TLV2171: Pin7 has some ripple output being added

Hi Zoe, 

Could you clarify what the component values (resistors, caps) are in the schematic around the op amp? I'm having difficulty reading them on the image provided. 

Thank you!

Best Regards,
Ashley

Hi Zoe,

looks like the phase lead compensation(s) is missing.

By the way, I do not see a classic differential amplifier in the schematic. A differential amplifier has two inputs and subtracts the input signals from each other.

Please explain in detail what the circuit is designed to do. Why are there two outputs? And what is the purpose of the switch?

Kai

Thanks Ashley, please see attached schematic. It is used as to Select current or voltage as analog output to another inverter's analog input. So one routine is for voltage output, another routine is for current output.

Thanks for replying Kai! It is used as to Select current or voltage as analog output to another inverter's analog input. So one path is for voltage output, another path is for current output. Voltage path is the upper one.

Hi Zoe, 

I ran a stability analysis for the first stage of the circuit and did find that the circuit was stable before the relay. I've attached a TINA simulation for reference. 

zoe_TLV2171.TSC

Could you help to clarify what the relay part number is and what the circuit components are after the relay? The additional capacitance caused by the relay or after the relay point may be causing a stability and ripple issue but would need to confirm with the part number of the relay and in simulation. 

Please let me know if you have any questions.
Thank you!

Best Regards,
Ashley

Hi Zoe,

Kai

Kai, attached is the component value of the low path circuit. Relay is DIP switch and it's Pin 3 output connected several capacitors C107/C136/C110 to the analog OUTPUT.

Ashley, attached is the component value of the low path circuit. Relay is DIP switch and it's Pin 3 output connected several capacitors C107/C136/C110 to the analog OUTPUT.

Kai, The BJT was 40V/0.6A/225MW SOT23 package

They tested 3PEAK  and SGM both on the same board,no error waveform exist. 

SGM8292XS8G and TP2582-SR

Kai,Do you have internal circuit schematic that could share via email? Thanks!

Ashley, Do you have internal circuit schematic that could share via email? Thanks!

Hi Zoe, 

Thank you for the additional information of the schematic. We do not share information our internal circuit schematic as it is TI confidential information.

Do you have the part number of the DIP switch that is used in the application?

Thank you!

Best Regards,
Ashley

Sorry I don't have the part number. When customer changed to SGM8292 and TP2582, there showed no ripple on the pin7 output.

Hi Zoe, 

Are you seeing this issue with multiple TLV2171 devices? And have you confirmed that the issue follows the part (A-B-A swap)? 

Best Regards,
Ashley

Hi Zoe,

the problem in the circuit is not the OPAmp but the feedback circuit. Other OPAmps will also experience problems after some time.

With a 100nF load capacitance you urgently need to add a phase lead capacitance to recover the phase margin and make the OPAmp run stably. 220pF directly from the -input of OPAmp to the output looks good in Ashley's simulation:

Kai

Dear Kai /Ashley, 

Customer feedback if add 220pF on Pin7 and -in there would be no difference. But if delete R136 and 0.1uF capacitor on the DIP output , the output is common and  no ripple sampled from both the pin7 and dip switch output. Would you please help comment on why and simulate on Tina with this configuration? Mine has trouble on running the Tina software and showed garbled. Thanks!

Hi Zoe,

this would be the step response of your circuit:

zoe_tlv2171_2.TSC

Kai

And it could be improved to this:

Kai

Kai, 

What you advice in the first place is correct after customer evaluation. They add 220pF C4 between Pin7 and -INPUT already fixed the problem. Now they want to know more about how to set a proper loop parameter to make the output fixed to the proper phase margin? Could you please help explain how to set different RC load outputs and phase margins that could make a curve which they could follow as a good reference design in the future? Thanks a lot!