Using FET as a switch to bypass a resistor?

Thread Starter

redbird_is

Joined Jan 13, 2008
11
Hi everyone,

After much experimentation and Spice simulations, I've realized I require some additional insight on a circuit that I'm trying to design :)

The attached picture provides a description & schematic of the circuit and problem. Basically, this is what I am wondering:

I'm trying to use a FET as a switch to bypass a resistor (the resistor is connected in parallel with the source and drain). I believe this should work under DC conditions, but the resistor is part of an AC audio circuit. Is this kind of FET switching possible?


Bypassing the resistor (i.e. switching on the FET) essentially sets its equivalent resistance to zero. I'm using this property to try and create a variable low pass filter.


Thanks for reading my post!

Michael
 

Attachments

Ron H

Joined Apr 14, 2005
7,063
You can't easily use JFETs, because you will get gate current whenever the gate voltage is more positive than either the source or the drain. This will short out your AC signal, and screw up the DC bias on your op amp. To turn on a JFET, the maximum Vgs is 0V, in order to avoid this problem. The voltage to turn them off is generally Vgs < 0 by several volts (see datasheet). Vgs(off) for the 2N3819 is -8V.
Take a look at MOSFETs such as 2N7000, 2N7002, or BS123. They have insulated gates. You will need about Vgs≈+10V to turn them fully on, and -1V (or more negative) to turn them off, given the ±1V signal source.
 

Thread Starter

redbird_is

Joined Jan 13, 2008
11
Thanks for the reply!

I hadn't realized that about JFET gate current.

I'll try a new spice simulation using the FETs that you mentioned (The gate resistor / diode is no longer necessary, right?).
 

eblc1388

Joined Nov 28, 2008
1,542
Take a look at MOSFETs such as 2N7000, 2N7002, or BS123. They have insulated gates. You will need about Vgs≈+10V to turn them fully on, and -1V (or more negative) to turn them off, given the ±1V signal source.
And then wonder about what would be the effect the intrinsic diode does to the signal....
 

Ron H

Joined Apr 14, 2005
7,063
And then wonder about what would be the effect the intrinsic diode does to the signal....
Good point! It totally slipped my mind.
My first thought was CMOS analog switches. I guess I should have stuck with it.
SD210/214 should also work, but they are very susceptible to damage from static discharge. SD211/213/215 should also work, but you have to contend with the gate-protection zener.
 

Audioguru

Joined Dec 20, 2007
11,248
Your filter needs to have both of its resistors changed to change the cutoff frequency.
But you are changing only its input resistor.
The values in your filter are odd and it has an extra very high value resistor. Look up Sallen and Key Lowpass filter.
 

Attachments

Top