Output error when using the AD8657 as a voltage follower

Dear Osamu, 

do you experience this behaviour in real life or only in the simulation? If you download the spice model from analog.com and open it with a text editor, it seems like it only models 18 V operation at 25 °C.

Dénes

Thank you for your answer.

The offset error is confirmed by actual measurement.
The measured value is larger than the offset error of the voltage follower confirmed by LTSPICE.
I don't know the cause, so I consulted.
Actual measurements are at room temperature.

Thank you.

Hi Osamu 

Can you try on your circuit to increase the supply voltage to 18V then measure again the offset voltage?

We are currently checking on this one as I am modelling it on Ltspice and there is a change in the offset voltage once we increase the supply.

Regards,

JE

Hi Osamu ,

Meanwhile just some questions that may help to isolate the source of the problem:

1. Are there any load on the opamp output? The datasheet says: "The amplifiers are capable of driving resistive loads down to 100 kΩ."

2. How do you measure the output voltage? DMM or oscilloscope? Are you sure there are no AC components, no oscillation at the output? Any capacitive load by the instrument? R19 seems to be rather low to isolate possible cap loads.

3. How do you measure the input on pin3 of U4A? It is a clean 1.8V DC?

4. Do not let the inputs of U4C float. Connect it as a follower:

For the reasons and possible problems see "What shall we do with an unused op-amp?"

5. It seems you have a supply voltage close to 10V and operate the opamp at room temperature. So the offset is specified to be less than 350uV:

Even for lower supply this is the same:

It is hard to believe that during normal operation one can get 100 times higher value the specified as maximum, so I do not think this amplifier can be the source of the problem. I hope the above can help to get closer to the reason.

Zoltan

Hi Zoltan

Thank you for your polite answer

1 R:100Ω C:0.1uF RC filter and Connected to the MAX11608.

2 Put probes on pin 3 (-IN) and pin 1 (VOUT).

3 and 4 　I'm using it as a power source for the bridge sensor. There are no NC pins.

5. I apologize for not including the time of actual measurement. There is an error of 1000uV to 700uV.

Regards

Hi,

Osamu said:

R:100Ω C:0.1uF RC filter and Connected to the MAX11608.

I think the AD8226 has better ADC drive capabilities than the AD8657. I would drive the ADC with the inamp.

Osamu said:

I'm using it as a power source for the bridge sensor.

If the bridge sensor has a resistive load below 100kΩ, the datasheet recommendation is not met, so the opamp specs may not be valid.

Zoltan

Hi Zoltan.

What is the mechanism that affects the offset when the load is not the recommended value?

Since then, we have been conducting reproducible experiments with even lighter load resistance instead of bridging sensors.
However, the phenomenon that the offset is larger than the regulation has not been confirmed yet.

Regards

Hi Osamu ,

I do not think the offset is simply related to the load. If you havve excessive load, the output may not follow the input properly. However, in any case one should always keep the operation within the specified range.

Osamu said:

The error range is roughly 30mV to 40mV

Osamu said:

There is an error of 1000uV to 700uV

I guess tha latter is the case ro they are not about the same thing. How do you measure the offset? Using a DMM? In what range, resolution?

Zoltan

Hi,

I believe that the questions have already been answered so I am closing this thread. But if there are any questions, you can post a new thread. 

Thank you!

Regards,

Mae