Hi all. I am fairly new to electronics and am trying to use some JFETS as voltage controlled resistors, but I'm having some difficultly figuring out how the JFETs should be wired. Most pages and books say that the gate voltage of a JFET should be negative, but negative with respect to what? Do I have to use a bipolar power supply every time I want to use a JFET, or does this just mean negative with respect to the source voltage?

For instance, say I wanted to control several JFETs with a single potentiometer whose output voltage was positive. Would this ever work with an N-channel JFET? Would I just need to make the potentiometer output more negative than the source voltage? Or do I have to somehow get P-channel JFETs?

Thanks everyone!

 

hi C,
Welcome to AAC.
JFET's N or P types, can Depletion or Enhancement mode.
Check this PDF and ask if you have a query.
E

 

The gate voltage is always referenced to the source. A J-FET is a normally-on device. When the gate-source voltage is zero (gate connected to the source) the device will be on. To increase the resistance of the channel (turn if off) on an N-channel J-FET the gate has to be negative with respect to the source. At a certain negative value (the pinch off voltage) the device will be completely off. With a P-channel all the voltages are reversed.

Most J-FETS are n-channel. P-channel J-FETs exist but are rare and hard to find.

 

@eric,

Nice description, but in Figure 6, the intrinsic diode of the mosfet is potentially forward biased.



Is that a problem? Or, does the AC coupling avoid that?

 

hi John,
The Fig 6 is shown using a MOSFET type without an internal diode, is this what you are referring too.?
Eric

 

Is the question about JFETs or MOSFETs?

JFETs are depletion mode devices. MOSFETs can be enhancement or depletion mode.

 

@eric,

Nice description, but in Figure 6, the intrinsic diode of the mosfet is potentially forward biased.

View attachment 148387

Is that a problem? Or, does the AC coupling avoid that?

Did you read the title? He is talking about a JFET, not a MOSFET. JFETs do not have an internal diode.

Bob

 

The PDF in post #2, has two sections, first section jFET's, about which the TS was asking and a following section on MOSFET's
I believe John was making a general enquiry about the MOSFET internal diode.

MOSFETs are/were available without internal diodes.

E
For ref Enhancement jFET's do exist.
https://scholar.google.co.uk/schola...ved=0ahUKEwjphMKAzO7ZAhXKyqQKHV_TCyMQgQMIJTAA

 

MOSFETs are/were available without internal diodes.

I think the internal (substrate) diode is inherent in the fabrication of MOSFETs so I don't see how one could be made without it.

 

hi Carl,
I feel sure way back in the 70's there were 4 terminal FET's .?
I agree in the way MOSFETs are fabricated the diode action must exist.

E

 

@ericgibbs
Yes, that is the figure. I didn't realize it lacked an intrinsic diode. What little I have read for DC power circuits agreed with crutschow's comment. Hence, I am used to putting the source at a higher potential than the drain, pulling the gate to the source's potental to turn it off and to ground to turn it on.

@dl324 , @BobTPH
Of course, the TS's question was about jfets; however, the review posted discusses both devices and gives 4 virtually identical circuits as examples. Two have N-channel devices and two are P-channel. The P-channel mosfet circuit stuck out, and I wondered why it wasn't installed with the drain to ground and a positive voltage to control it.

 

I feel sure way back in the 70's there were 4 terminal FET's .?

Yes there were (are?) a few of those.
They bring the substrate connection out to a fourth terminal in case you want to bias the substrate at a different voltage from the source. You could do that to change the Vgs(th), for example or, I think it could be used to turn on the FET with a voltage of opposite polarity to the gate, e.g. a negative bias could turn on an N-MOSFET with the gate gournded.

The P-channel mosfet circuit stuck out, and I wondered why it wasn't installed with the drain to ground and a positive voltage to control it.

A MOSFET will conduct equally well in either direction when on, that's why they can work as variable resistors.
So the circuit has the P-MOSFET source to ground and its resistance is controlled by a variable negative voltage to ground. As long as the peak AC signal does not exceed the substrate diode's forward voltage, the circuit should work fine.
But even if you reversed the source and drain, you would still need a negative voltage to ground to control it.
In that case, the drain now acts as the source, and the source acts as the drain.
A positive voltage would only work if it were referenced to the signal level to make the top terminal act as the source, (since you want a constant Vgs to control the MOSFET's resistance).

 

Hi all,

Thanks so much for all the responses.

The gate voltage is always referenced to the source. A J-FET is a normally-on device. When the gate-source voltage is zero (gate connected to the source) the device will be on. To increase the resistance of the channel (turn if off) on an N-channel J-FET the gate has to be negative with respect to the source. At a certain negative value (the pinch off voltage) the device will be completely off. With a P-channel all the voltages are reversed.

This seems to be the closest answer to what I was asking. So if I placed a resistor after the source and made sure the voltage drop across it was greater than or equal to the gate voltage, would that make the gate voltage "negative", for all intents and purposes?

Frankly, I'm thinking enhancement mode MOSFETS might be the way to go with this anyway, after reading that pdf ericgibbs posted. Seems more intuitive at least, and the PDF says they're more common as well.

 

Hi all. I am fairly new to electronics and am trying to use some JFETS as voltage controlled resistors, but I'm having some difficultly figuring out how the JFETs should be wired. Most pages and books say that the gate voltage of a JFET should be negative, but negative with respect to what? Do I have to use a bipolar power supply every time I want to use a JFET, or does this just mean negative with respect to the source voltage?

For instance, say I wanted to control several JFETs with a single potentiometer whose output voltage was positive. Would this ever work with an N-channel JFET? Would I just need to make the potentiometer output more negative than the source voltage? Or do I have to somehow get P-channel JFETs?

Thanks everyone!

The JFET is quite similar to a tube triode, but works at lower voltage. N channel takes a negative bias on the gate to pinch off the channel.