LM4871: Discrepancy between Spice and data sheet

Vincenzo Costanza

Part Number: LM4871

Hi,

we were modeling the basic configuration for the LM4871 both in PSpice for TI and LTSpice as shown in the part's data sheet (input signal to the inverting input of the amplifier). However, the simulation results do not match with what we were expecting (signal across the speaker to the positive rail). When we send the input signal to the non-inverting input of the amplifier we obtain the expected waveform. We saw other posts on the forum mentioning similar issues and we wanted to double check: are the inputs of the spice model flipped?

Thanks for the support

over 1 year ago

0 Ivan Salazar over 1 year ago

TI__Guru** 100575 points

Hi Vincenzo,

Do you have a link to that other post with similar issues? Also, if you can add any captures of your simulation and show what seems to be wrong it would be helpful.

Best regards,
-Ivan Salazar
Applications Engineer

0 Vincenzo Costanza over 1 year ago in reply to Ivan Salazar

Prodigy 10 points

Hi Ivan,

this is the post I was referring to: https://e2e.ti.com/support/audio-group/audio/f/audio-forum/559924/lm4871-reference-design. I'm attaching the sim where the input is connected to the non inverting input and working properly

while this is when the input is connected to the inverting input (input signal is green, voltage across the speaker is blue)

we also notice that in the spice part the PIN numbers do not correspond to the ones reported in the datasheet

0 Ivan Salazar over 1 year ago in reply to Vincenzo Costanza

TI__Guru** 100575 points

Hi Vincenzo,

Thanks for the extra details. I'll double check the spice model and let you know the observations during today or tomorrow before end of the day.

Best regards,
-Ivan Salazar
Applications Engineer

0 Vincenzo Costanza over 1 year ago in reply to Ivan Salazar

Prodigy 10 points

Hi Ivan, was wondering if you had any update on this.

Thanks

0 Ivan Salazar over 1 year ago in reply to Vincenzo Costanza

TI__Guru** 100575 points

Hi Vincenzo,

Apologies for the longer delay than expected. I'll make time to check on this today and provide some feedback.

Best regards,
-Ivan Salazar
Applications Engineer

+1 Ivan Salazar over 1 year ago in reply to Ivan Salazar

TI__Guru** 100575 points

Hi Vincenzo,

After review I can confirm that:

Proper device implementation is as features in the data sheet. Input signal must be fed to inverting input while non-inverting input is shorted to bypass pin/cap.
Simulation model functionality is correct, however the input labeling is incorrect, +IN should be -IN and -IN should be +IN.

Best regards,
-Ivan Salazar
Applications Engineer

0 Vincenzo Costanza over 1 year ago in reply to Ivan Salazar

Prodigy 10 points

Hi Ivan, thanks for confirming that. I was wondering if we can get the updated spice model with the correct labeling.

Thank you

0 Ivan Salazar over 1 year ago in reply to Vincenzo Costanza

TI__Guru** 100575 points

Hi Vincenzo,

This wouldn't be an officially released model, but I can look into modifying the symbol, I'll send it back within a couple of days.

Best regards,
-Ivan Salazar
Applications Engineer

0 Ivan Salazar over 1 year ago in reply to Ivan Salazar

TI__Guru** 100575 points

Hi Vincenzo

I've modified the symbol in the attached file, you may use this project or the symbol/PSpice associated object itself.
Let me know if any further questions.

snam137_edit.zip

Best regards,
-Ivan Salazar
Applications Engineer

Audio

Audio forum

LM4871: Discrepancy between Spice and data sheet