What is the max current consumption of AD8667?

Hi,
I am curious about the following AD8667 max current consumption.
From datasheet, it says that power supply is less than 1mA, but what is the max current consumption? Pin7 max output is a 0V ~ +3.3V waveform. And ADC1 link to STM32.
Any comments?

Best,
Tony Liu

It depends by the load, and cannot be more than the short-circuit current. The datasheet shows that the short-circuit current is +/- 7mA.

More theory about this topic:
http://www.eetimes.com/document.asp?doc_id=1272353

Hi,

I can′t find the "1mA" specification.
It says max. 275uA (@25°C) per amplifier = 550uA.

The first OPAMP drives into a 20k load. Let′s say max. 3V --> 150uA
The second OPAMP drives into 20k load --> again 150uA
(20k feedback, and I assume the ADC input is high impedance)

The second OPAMP is inverting, therefore when the first OPAMP′s output is positive then the second OPAMP′s output is negative. Therefore you will never see both 150uA at the same time.

In detail it depends on signal waveform, signal voltage level and signal frequency.

And it depends on what "current" you mean: average? peak?...

I estimate the total average current in each power supply line will 400uA ... 700uA.

****
BTW: the 100R / 100n + 100p is not suitable here.

What do you want to achieve?
HF suppression in the MHz range? --> then it′s OK. (but where do you expect the HF come from?)

To stabilize the supply for the OPAMP? --> 700uA will cause a drop of 70mV .. this may be good or not.
The cutoff frequency is about 16kHz, and the PSRR at this frequency is less than 40dB... so a voltage ripple at the supply may cause a reasonable coupling to the output...
Indeed all frequencies above 16kHz will be attenuated by 40dB only (because rising slope of PSRR combine with fallsing slope of fiilter)

****

--> I recommend to use a 10u (or even 100u) instead of the (i think: unnecessary) 100p. This causes a cutoff frequency of 160Hz and an attenuation of 80dB.

Klaus

Added:
Short circuit current:
For sure this is 7mA per amplifier. But I didn′t take this into account. My estimation is for normal operation.

Hi, KlausST,
Your analysis is surely right.
Yes, the 100p is for decoupling RF which coexists in the same pcb board.
Thanks for your PSRR analysis.
Many thanks.
Best,
Tony Liu

I'm a bit confused.
OpAmp average current depends on signal waveform and signal frequency?

I just tested a single supply 1MHz gain bandwidth OpAmp at various frequencies (inside of the gain bandwidth) using various input signal waveforms.
During measurements I kept the output resistive load constant.

For all frequencies and waveforms I don't see any significant variation in the DC average current consumption (less than +/- 5% variation).
My setup has the ability to measure down to 1uA average current.

But if I change the resistive load value, I get huge average current variations.

Hi,

It depends on the circuit...

But even without external load the supply current will change with frequency. Internal capacitors need to be charged and discharged, causing increased supply current.

External load current
... is added to the internal circuitry current.
* Imagine a capacitive load, wich depends on frequency
* imagine a speaker as load... it is obvious, that the speaker current needs to "travel" through the supply pins of the amplifier.
* or any other (resistive or not, linear or not) load current needs to travel through the supply pins.

And the load connection matters:
Let′s assume an opamp supplied with 0V / 5V.
Now consider a 1k resistor connected to the output and 0V.
All the resistor current is added to the positive supply current, but the negaive (0V) supply current will NOT be influenced that much.
If you now connect the same resistor beween output and 5V, then the 0V supply current will be influenced, but not the 5V supply current.

Klaus

On my bench doesn't depend by the circuit.
In standard inverting, or non inverting, or just buffer configuration, I don't see much current variation with signal frequency or waveform.
For the same single Vcc, the significant current variation happen ONLY when I change the resistive load, wherever is connected.
Is a fact, all the other factors don't affect much the DC average current consumption. I'm speaking here about tens of times difference.

The reason that I am pointing to this situation is to do not spread the confusion, because I remember I did this test for the first time many many years ago, when I believed that high frequency tones will make my HiFi amplifier to surge more DC current