I have been using in my application one ADG701 switch for the following circuit: I need from time to time to pull a 2MOhm resistor to the 5V rail (it acts as a pull up resistor to detect an open thermocouple). In normal operation the 2Mohm resistor should be left floating(it induces an approx 5 degree C error if it is connected to 5V because the thermocouples have very long wires).
I have used the ADG701 to connect the resistor to 5V or to disconnected. The Vcc of the ADG is the same 5V rail that is connected to the Source of the IC, the drain is connected to one terminal of the resistor. The input pin(controlling the state of the ADG) is controlled by an atmega644 Digital out pin (either pulled to High level - 5V, either to 0V).
When in on state the resistance of the ADG (S to D) is about 2ohms. When in off state the resistance is about 10-15k.
To test the behavior I have done the following: when in off state connect an microampmeter from the drain to GND. The value is approx 220uA. I have tested with 2 multimeters(Hioki and axiomet)
The second test: soldered a resistor from D to GND (approx 10K). measured the voltage accross the resistor(approx 1.8V-2V)
So, the current is there.
The ATMEGA for sure pulls the input of the ADG to 0.
Since the voltage through an analog switch can be bidirectional(as stated in the datasheet) and the voltage is in the range of Vcc / GND(as stated in the datasheet), what can be the problem?
The tests were conducted with the 2Mohm resistor disconnected, the source connected to 5V rail .
I have tested al the things regarding software issues, digital signals, etc with the oscilloscope and ohmmeter, even with pull up resistors to be sure that it's not a false 0V or left floating.
It is a little bit urgent because the design is used for production.
All the other mux-es(ADG632 and adg608) work good, but they propagate a voltage of 2.5V plus or minus 0.5V)