0
\$\begingroup\$

I am trying to do a high bandwith and high imput impedance analog front end for a mini oscilloscope project. I am using a JFET for this purpose. I accounter a weird behaviour and i don't understand it.

When i applied 50Hz, 2Vpp square wave to the input of the source follower when the gate resistor is 4M ohm I get results below both in simulation and on the circuit.

enter image description here

But when I remove the gate resistor low frequency response gets very nice both in simulation and on circuit. I belive this is not proper usage of JFETs but I don't understant the behaviour. Do you know the cause for this?

When diodes revomed results are still same.

enter image description here

\$\endgroup\$

1 Answer 1

0
\$\begingroup\$

With no resistor, there is nothing to set the DC voltage on the gate. It is still AC coupled to the input.

A JFET's gate is actually a pn junction, so leakage currents (<< nA) will eventually cause the gate to charge up, but this could take many minutes and will depend on the leakage on your capacitor and/or the other diodes you have connected. You will need a resistor to set the gate DC bias voltage.

\$\endgroup\$
5
  • \$\begingroup\$ if I periodically clear the charged voltage from gate pin after each sampling, with a relay or analog multiplexer is it usable or will there be other problems? \$\endgroup\$ May 7, 2022 at 7:56
  • \$\begingroup\$ Yes, that would work, but might give unexpected outputs on your waveform. You should also use a larger capacitor (10's of uF) to give a longer time constant. Why are you AC-coupling the input anyway ? \$\endgroup\$
    – jp314
    May 7, 2022 at 16:14
  • \$\begingroup\$ I want both ac and dc coupling capability so there is another path with an opamp which does dc coupling. But when i feed gate with dc coupling i couldnt solve the oscillation and setling time problems. \$\endgroup\$ May 8, 2022 at 0:05
  • \$\begingroup\$ so just use DC coupling and eliminate the capacitor. \$\endgroup\$
    – jp314
    May 8, 2022 at 0:09
  • \$\begingroup\$ main opamps input bias current is high so i can not get 1Mohm input impedance with it. \$\endgroup\$ May 8, 2022 at 0:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.