LMC662: Measuring the voltage from a very high Input Impedance Source

HARI T O1

Part Number: LMC662
Other Parts Discussed in Thread: LMP7721

Dear Team,

We need to measure the voltage from a charge pump that is present inside the ASIC.

The Impedance of the charge pump is 200T Ohm. The team has designed a circuit with LMC662A which has an input impedance of 1Tohm.

Since the Impedance of Opamp is very less compared to the source, it will load the source. May I know the below circuit will be able to measure the voltage properly?

BOOT_STRAP.TSC

Regards

HARI

over 3 years ago

0 kai klaas69 over 3 years ago

Guru 140200 points

Hi Hari,

an output impedance of 200TOhm??

Forget all measurements...

Kai

0 HARI T O1 over 3 years ago in reply to kai klaas69

Expert 1215 points

Hii Kai,

I will check with the team one more time.

Will update you

Regards

HARI

0 Ron Michallick over 3 years ago

TI__Guru*** 140146 points

Hari,

By inside the ASIC, I assume you want to land a needle probe to the silicon die. Just landing the probe will change the circuit if the impedance is even remotely close to what is claimed.

I suggest a different approach; measure something else (ideally an external pin) that varies in relationship to the charge pump voltage. Then force a voltage on the internal charge pump node. Then vary that voltage until the 'something else' measurement is same as the no probe value. The forced voltage at this point should be the natural (unloaded test) voltage.

0 HARI T O1 over 3 years ago in reply to Ron Michallick

Expert 1215 points

Hi Ron,

We have a breakout board in which the signal is coming to a connector .The output of the ASIC will go to the circuit shown above and we will be measuring at the output of the circuit shown above.

The impedance varies from 10Gohm to 10Tohm

Regards

HARI

0 HARI T O1 over 3 years ago in reply to HARI T O1

Expert 1215 points

Hi Kai,

Input impedance varies from 10Gohm to 10Tohm.

For 10Gohm I think the above circuit is okay,but for 10Tohm ?I don't think it will work.

May I know your thoughts.

Regards

HARI

0 Ron Michallick over 3 years ago in reply to HARI T O1

TI__Guru*** 140146 points

Hari,

For 10T, I hope the signal length is very short. Any length will pick up external noise signals or if shielded the capacitance will make the charge and discharge very slow. There is also a possibility that the coupling from op amp's variable V- and V+ though op amp or board will over power the input signal.

Are you sure the ASIC has an enabled pin impedance so high, that value seems ridiculous to me.

0 HARI T O1 over 3 years ago in reply to Ron Michallick

Expert 1215 points

Hi Ron,

Thank you very much.

I checked with the ASIC team. The impedance varies from 10Gohm to 10Tohm depending on the leakage current.

May I know what are layout precautions we need to take. Will guarding the input will help us?

Regards

HARI

0 kai klaas69 over 3 years ago in reply to HARI T O1

Guru 140200 points

Hi Hari,

a charge pump with a source impedance of 10T makes no sense to me at all. You generate a voltage within an ASIC but you cannot load it because the source impedance is 10TOHM? What shall this charge pump be good for?

Kai

0 Ron Michallick over 3 years ago in reply to HARI T O1

TI__Guru*** 140146 points

Hari,

I think this spec is the impedance of the charge pump pin when it is turned off, in other words open.

You will need to measure a current, not a voltage.

0 HARI T O1 over 3 years ago in reply to kai klaas69

Expert 1215 points

Hi Kai,

This charge pump is used for providing voltage to a MEMS.

That is why it's impedance is very high.(from design team)

Regards

HARI

0 HARI T O1 over 3 years ago in reply to Ron Michallick

Expert 1215 points

Hi Ron,

This charge pump is used for providing voltage to a MEMS.The ASIC is a part of MEMS microphone.

We nned to measure charge pump voltage only.

Regards

HAri

+1 kai klaas69 over 3 years ago in reply to HARI T O1

Guru 140200 points

Hi Hari,

I would give the resistive voltage divider a try. Put one or more 10T resistors in series and form a voltage divider in combination with a 10T resistor. Sense the voltage across this 10T resistor by the help of LMP7721. 3fA input bias current times 10T gives an acceptable offset voltage of 30mV. And 20...100T might provide an acceptable load to the charge pump.

Kai

Edit: Of course, the voltage at the +input of LMP7721 must not exceed the common mode input voltage range of LMP7721! So the voltage divider must be modified accordingly.

+1 Ron Michallick over 3 years ago in reply to HARI T O1

TI__Guru*** 140146 points

Hari,

Here is an article written by Paul Grohe that will be helpful. https://www.edn.com/design-femtoampere-circuits-with-low-leakage-part-one/ Be sure to read all four parts. The op amp chosen for final experiment is the LMP7721 which has IIB limits, not just a typical value. However max supply voltage is lower.

Amplifiers

Amplifiers forum

LMC662: Measuring the voltage from a very high Input Impedance Source