+ Post New Thread
Results 1 to 4 of 4
  1. #1
    Junior Member level 1
    Points: 1,241, Level: 7

    Join Date
    Jul 2006
    Posts
    19
    Helped
    0 / 0
    Points
    1,241
    Level
    7

    problem in boost Current mode dcdc

    i tape out one PMU chip and when it coming back , i find in one boost dcdc the output ripple is so big , and seems not stable, please see the waveform of test.

    i wonder if it is system loop stability issue ,and what is the root cause, can you give me some suggestion?
    BTY, the simulation results and small signal model caculation shows the system is stable and PM>60degree, so i am so confused.

    •   Alt20th May 2011, 08:58

      advertising

        
       

  2. #2
    Advanced Member level 5
    Points: 19,685, Level: 34
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,269
    Helped
    1159 / 1159
    Points
    19,685
    Level
    34

    Re: problem in boost Current mode dcdc

    Yeah that looks like instability. Is the yellow trace your output, and the blue trace the error signal? Is this behavior line/load dependent? Are you seeing your inductor current track your error signal?

    Also, how are you inferring your phase margin from a transient simulation? Usually you have to do a small signal analysis of your switching converter (which simulators generally can't do) in order to get PM.

    Also I can't find a datasheet for your controller IC.



    •   Alt20th May 2011, 17:12

      advertising

        
       

  3. #3
    Banned
    Points: 8,487, Level: 22
    Achievements:
    7 years registered

    Join Date
    Jun 2008
    Posts
    1,180
    Helped
    37 / 37
    Points
    8,487
    Level
    22

    Re: problem in boost Current mode dcdc

    You cannot realistically do feedback compensation by calculation.

    there are too many unknowns as per ESR tolerance , opto-coupler tolerance, and resistance of inductors etc.

    Even if you did calculate it........you would still have to measure it with a gain-phase analyser to confirm the calculation. (eg AP200 or venables type etc)

    -so there is no point in calculating it,
    -just use your analog experience to put down R's and C's to to compensate.
    -then measure it with a gain-phase analyser.
    -or like most places, just do transient testing.

    i cannot easily make much sense out of your circuit.

    i cannot see compensatory R's and C's
    I cannot really see where your Vout divider feedback goes.

    I am nos sure why you have two separate chips.

    Is that an smps simulator.....because most simulators cannot do smps



    •   Alt20th May 2011, 20:40

      advertising

        
       

  4. #4
    Junior Member level 1
    Points: 1,241, Level: 7

    Join Date
    Jul 2006
    Posts
    19
    Helped
    0 / 0
    Points
    1,241
    Level
    7

    Re: problem in boost Current mode dcdc

    thanks for your replay. first of all, i will caculate the system loop transfer function when L,C current gain, slope compensation, loading is given, i will get poles/zeros, Q, and gain, then i will simulate error amplifier transfer function, and also get poles/zeros and gain, at last i will get the overall transfer function, it shows the PM >60 degree, and crossover frequency is 30~50khz which the switching frequency is 1.5Mhz, it seems reasonable stable.
    The transient simulaiton also give load transient, and see output voltage setting, it is also ok!
    But the test is like that, ans seems the oscillation frequency is above 100khz, is that anything that i ignored or some strange thing happen.
    BTW, i have not add error amplifier clampping circuit,the EA circuit is very simple, one stage opamp.



--[[ ]]--