New JFET by TI - very tightly specced Vgs(off)

I have bought some Toshiba 2SK209 jfets in the past and I was amazed at how small the variance of parameters was. They have several groupings of parameters marked by colour code in the name (Green, Red, Yellow, Blue). Might be worth looking at. If I remember correctly, the Red one looked quite similar to J201.
 
I understand how the circuit works. The JFET, opamp gain stage and peak detector are inside a feedback loop which should help mitigate variations in JFET specs, at least up to a point. Additionally, the Rockman has two SIT resistors connected to the JFET gate, apparently for tuning the JFET's R vs Vgs transfer function.

According to the Rockman drawing I have (x100 rev 10), the JFET part no. is 2N4339. The SMD part (SMP4339) is available from Mouser if you're hard-over on using the same part that Scholtz used. The datasheet specs the max Rds,on at 1.7K. There is no min Rds,on spec.
 
Last edited:
Chuck D. Bones said:
According to the Rockman drawing I have (x100 rev 10), the JFET part no. is 2N4339. The SMD part (SMP4339) is available from Mouser if you're hard-over on using the same part that Scholtz used. The datasheet specs the max Rds,on at 1.7K. There is no min Rds,on spec.
Click to expand...
I had good luck with a batch of MMBFJ202s. Of course, theres a lot of parameter overlap with J201, so YMMV. The particular 2N4339s I have are all quite off.

The 5Meg-ish resistors are there to liniarize the JFET's transfer function in the ohmic region, reducing distortion: right off Siliconix AN105, page 3. I've consistently achieved lowest distortion with equal R values about the gate.
 
JTEX said:
The 5Meg-ish resistors are there to liniarize the JFET's transfer function in the ohmic region, reducing distortion: right off Siliconix AN105, page 3. I've consistently achieved lowest distortion with equal R values about the gate.
Click to expand...

That was my point. Scholtz selected those resistors based on some secret criteria and they're not always equal. We can guess that it has to do with linearity, dynamic range and maybe transient response.
 
You must log in or register to reply here.
Back
Top