Calculation of drain current of a JFET

[Kewell]

Hi guys,
I am struggeling to calculate the drain current of a JFET. I use the formula $$I_D = I_DSS * (1- \frac{V_GS}{V_GS,off})^2$$
In the datasheet for f.e. IF302 i get  $$I_DSS=30 mA$$ and $$V_GS,off = -0.35$$
Calculating the drain current for $$V_GS = -1.18$$ gives me 168.71 mA. But I simulated a very basic circuit in LTSpice and get for the same VGS a drain current of 44.8mA.
Can someone help me understanding this? What am I doing wrong or is my model broke?
I also did the same calculations for different I_DSS (300 mA) and VGS,off (-2V) but there I have the same problem

Thanks for any help

[T3sl4co1l]

Datasheet: https://www.interfet.com/jfet-datasheets/jfet-if3602-interfet.pdf

-0.35V is only the minimum.  -2V is the max--

Wait, they have these backwards, -2 is minimum, -0.35 is maximum.  Damned datasheet writers can't ever math.

Anyways, it could be anywhere inbetween there, and you'd have to read the SPICE model to see which value they are representing.

This is fundamentally why you don't bias JFETs with fixed voltages; you use a negative feedback or fixed-current method instead.  And the transfer function is mostly of academic interest; it's more useful when differentiated, to get transconductance (dId/dVgs) at a given operating point (Id).

Tim

[Kewell]

Thanks for your answer Tim.
Is the spread in the VGS,off due to the fabrication of the JFETs?
So every IF3602 has another VGS,off wich is in the range of -0.35V and -2V?

[Kleinstein]

Yes there is quite some spread in the VGS,off and other related parameters.  A high high VGS,off also come with a high Idss.
This are just random / charge variations in the process. If you by multiple one from the same production run, chances are that they are quite similar, but no guarantee and you still don't know if more -0.4 V or -1.5 V.
I have not tested part in question here, but some other parts (2N4391, J202). The parts from a batch were relatively similar, but still with some variations (like 0.5 V range).

[David Hess]

Thanks for your answer Tim.
Is the spread in the VGS,off due to the fabrication of the JFETs?
So every IF3602 has another VGS,off wich is in the range of -0.35V and -2V?

That is right, although there is consistency between JFETs manufactured at the same time.  Bipolar transistors have a similar variation (and consistency) in Vbe and hfe.

Matched JFETs are available if you can afford (and find) them, or you can match discrete parts yourself with a little bit of testing.  It does not take much effort to find a couple which match within millivolts.

[Zero999]

Some ranges of J-FETs are binned for different pinch-off/drain currents, take the 2N4117/8/9 for example.
https://www.interfet.com/jfet-datasheets/jfet-2n4117-2n4118-2n4119a-interfet.pdf

Wow the IF3602 is expensive. The JFE2140 is a much more reasonable price.
https://www.ti.com/lit/ds/symlink/jfe2140.pdf

[Benta]

JFETs can be problematic in that the V_GS to I_D has a huge span. For unbinned types it can be up to 1:10, for binned types 1:3 is common (eg, BF245A/B/C, 2N4117/8/9).
If you want to go lower (even more precise binning) it gets expensive.

Processing plays a big role, but also that JFETs today are a dying race (which is sad, it's a brilliant device) and no manufacturer invests money in improvement.

[mawyatt]

Processing plays a big role, but also that JFETs today are a dying race (which is sad, it's a brilliant device) and no manufacturer invests money in improvement.

Agree, sad but true, think CMOS is taking over everywhere.

I recall over 25 years ago, was an invited member of IEEE panel session with Dr. Thomas Lee from Stanford. Tom indicated that CMOS would rule just about everywhere because of digital, I countered that SiGe BiCMOS would survive because of the advantages of the SiGe bipolar transistor.

Tom was right

Best,