TLV9054: I would like to get an advise of OP-AMP for replacement of 9054.

Hello Hironari-san,

It looks like U1B is in a very high gain configuration. This design would have second stage DC gain of  1+ (R19/R17) = 1+(13k/2.2) ~ 5900 V/V

Do you mean to implement a circuit like this? I believe the circuit could oscillating because of this configuration. 

When I try to run this circuit in SPICE simulation the internal offset voltage of the device causes the output stage to saturate in U1B.

What do you want the filter to accomplish?

Thanks,

Jacob

Hello Mr. Jacob

Thank you very much your reply.

First of all, I apology my typo. 2.2 should be 2.2kohm.

I need this 2 stages filter and DC gain on 2^nd stage for system reason.

Those filters actually need to get DC out-put.

And the second stage DC gain is to eliminate one more op-amp.

As I mentioned, DC gain (13k and 2.2k,not 2.2ohm) need to add capacitance parallel to 13k to suppress oscillation.

However, to suppress the oscillation, the capacitance need more than 0.022uF.

I would like to change this value much small because of the response faster.

Filter cut off frequency is around 35kHz.

0.022uF x 13kohm = 0.286ms

0.286ms is slow to 35kHz.

That is a reason I need another OP-AMP if available.

I appreciate if you could some advises.

BEST REGARDS,

Hironari

Hello Hironari-san,

Thank you for the additional details, this makes sense to me.

I have checked the stability of the circuit in TINA-TI, and I see similar results with TLV905x:

Anything less than 30nF seems to cause rapid phase shift in loop-gain phase

Simulations with the TLV9064 look to be slightly more stable than the TLV9054, but it looks like we still suffer from rapid phase shift with anything less than ~30nF

What is the filter driving on the output?

Knowing the output load may help us understand if we can change the design of the circuit.

Thanks,

Jacob

Thank you very much your reply.

The output load will be 5kohm or 10kohm and no capacitance.

Again, I would like to put much smaller C13 to make response faster.

BEST REGARDS,

Hironari

Hello,

This additional pole in the feedback path is mostly dominated by the equivalent feedback resistance forming an RC time constant with the feedback capacitor. In other words, my simulations suggest that changing the amplifier may not significantly change the performance of this filter. If you want another part to test, I would give the TLV9064 a try. In my simulations the TLV9064 slightly improved the stability of the circuit, but not my much. 

The problem is that this interaction is causing further phase shift in the amplifier, and as a result, destabilizing the amplifier. I tried on my end to change the RC values in the feedback path, but this will not change the time constant of the circuit unless we alter the DC gain of the circuit.

Can you try to configure the amplifier with less DC gain? I understand that this will not perform the function you desire, but this will help me verify that the problem is sourcing from this particular portion of the circuit. 

Thank you,

Jacob

Thank you your reply Mr. Jacob,

And Sorry to late reply since I was out of office on Friday last week.

BTW, I actually need 7 times gain through 2 stage filter.

This time, I put this 7 times gain on 2nd stage filter.

So, What do you think if I divide 3.5 times gain on 1st staged filter and 2 times gain on 2nd staged filter?

 Is this something help?

Or,  I will add another amplifire just for 7 time gain after 2nd stage filter that I tried not to. But I might have to.

I appreciate if you could provide any advise.

BEST REGARDS,

Hironari