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.

LM4702 high frequency instabilities?

Hello everyone,

I'm writing because  I've got a strange result in the frequency vs gain graph when testing an LM4702 application, with IRFP240/9240 vertical MOSFET output devices, I've got a strange gain peak at high frequencies, this can be caused by defficient Miller compensation?

I can't also simulate output triples and output stage with driver bipolar transistors.

TINA diagram and Bode plot attached, the circuit has a multislope VI limiter but without it, the same problem occurs.

Best regards,

Daniel Almeida

8535.LM4702_multislope.TSC

8838.LM4702_multislope_driver.TSC

 

  • Hi Daniel,

    What value of compensation cap are you using for the LM4702?

    Could you try increasing this value and re-simulate? Possibly 15pF or 20pF.

    Do you have any base resistance from the LM4702 outputs to the FET inputs? If so, what value?

    I'm not sure what you mean by you can't simulate with output triples or driver bipolars; as long as you have a model it shouldn't be a problem.

    Best Regards,

    JD

  • Hello Engineer John DeCelles,

    I'm using 10pF, but increasing the value doesn't help, the graph is always similar.

    I don't know what you mean with base resistors, are you talking about the bias resistors, gate stopper resistors or the resistor connected between sink and ground to improve 2nd harmonics distortion characteristics?

    With driver transistors to improve the slew rate, the simulation stops and says "Operating point not found. Component: U1.XU1.XD3.D1"

    The TINA schematic is attached to the previous post,

    Thank you very much for your support,

    Best regards,

    Daniel Almeida

  • Hi Daniel,

    Try decreasing R14, R15, R18 and R19 to about 225 to 250 ohms.

    JD

  • Hello Engineer John DeCelles,

    The problem only disappears totaly when the resistors (R14,R15,R18,R19) have values inferior to 10 ohm (actually I don't have any gate stopper resistor), but without gate stopper resistors parasitic oscillations could occur in the MOSFET gate junction, I don't understand what means that "gain peaking" in the graph I've sent and what it's causing it.

    I hope you can help me.

    Thank you very much for your support,

    Best regards,

    Daniel Almeida

  • Hi Daniel,

    When I substituted 220 ohms for R14, R15, R18 and R19 the peaking has been minimized as shown in the attached. This result is acceptable in my opinion.

    Best Regards,

    JD

  • Hello Engineer John DeCelles,

    So I can use 220ohm instead of 470ohm gate resistors, but what causes this gain peaking?

    I have a BUZ900/905P model in a .txt file and I wanted to use it instead of the IRFPs, but I don't know what I have to do to use the model with TINA.

    Thank you very much for your help,

    Best regards,

    Daniel Almeida

  • Hi Daniel,

    If you follow section 7.1 in SNAA045, it describes selecting the proper gate resistor value. I presume that while it may appear that 220 is an optimum value, you will find through real-world bench testing after you build your prototype that the value may change, especially if you are changing output power devices. 

    I do not know what is specifically causing the peaking in the frequency response; I would presume it has to do with the FET model. Maybe others on the forum wish to comment. I do know that the models have limitations; Bob Cordell's book has a whole chapter on models and goes over some of the differences between them.

    If you need help utilizing a different model for TINA, I would suggest you post in the modeling forum; they will know how to better help you with that.

    Best Regards,

    JD

  • dont do anything just put a 8 ohm resistor as load i believe you have forgotten to put the load resistor before you simulate...

    now what you seem you have done is shifted the pole to RF which will create fireworks.. 

    dont panic there is nothing wrong if you are using IRFP240 9240 use LME49830 which is actually ment for the mosfet and hence ment for more VBE which is actually essential for the mosfet or hexfets..

    if its transistors then use lme49810 for high power but for lower power class A and pure operation use 811 without a additional driver it can drive one pair of transistors 9ma output

    lme49830 can drive two pairs of irfp240 9240

  • Hello everyone,

    The output resistor is the one with 4 ohm.

    This is only a simulation with IRFPs, in my application I'll use two pairs of BUZ900/905 instead.

    I don't know really what is causing this gain peaking.

    I'm not sure if the very high frequency oscillations will "blow" the output devices, because the speaker is an inductive load so the impedance increases proportionally with frequency (Zinductor = jWL), W = 2*pi*f [rad/s], L is the inductance value in Henry, so the only drawback is an excessive power dissipation which could trigger the thermal protection circuitry included.

    Thank you very much for your support,

    Best regards,

    Daniel Almeida

     

  • How did you measure the bias current ? I put the ampermeter to R16,R17,R20,R21 but I couldn't see any change when i adjust R1,R4,R14,R18,R15,R19. I have same problem at my circuit ?

    e2e.ti.com/.../2281632