I am designing an integrate and dump circuit using an Op amp integrator, but I don't know how to reset the output every T seconds, any one got an idea?
A CMOS switch should work - it depends on which one, how large your capacitance is and how quickly you want it to discharge. If you gave more details I could be more specific.
Using a single MOSFET depends on the voltage you have to discharge and where you are trying to discharge it to. That's why analogue switches have both PMOS and NMOS.
I am discharging 1 volt from an 0.5 nF capacitance. I want it to discharge as fast as possible, ideally I want instantaneous resetting as I am implementing this circuit as a part of a BFSK receiver, and I have to reset the output every 1 ms to detect a new symbol that lasts for 1 ms, any delay in the discharge will cause ISI.
What switch have you tried? Instantaneous doesn't exist - you need to put a number on it. To discharge 500pF from 1V in say 1us only takes 500uA so a fairly low grade analogue switch could be able to do that. You have to watch out for charge injection though. By coincidence I am looking for an analogue switch at the moment to switch in <10ns with <10pF of capacitance.
Attached pics are for the switch I'm currently testing. Try using it if you're interested.
The high control voltage opens the switch and low voltage closes it.
I don't understand your circuit or simulations. Where is the capacitor you are discharging? While your simulation may work, you seem to have a 1 ohm resistor in the base of the transistor. Depending on what you drive the transistor with, you would blow it up. I would have used a MOSFET anyway - a bipolar transistor won't discharge to zero volts.
I am sorry, I only included the switch and a simple test for it with no capacitors, I attached now the whole Integrate & dump circuit and its simulation, the switch subcircuit works well with the capacitor, but I don't think you'll be interested in viewing the whole schematic !
I shall consider the 1 ohm resistor issue also.
Thanks a lot for your time and help, I appreciate it.
may I remember you that Keith has recommended an "analog switch" (that's an analog gate!). ?
Does your BJT switch work for both polarities of the capacitor voltage?
I think you should look at a small signal MOSFET instead of bipolar. As I mentioned, a bipolar cannot discharge to zero volts - it has a saturation voltage. You can see the problem on your simulations.
Also, bear in mind that with a single transistor you cannot discharge any voltage in any direction. That is fine if you only ever have positive integration voltages, for example - you can put the right transistor the right way round. If you need to be able to handle positive & negative voltages then you will need two transistors.
I am integrating unipolar symbols, thus, I have no negative polarity in consideration.
Yes, I realized the problem of saturation voltage and I'll switch to a MOSFET transistor.