0
\$\begingroup\$

I have 2 MOSFETs :

enter image description here

NMOS -->

     Ref :BSZ520 
  Rds_on : 52 mohms

PMOS -->

     Ref :IRF9640 
  Rds_on : 500 mohms

My load (Cact1) is drawing around +/- 250 mA. I calculated the power dissipation with P= R*I² which gives me the following results :

BSZ520 -->

       P = 3 mW

IRF9640 -->

       P = 31 mW

Then why when I use ALT+right click, LTspice gives me hundreds of watts? I hope it's not relevant.

enter image description here

\$\endgroup\$
2
  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$
    – Voltage Spike
    Apr 30, 2021 at 0:42
  • \$\begingroup\$ Sounds about right for an amplifier that deep in Class A. \$\endgroup\$
    – user16324
    Apr 30, 2021 at 12:00

1 Answer 1

3
\$\begingroup\$

LTspice gives me hundreds of watts? I hope it's not relevant.

Unfortunately, it is relevant. You have shoot-through, i.e. current flowing through both transistors at the same time.

schematic

simulate this circuit – Schematic created using CircuitLab

With the biasing resistors as in the schematic, significant current flows through the transistors even when the input voltage is 0.

enter image description here enter image description here

By decreasing the 5k\$\Omega\$ resistors to 2k\$\Omega\$, the lost current became dramatically reduced.

enter image description here

There may be some cross-over distortion, but it is "small" for some definition of "small". If this is for audio purposes, it might be audible. My own ability to discern "high fidelity" is rather weak.

enter image description here

However, I simulated at 5V, 2V and 1V, and even at a 1V input, cross-over distortion is not visible in the graph. enter image description here

\$\endgroup\$
6
  • \$\begingroup\$ Does it mean that my biaising is not correct ? Because i need the mosfets to react quickly to avoid cross over distortion \$\endgroup\$
    – AF2021
    Apr 29, 2021 at 21:12
  • 1
    \$\begingroup\$ Yes, you are biasing the MOSFETs too strongly. Unfortunately, MOSFET threshold voltages vary from device to device, so it is hard to get an exact value for the R7/8 resistors that will be "ideal". I decreased them to 2k in a simulation, and the shoot-through fell to micro-amps. The cross-over distortion was not visible to me on the output voltage plots. However, it may significant to you, especially if this is an audio circuit. Some people may or may not notice. \$\endgroup\$ Apr 29, 2021 at 21:48
  • \$\begingroup\$ Thanks a lot for all the effort ! I tried your suggestion on Ltspice but unfortunately the distortion is not acceptable ( and you guessed right, it is for audio, so i have to get very clean signals) \$\endgroup\$
    – AF2021
    Apr 29, 2021 at 22:07
  • 1
    \$\begingroup\$ If it were not for the fact that over-current will make your mosfets toast, I would suggest a trimmer pot (variable resistor), so you could adjust the bias to just the right spot. Unfortunately, one movement of the trimmer pot too far, and poof! \$\endgroup\$ Apr 29, 2021 at 22:11
  • 1
    \$\begingroup\$ Also be aware that the exact bias resistor value will depend on the individual MOSFET characteristics so you can't just set it by design. Furthermore it may vary with temperature in the right direction for thermal runaway. Not an easy problem : you might consider using those current sense resistors R12,R16 to control the bias voltages with a negative feedback loop (perhaps by replacing those 100K resistors with controlled current sources) but it's not a trivial piece of design. \$\endgroup\$
    – user16324
    Apr 30, 2021 at 12:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.