rates will degrade performance.  It is important to design for the actual slew rate conditions of your application.  If you do in fact have fast slew rate requirements, then additional output capacitance may be required.  We do have ICs that feature improved transient response.  You may want to consider a device with DCAP2 control such as TPS54227 or a device with "Turbo-Trans" (I do not know a part number from memory).