LMP2012: Settling time

Tamara Slim

Part Number: LMP2012
Other Parts Discussed in Thread: OPA387, OPA388

Hello,

I have witnessed that the LMP2012 takes a long time to settle after the input signal switches from one value to another. This settling time and error depend on the amplitude of the signal (it can go up to 700 us and 100 mV of error for a dVout of 4.5V). The gain was also found to have an influence (the greater the gain, the greater the error).

I have observed this phenomenon under different supply voltages (dual/single supply), configurations (inverting/non-inverting), different loads and different feedback impedances. (These waveforms were obtained with a LMP2012 in a non-inverting configuration with a gain of 20 (19k feedback resistor and 1k resistor to ground), and a 1k resistor as load)

I was wondering if this could be caused by the auto-zeroing technique: could it be overcompensation of the offset voltage as the input voltage changes rapidly? Is it linked to the offset calibration time parameter in the datasheet?

Regards

Tamara

over 1 year ago

0 Chris Featherstone over 1 year ago

TI__Mastermind 20490 points

Hello,

Our team will be looking into this and will respond within two business days. Thank you for your patience.

Best Regards,

Chris Featherstone

0 Michael Hartshorne over 1 year ago in reply to Chris Featherstone

TI__Genius 11795 points

Hi Tamara,

Sorry for the delay.

Yes, I believe you are right on with the offset calibration time parameter. You may have seen this post below that describes a similar question:

https://e2e.ti.com/support/amplifiers-group/amplifiers/f/amplifiers-forum/876826/lmp2012-offset-calibration-time

There are some additional details and a good app. note in the link above, but basically this auto-zero amplifier has two signal paths: a high speed path that will pass high-frequency signals (i.e. a high dV/dT edge) and a DC path that will correct for any offset errors. The high speed path allows the output to rise very close to the final value, and the settling to the accurate DC accuracy comes from the settling in the DC path. This settling is what the "offset calibration time" is referring to, and what you are looking at. I would expect this settling to account for a few mV of input-referred offset, which, when gained by 20, can get close to 100 mV.

I hope this helps answer your question. If this is a problem, we have other chopper-type amplifiers that have much faster settling, see OPA387 or OPA388.

Best Regards,
Mike

0 Tamara Slim over 1 year ago in reply to Michael Hartshorne

Prodigy 10 points

Hi Michael, thank you so much for your detailed answer.

I have one more question about the offset calibration time: in the datasheet the typical calibration time is rated at 500 us but the maximum value is 10 ms. Is there any information on the parameters that could raise the calibration time to 10 ms?

Best regards,

Tamara

0 Michael Hartshorne over 1 year ago in reply to Tamara Slim

TI__Genius 11795 points

Hi Tamara,

Unfortunately I don't have any more data on this parameter. This is an older device that came from National Semiconductor, we don't have the historical records to dig any further.

Sorry,
Mike

Amplifiers

Amplifiers forum

LMP2012: Settling time