Getting Loop Gain of Boost Converter with UC3843 controller using SIMetrix software

Discussion in 'General Electronics Chat' started by Julian Marvick F. Olivero, Jul 20, 2013.

  1. Julian Marvick F. Olivero

    Thread Starter New Member

    Jul 20, 2013
    5
    0
    Hi everyone I am trying to measure the loop gain of my Boost Converter (Vin = 15V, Vout = 30V, up to 10A load) using the free version of the SIMetrix software, SIMetrix/SIMPLIS Intro 7.00. I am measuring the overall loop gain (outer voltage loop, including the current loop) in order to check the phase and gain plots for stability purposes. The controller used is the UC3843 controller also available in SIMetrix.

    The problem is I can't believe that I am doing it right since the resulting plots are exactly the same regardless of whether I shut down the IC, change the compensation values drastically. The magnitude gain is very low ~ -50dB below at 1Hz up to 100kHz which is weird. It is weird since the resulting plots of output voltage curves are totally good and ok from no load up to 10A.

    I used AC analysis. Attached here are images of the schematic, the options I chose in the Simulator->Choose Analysis window, and the plot which remains the same regardless of compensator values. The plots change when I change the load, but the magnitude gain is still very low.

    Some explanations to the schematic: input voltage is V1 (displayed as 15V in schematic), load is I1 (displayed as 8A in the schematic), C4,R15,C5 are the compensators Type II compensator, Q2 and R10 comprise the slope compensator, I used 3 parallel NMOSFETS IRF32105 each having Rg = 10 ohms and Rgs = 30kohms. V3 is the perturbation, and the probe used is the Bode Plot Probe.
     
  2. LvW

    Active Member

    Jun 13, 2013
    674
    100
    I didn`t check your circuit in detail - thus, I cannot detect if right or wrong.
    However, here are some basics for loop gain measurement.
    * The loop must be opened at a suitable point - without disturbing the operating point
    * As this is problematic in most cases, some specific provisions are to be made
    a) ac opening only using a huge inductor and voltage feedung in via a huge capacitor or
    b) test voltage Vac in SERIES with the loop (insertion of Vac at a low impedance point)
    * The whole circuit must be linear (or linearized using special components).

    Is your method in accordance with these requirements?
     
  3. Julian Marvick F. Olivero

    Thread Starter New Member

    Jul 20, 2013
    5
    0
    Hey there thanks for replying! To answer your questions:

    1.) I opened the loop and inserted the disturbance V3 where one end is connected directly to the output of the converter (low ac impedance), and another end is connected to the feedback resistors path (higher ac impedance)
    a.) I did not try doing this. Is this similar to how one would open up the loop in real life using a network analyzer? Like pass the noise/disturbance through a transformer then inject it at the circuit? What I did was simply put in the AC source there. I'm wondering what's the physical explanation of the need for the inductor and capacitor.
    b.) The test voltage is series with the loop. I also found out after searching some online literature that switch-mode supplies can't just simply analyzed with AC analysis since they have different linearizations per switching state. This may have been the problem all along. Transient analysis using step responses and insertion of sinusoidal signals showed that my boost converter is good which makes my results really unreasonable.

    I found this website http://www.simplistechnologies.com/simulation/identify/example/bode which shows that the AC analysis can be done using the SIMPLIS engine, which I did use. The results were more realistic. However, at points greater than half the switching frequency, the resulting plots were very unrealistic (zigzags and some noisy looking stuff). That's also what I experience when I use real physical network analyzers to check power supply feedbacks, at higher frequencies it becomes realistic. Do you know why this is? My switching frequency is around 100kHz by the way.
     
  4. LvW

    Active Member

    Jun 13, 2013
    674
    100
    In case of series test voltage (V3 in your circuit) the loop gain is the RATIO Vin/Vout, did you consider this?
    As I have no deep experience in switch-mode supplies I cannot answer the other questions, sorry.
     
  5. Julian Marvick F. Olivero

    Thread Starter New Member

    Jul 20, 2013
    5
    0
    I am not sure what you mean by Vin/Vo, can you explain it? What I was measuring is this: In feedback response there is open-loop gain A and there is feedback gain F. The loop gain I measured is basically the product AF, since the net gain would be is A/(1+AF) and for stability purposes it's the AF in comparison to -1 that would matter. That AF term is what I was trying to measure.

    The measurement procedure I based this from is how we would do it using a network analyzer. I also based it from the video in the SIMPLIS link I showed earlier. What it does according to my understanding of the explanation in the video is that the disturbance V3 injects a noise, and then goes around the loop, and then when it reaches back to the output the simulator measures Vo/V3.
     
  6. LvW

    Active Member

    Jun 13, 2013
    674
    100
    Ok - let me explain.
    There is one overall loop which determines mainly the behaviour of the circuit.
    Then, find a node A where where a low output impedance is connected to a large input impedance (at least a factor of 100 between both).
    Then, you can open the loop at this node without disturbing the load conditions too much.
    Well, open the loop at this point and insert a series ac source Vac=1v at this place. Now, name the neg. side of the source Aout and the pos. side Ain.

    After performing the ac analysis the loop gain is V(Aout)/V(Ain).
     
    Last edited: Jul 21, 2013
  7. Julian Marvick F. Olivero

    Thread Starter New Member

    Jul 20, 2013
    5
    0
    ahh ok now I get you. That's what I did. Thanks for the reply!
     
  8. crutschow

    Expert

    Mar 14, 2008
    13,003
    3,232
    That may be due to the fact that, at any frequencies above the Nyquist frequecies (1/2 the sample frequency), you get aliasing effects.
     
  9. Julian Marvick F. Olivero

    Thread Starter New Member

    Jul 20, 2013
    5
    0
    Thanks for the replies! I think I have found a book that discusses the matter more theoretically and mathematically. I will read it then ask questions again when I get stuck. But for now, I will just accept the fact that simulations up to 1/2 the switching frequency is accurate. Further above that is not so accurate anymore XD. I was just wondering how this could happen in a physical situation and not just in a simulator? Would aliasing effects also occur in a physical situation with a Network Analyzer and a hardware power supply,, because that's what my experience 2 months ago is? I am not so much knowledged with regards to aliasing, only thing I know of it is that if you try to reconstruct a continuous-time signal from its sampled version, then the frequency content of that continuous-time signal must satisfy Fmax, the maximum frequency must be greater than or equal to twice the sampling frequency of else aliasing occurs. I do not know how that is related to power supplies. If you could explain it to me it would be really helpful.
     
Loading...