NJFETs for Source Follower Applications

AP is not the best tool to measure noise.
A Wurcer LNA and a good sound card can do better.
Or a LNA + a proper Spectrum analyser with high resolution.


Patrick

Patrick,

Measuring noise is well within the skillset of the AP2722. On my bench the AP2522 using the same software as the2722 measures to low single digits uV’s. Turn off the generator; run the FFT and autorange will zoom right in on the noise floor showing 1/f noise and the rest. You can adjust the Circuit Under Test to tailor the distortion and noise profiles to your liking. Tune the sweet spot.

Take a look at Linear Audio Volume 4, John Walton uses his 2722 plus a custom 40dB gain LNA to zoom in on regulated power supply noise.

Scott's 60dB LNA is also a fine tool with resolution headroom to spare.

Thanks DT
 
sound warmth

Feel free to use this image, it might be easier to interpret than unformatted ascii text.

_

Hi. 2sk209gr is interesting part.It has about 10dB more H2 but has less H3 than 2sk170bl.Maybe it sounds better than 2sk170 because of more H2 adds warmth to sound and H3 is lower so sound may be nicer? Anyone compared them and could tell about sound? I have reel of 3000pcs of 2sk209gr.Did not have a spare time to play with them yet but i want to try them a lot.
 

Attachments

  • 2sk209h2.png
    2sk209h2.png
    7.5 KB · Views: 588
Further to the publication of DimitriDanyuk on SMD JFET noise, I was considering how to use the BSR58 in actual applications.
Measurements Rate New SMT Low-Voltage JFETs Under Consistent Conditions - An Update Using Modern JFETs | audioXpress
No curve tracing data can be found on the net, so I bought some and did my own measurements.

This is just a random sample. So not sure how typical it is.
Also to limit dissipation, I restricted the drain current to below 12mA.

Vgs at 10mA = 0.92V
Yfs at 10mA = 12.8mS

In order to limit dissipation to realistic values, one would need a source resistor of 100R or more.
And the output impedance of a source follower would be almost 200R.

2x 2Sk209GR in parallel is for me still better choice.


Patrick

.
 

Attachments

  • BSR58 Id vs Vgs.png
    BSR58 Id vs Vgs.png
    16.8 KB · Views: 314
While there are continuous discussions about JFET noise and what can we use to replace K170 etc.
It is perhaps useful to post a table published by Scott some time ago.

And the surprise winner is not the popular K170 or BF862.


Patrick

.
 

Attachments

  • Wurcer JFET Noise Table.jpg
    Wurcer JFET Noise Table.jpg
    82.2 KB · Views: 481
  • Like
Reactions: 1 user
I have always wondered how the much-talked-about J113 performs in terms of noise.
Just found this wonderful website which has all the information one likes to have :
https://www.mvaudiolabs.com/diy/modern-jfet-noise-measurements/
I leave you to draw your own conclusions.
For me, no surprises with 2SK170 and 2SK117 ( = 2SK209 SMD ).
So I reafirms what I woulld use for source followers / buffers in applications like active crossovers.
And the J175 / 176 are actually better than I thought.


Cheers,
Patrick
 

Attachments

  • noise_J113.png
    noise_J113.png
    23.6 KB · Views: 274
  • noise_2SK117.png
    noise_2SK117.png
    12.2 KB · Views: 263
  • noise_J175.png
    noise_J175.png
    26.5 KB · Views: 241
  • noise_J176.png
    noise_J176.png
    23.4 KB · Views: 254