2

The circuit shown below (with actual measured values) simulates correctly in as much that it provides some gain.

When I have built the circuit connected it to a signal generator using a 1mVp-p zero dc 1KHz sine input it does not work or at least on the oscilloscope there is no output signal.

So I was wondering if it needed a bias after the 39.7p capacitor. But then would this not spoil the JFET high impedandance?

What is the minimum components necessary to be added to get this to work in reality rather than just in simulation?

preamplier

onepound
  • 185
  • 8
  • 2
    Can't have both. Either a very large (Gohm resistors are available) from gate to GND, or a direct input connection. The coupling capacitor seems to imply a highpass response is desired, but that only makes sense with a resistor. Lack of gate bias also makes the source bias network redundant. – Tim Williams May 22 '22 at 18:48
  • "Lack of gate bias also makes the source bias network redundant." Well.... That part has considerably current variation from part to part at \$V_{GS} = 0\$. So once the OP gets a bias resistor to ground, the network on the source is still a good idea. – TimWescott May 22 '22 at 18:53
  • 1
    If this is a capacitor microphone, then C1's value is representative, and the 10Mohm suggested in an answer is hopelessly low. A typical value is 1Gohm (I resorted to slivers of CrO2 video tape when I couldn't find any resistors > 100M but that was a long time ago!) –  May 22 '22 at 19:40
  • @user_1818839 1000M Ohm are available but not cheap relatively speaking. – onepound May 22 '22 at 20:24
  • @onepound I promise you : "not cheap" is an improvement over "roll your own using videotape, razorblade, superglue and silver-loaded paint". –  May 23 '22 at 12:46

2 Answers2

5

You need a resistor from gate to GND to setup the dc operating condition of this FET. A typical value is 10 MOhm.

Jens
  • 5,598
  • 2
  • 7
  • 28
  • I'll solder that in and report back tomorrow evening. – onepound May 22 '22 at 20:13
  • 1
    Larger resistor values produce more noise. You may optimize this if you calculate the minimum resistor for your input frequency. DC setpoint won't be affected by this. – Jens May 22 '22 at 21:03
  • @Jens That's a common (potential) misunderstanding. Larger resistors produce more noise ... unless they are in parallel with a relatively low impedance https://electronics.stackexchange.com/questions/264209/how-to-avoid-johnson-noise-in-high-input-impedance-amplifier/264213#264213 Now we don't know the OP's context. It *looks like* a capacitor mic - in which case C1,R1,V2 dominate except at LF, and the bias resistor (fka grid leak) determines the LF rolloff but not the wideband noise. But it may be something else altogether - in which case your warning might very well apply. –  May 23 '22 at 11:46
  • @user_1818839: yes, perfect, thank you for the hint – Jens May 23 '22 at 12:16
3

Here is a simple vacuum tube replacement amplifier

File source shown. enter image description here

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182