JFET question

Thread Starter

mbohuntr

Joined Apr 6, 2009
446
Can a JFET be configured as a switch using only the circuit voltage instead of a separate supply?? Thanks!
 

Adjuster

Joined Dec 26, 2010
2,148
You could build a coupling / bias network so that fairly small supply voltage variations could switch the JFET on and off, but what would be the point?

Why would you want to use a JFET, rather than some other active device?
 

Thread Starter

mbohuntr

Joined Apr 6, 2009
446
I'm having stability issues with a circuit I'm building. http://forum.allaboutcircuits.com/showthread.php?t=48266&highlight=555+edge+trigger

If I disconnect the trigger wire for a few seconds. the PIR settles down. I was going to put the JFET in series in the trigger wire, and use the pulse from the 555 output to "disconnect " it for about 10 seconds. Is this bias network fairly simple, or is there another very low power component option? I am even looking at using the inverter to bias a 2n2222 at about .1mA. (using 400K resistor) Any thoughts?? (it's battery powered, so low current is really important)
 

Adjuster

Joined Dec 26, 2010
2,148
No, I thought you were asking an abstract question. The circuit I was thinking of would not suit your application.

To gate the input from your PIR, you need something like an analogue switch. You may be able to realize that using the channel of a JFET as a switch element, with a control voltage fed into the gate. Although there is some commonality between the input and control signals, that's not quite the same thing as "using only the circuit voltage instead of a separate supply".

Unfortunately, there may be some leakage from the control voltage into the signal path with this scheme. A dedicated analogue switch IC would probably work better and would be easier to design with.
 

Thread Starter

mbohuntr

Joined Apr 6, 2009
446
OK!!! Those nice folks at MAXIM sent me samples of their MAX324 analog switch. I read the datasheet, and here is my intended circuit. Multisim doesn't offer an analog switch component, so I used something that looked similar. The 5 sec pulse from the 555 will disable the PIR for the same period, hopefully giving the circuit time to stabilize. Is my placement and use of the diodes correct, and is there something I missed/ should add or change?? No coil means no back emf, right?? The quest continues.... Thanks for the help.
 

Attachments

Top