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.

LM3488 external slope compensation's effect on current limiting and resistor selection equations

Other Parts Discussed in Thread: LM3488

Hello,

I am having a bit of difficulty understanding the relationship between the current sensing resistor (Rsen) and the external slope compensation resistor (Rsl) as well as were the design equations are coming from.

I am using the LM3488 as a SEPIC controller that boosts from 12V to 24V and is rated at 100W. At no-load, operation was fine, but as I increased the load, I started seeing severe sub-harmonic oscillations. By increasing the slope compensation resistor, Rsl, I was able to remedy the oscillations, however the addition of Rsl = 3k OHMs seems to have significantly reduced the current limit and, thereby, the output voltage and power. I believe the solution is change the sensing resistor, but the methodology for doing so confuses me. 

The datasheet gives some guidance on how to select Rsl, but it seems there should be a more analytical way to calculate its value. Additionally, the equations given always give negative values and I don not understand how to derive them. I am also having a great deal of difficulty understanding the waveforms in Figures 21 and 22. If Io is the switch current, these waveforms make no sense, and in Figure 22, neither waveform aligns with the switching times.  How are M1 and M2 calculated and where does the equation I1 = -(M2/M1)*Io come from? Is there additional literature covering the derivations of these equations? 

Also in the datasheet, are the calculations for Rsen, taking into account the effect of Rsl, but the equation is the opposite of what I expect. If the current limit needs to be higher, I would expect the sensing resistor to become smaller. Yet in the equation:

Rsen = (Vsense - D*(Vsl + dVsl)) / ISW_limit  such that dVsl = 40uA * Rsl

Rsen is directly proportional to Rsl. As such, I am calculating that I need a larger sensing resister, which seems counter intuitive. How is the logic that governs the current sensing circuit functioning? 

The check for stability, not considering Rsl, is also confusing for my system, as it results in division by zero:

Rsen < (2 * Vsl * fs * L)/(Vo - 2*Vin) = (2*0.092*500e3*75e-6)/(24 - 2*12) = infinity!

When rearranged to give the guideline for Rsl this becomes an equally frustrating Rsl > -Vsl/40uA... which is everything.

I believe the core of my difficulties is my confusion over how exactly the current mode control and slope compensation are working and what Figures 21 and 22 are trying to tell me. Any deeper understanding you can impart to me that will allow me to intelligently select the correct components for this circuit would very greatly appreciated.

Thank you!

 

  • Hi Mike,
    There is an application note available that explains very well principles of slope compensation in current mode converters. Please see link below.

    www.ti.com/.../slua101.pdf

    Just a couple of comments on the equations in your post:
    1. Rsen --> the assumption that higher current will result in higher Rsen is not correct. In this equation, if you include dVsl=40uA*Rsl in second part(that multiplies with D) will get smaller resulting in smaller overall Rsen.
    2. Rsen<infinity --> What this means in practice is that for D=0.5 (vin=12, Vout=24) there should be no sub-harmonic oscillations and part will be stable pretty much for "any" value of Rsen. If you would change your calculations so D>0.5 (ever so slightly) the resulting values in this equation start to make sense.

    I hope this attached application note will shed some light on the functionality of current mode converters and slope compensation.
    Best regards
    Brani