# JFET current source drain current not constant

#### ray242

Joined Nov 27, 2016
44
Hi, I learned about JFET current source and just built one on my bread board using J109. It's the simplest form with source and gate shorted together. According to my understanding of JFET behavior, when I change the drain resistor from 500 ohm to 300 ohm the current should not change but I got different result when I measured it. The current is 9V(my power supply) divided by the resistance. Is their anything wrong with my understanding?

Thanks very much !

#### hp1729

Joined Nov 23, 2015
2,304
Hi, I learned about JFET current source and just built one on my bread board using J109. It's the simplest form with source and gate shorted together. According to my understanding of JFET behavior, when I change the drain resistor from 500 ohm to 300 ohm the current should not change but I got different result when I measured it. The current is 9V(my power supply) divided by the resistance. Is their anything wrong with my understanding?

Thanks very much !
Without a circuit example can only give a general answer. The "constant current" circuit does not cause a specific current to flow, just a limited current. You can draw less than that limit.

#### dannyf

Joined Sep 13, 2015
2,197
Is their anything wrong with my understanding?
you could have connected it wrong, you could have used a defective device, you could have measured it wrong, ....

#### OBW0549

Joined Mar 2, 2015
3,566
If you'll look at the data sheet for a J109 FET you'll see that its Idss is 40 mA minimum; that is the drain current it will attempt to regulate at with the gate shorted to the source. However, to get 40 mA through a 300Ω resistor you'd need 12 volts across the resistor plus at least another volt or two across the FET to put it into constant-current mode, or at least 13-14 volts total. Your supply is only 9 volts, so the FET isn't going to be doing any current regulating and the current will be determined entirely by the resistor.

#### dl324

Joined Mar 30, 2015
16,160
Is their anything wrong with my understanding?
+1 for what @OBW0549 said.

What voltage drop are you measuring on the resistors and what do you expect? Constant current sources can't function if the operating voltage isn't sufficiently high.

#### crutschow

Joined Mar 14, 2008
33,358
And the current won't be exactly constant even when in the constant current mode.
It will vary some with a change in voltage across it.

#### DickCappels

Joined Aug 21, 2008
10,104

#### Attachments

• 6 KB Views: 2

#### ray242

Joined Nov 27, 2016
44
If you'll look at the data sheet for a J109 FET you'll see that its Idss is 40 mA minimum; that is the drain current it will attempt to regulate at with the gate shorted to the source. However, to get 40 mA through a 300Ω resistor you'd need 12 volts across the resistor plus at least another volt or two across the FET to put it into constant-current mode, or at least 13-14 volts total. Your supply is only 9 volts, so the FET isn't going to be doing any current regulating and the current will be determined entirely by the resistor.
When I connect the drain directly to the power supply I measured 220 mA current rather than 40 mA, is it due to the manufacturing spread or is it a wrong way of measuring Idss?

In the data sheet fig.3 and fig.6 have the same parameters except for the Vgs cut off voltage, which one should I use ?

Thank you!

#### OBW0549

Joined Mar 2, 2015
3,566
When I connect the drain directly to the power supply I measured 220 mA current rather than 40 mA, is it due to the manufacturing spread or is it a wrong way of measuring Idss?
My guess would be manufacturing spread. The J109 is designed for switching applications where low Rds(on) is important, so they don't specify any maximum Idss. 220 mA seems entirely plausible.

In the data sheet fig.3 and fig.6 have the same parameters except for the Vgs cut off voltage, which one should I use ?
From the Idss value you measured, looking at Fig. 4 it appears your part probably has a Vgs(off) of around -4 volts so the curves in Fig. 3 would be more representative than those of Fig. 6.

Overall, I think you're seeing why people don't use FETs very often for current sources: the manufacturing spread makes them very unpredictable and thus a real PITA to use.

#### dannyf

Joined Sep 13, 2015
2,197
I measured 220 mA
at that kind of current and 300ohm resistor load, there are 3v on your jfet (assuming that the 9v source holds up), your jfet is toast.

put a resistor on the source to limit the current.