LT1498

 

Hello,

I have this circuit in datasheet of LT1498.

I would like to know, why have these 2 resistors and capacitors been added, (100 Ohm,1K and 500pF)

and how do you choose these values,

and maybe there's an article that clarifies that,

Thank you.

  • Hi,

    The 2 resistors and capacitor are added to avoid oscillation of the circuit especially when the sense resistor (0.5 ohm) is fairly low.

    The 100 ohm resistor is a MOS gate resistor to avoid parasitic oscillation or ringing caused by the gate capacitance in series with an inductor like a connecting wire.

    The 500pF and 1k ohm are for loop compensation and stability of the amplifier when driving the highly capacitive inputs of MOSFETs.

    You can fine tune the values depending on your target specs for settling time, step response time, phase, noise etc. You can use the LTspice for easy tuning of your circuit.

    Try transient simulation with C=500pF and C=50pF in LTspice and you will see the difference in settling time, ringing, overshoot of Iout.

  • Thank you for your response,

    but are there certain formulas? which one can calculate these resistors and capacitor, with this Mosfet

  • I've tried searching around but can't find any specifics or formulas on how to choose the values. I'll try to reach out to the part owner and get back to you as soon as I got something.

  • Hi Sisamyo:

    The series 100 Ohm resistor into the gate is a nominal value to avoid gate-drain capacitance interfering with the opamp output, and also to prevent oscillations local to the MOSFET.   In this case it may not have been absolutely necessary, due to both the C-load nature of the LT1498 output as well as the relatively well behaved nature of the Si9410.  But prior years of problems due to the lack of series resistance led to the rule-of-thumb of putting 100 Ohms there.

    The compensation rule-of-thumb is that the 1/2pi*RC should be well within the 10MHz gain bandwidth of the opamp.  The 1k : 500pF gives a crossover point of 300kHz or so, well within the 10MHz of the opamp.  Beyond that, one could tune it in as an empirical exercise dependent partly on the series inductance in the 0.05 Ohm path as well as whatever reactance is in the drain path.  The input would be stepped from 0 command to some value, or from one value to another, and the output checked for rise time and overshoot, whether from the perspective of the source or the drain.

    Glen Brisebois