Since the very beginning of the project I've been using the AD620 instrumentation amplifier from Analog Devices, it works very well, the only thing that's quite annoying is that it operates in single supply on voltages above around 5V, but it doesn't matter a lot at the moment.

The AFE (analog frontend, I'll be using this word a lot) is almost completed, the only thing that I require is a Lead off detection system, imagine if you're using a wheelchair controlled by the EyesDrive AFE and an electrode gets disconnected: the AFE amplifies random noise and the wheelchairs begins to go around randomly, possibly hurting the user.

So I've used what's called a DC Lead off detection system, it requires just 2 pull up resistors and two comparators: when an electrode gets disconnected, it pull's up the input pin: here the comparator will check the voltage and, if it is above a certain threshold, it will trigger the system doing an emergency stop.

The Lead off works, but when i put the resistor between the input pins of the AD620 and the V+ rail the amplifer begins to behave strange, it amplifies random noise and the EOG isn't present at all. 

If anyone has any advice (even if you know a better chip, it could be actually the real problem), please tell me, it would be very helpful!