The datasheet information (Page 4) mentions 3 to 4mA for Idd @VDD=3.3V. While not switching (IN1..IN4=0V), the measured power consumption matches with the specs.
However, while switching, the power consumption is much higher - I am not sure if there is something wrong with my circuit or if this specific information is missing at the datasheet (or if I was not able to find it..).
ADGM1004 consumes only ~3mA (typical) of current during normal operation. IDD remains around 3mA during switching cycle and also when you are not performing switching.
Can you please tell me how much IDD you are measuring during switching and while you are not performing switching.
Also can you please share your schematic to get better understanding of your circuit.
I did not measure the current at ADGM1004, but before a voltage regulator - I will check that circuitry and I will also measure IDD with and without switching and let you know the numbers.
By the way, do you know how much IDD drops if we set EXTD_EN to high thus disabling the internal 11.5 MHz oscillator and driver boost circuitry?
Also, what is the expected "ready-to-switch" latency if one controls EXTD_EN pin between switching operations only to save energy (i.e., no intention of using the scheme "VCP pin must to be driven with 80V dc from an external voltage supply")?
If you set EXTD_EN high, the internal oscillator and driver boost circuitry shuts down. In the condition ADGM1004 consumes around 50uA max current.
You can set EXTD_EN high when you are not performing any switching to save power but you need to wait for 0.55ms (typical) for power up to make sure the 47pF VCP cap is fully charged before you perform switching.
I hope this helps.
this information was extremely helpful. I was using a "load-switch" to turn-on/off ADGM1004 while not being used by the MCU. But 50uA is not bad - almost a device "shutdown" pin.
In fact, 50uA and 0.55ms are acceptable parameters for my application - I will eliminate the load-switch and directly control the EXTD_EN.
It seems that the high current I reported before was due to an issue with the voltage regulator, not with the ADGM1004. But I still need to put a sourcemeter to carefully measure Idd with/without switching.
Can you please share schematic or even block diagram to get better understanding of your application. By the way for what type of application you are using ADGM1004.
sure. Give me a couple of days to conclude the final details of the circuit and corresponding energy profile for ADMG1004.
The device I have been designing is called "HF Cap-Cal" and it is used to provide "calibrated" values of capacitances. Since the target devices (sensors for dielectric measurement of materials) to be calibrated operate at frequencies higher than 50MHz, off-the-shelf calibrators at the market cannot provide a proper response.
To make the design more complicated, the MUTs (e.g., soil, food) for those sensors present a very high dynamic range in terms of impedance and 50-ohm transmission lines cannot be applied. Any wire going outside the "volume of influence" of the sensor becomes a problem due to EMI issues. ADGM1004 provides to me a chance of having the capacitance-switching in a physically very tiny space that is paramount for the calibration without having unacceptable noise levels. However, if wires are used even to power the curcuit, the noise level then jumps beyond our limits.
Therefore, the solution I envisioned was to use supercapacitors that have enough energy for almost 2min. of ADGM1004 and MCU operation which is more than the double the necessary time to conclude the calibration. That's why the energy consumption is also critical. In short, when the calibration fixture is "open", the ADGM1004 circuitry is quickly charged (few seconds) and, when it is closed, the MCU starts the switching operation that allows the RFC to reflect 16 combinations of capacitances, since a different capacitor is connected to RF1,.. RF4. In short, no external wires and no batteries are involved in this solution.