Glitch from ADG1633 spice model switch

Hi Leonardo,

I keep on seeing different response on my simulation. Can you check the attached files if you have attached the correct circuit that has the same analysis with your screenshot?

Also, I noticed the ENb has is floating. Do you have a working spice model for TL072? Have you uploaded it in your LTspice library?

Regards,

May

Hi May

I apologize for the delay in reply !

Yes, screenshot attached belong to my original circuit.

I discovered some interesting things about the ADG1633 macromodel.

I plotted the waveform of Ix(U5:D) and it seems to me, at first glance, that there is a switch that is causing these glitches.

Note that current reaches 622 mA. Would this be normal behavior ?

Hi Leonardo,

This is what I see when I run your simulation file.

Is there anything that I need to change in the file?

Also these are some observations with regards to the ADG1633 operation:

• EN pin is floating.
• Vss pin = -12V. Datasheet specifies Vdd/Vss = ±3.3 min to ±8max only.
• Digital Input = 12V. Is this correct?

One of these conditions might be the cause of the model's unexpected behavior.

Regards,

May

Hi May

I forgot to say that to run the simulation is necessary to switch to alternate solver. In this way, it converge.
I modified the circuit to match your observations:

- Connected the EN pin to ground;
- Reduced the power supply voltage to 6V;
- Digital pin is now with 6V.

However, glitches still appear in the output.

Regards,

Leonardo

6735.sync_filter_to_full_elimination_ripple.zip

Hi Leonardo,

Thanks for the clarification.

The glitch that you see is caused by charge injection parameter of the model.

May I know if charge injection is an important requirement in your application? If not, would you prefer to remove it to the model so you may proceed with your simulations?

Regards,

May 

Hi May

Charge injection isn't a critical parameter to my simulation.

To "clean" the signal, I put an RC filter on the output of U4. So, i was able entirely eliminate the glitches after some milliseconds.

Now, I´m satisfied with the simulation results and there is nothing more for me to add.

Regards,

Leonardo

7382.sync_filter_to_full_elimination_ripple.zip 

Hi May

Charge injection isn't a critical parameter to my simulation.

To "clean" the signal, I put an RC filter on the output of U4. So, i was able entirely eliminate the glitches after some milliseconds.

Now, I´m satisfied with the simulation results and there is nothing more for me to add.

Regards,

Leonardo

7382.sync_filter_to_full_elimination_ripple.zip