I am new to using jfets and need help,with a grid dip oscillator design that I have built,from a design I found online.The circuit built can be found in the enclosed attachment.The meter part of the circuit works fine,but the front end from the tuned circuit forming the oscillator does not.Please can anyone advise the correct connections for the jfet as the two mpf102's in the circuit description have their source connections joined together.The power supply as seen has a positive ground connection.Thank you and I look forward to any help and suggestions anyone may have.

 

Wow, that was a confusing looking circuit, using positive ground, a holdover from the days when most transistors were PNP.

The circuit looks "real" and we should give the author the benefit of the doubt (his web page has the same schematic on it, so it probably is not in error). If you properly identified the leads on the JFETs, the oscillator should oscillate, at least at some frequency or another.

How do you it doesn't oscillate (lack of meter reading?) Can you hear it on an AM, CW, or sideband radio receiver?


Double-check your connections (my bet is that something is wired wrong). Make sure C3 is near the coil and the entire three JFET circuit is compact. Check the gate-to-source voltages, the sources should be a couple of volts more positive than the gates.

The ON Semiconductor datasheet for the MPF102 does not specify a pinch-off voltage, but the Fairchild datasheet says the pinchoff voltage can be as high as 8 volts. How do you get that with a 3 volt power supply? The fets don't run pinched off, but with a gate-source voltage somewhere between 0V and pinchoff. I suspect that one or more of your JFETs need more voltage.

Once you are sure everything is connected correctly, you can try a higher battery voltage. If that doesn't do the trick, you might be able to select some JFETs with lower gate voltages, or you can have a look around Digikey for low pinch-off JFET (pinchoff specs below 1.5 volts are rare but they can be found).

 

Thank you for your reply Dick.A non oscillation constituted exactly that.No meter reading at all. I used another circuit again,based around a single Mpf102,without the weird positive ground connection,leaving the meter half of the circuit as is.This seemed to work as expected,however with no buffer between the tank circuit,and the rf output / meter part of the circuit,the oscillation soon died out. I plan on experimenting with a high speed op amp buffer,or jfet buffer to solve this problem.The only thing I have been advised to watch out for,is the bandwidth of the components used.