I want a voltage controlled resistor for high voltages for a project I'm working on and using a common emitter/source transistor to shunt current into ground in order to get a desired voltage is not working well in my circuits because it is generating its own current. I need a true, end to end voltage controlled variable resistor with input voltage through the top and load on the bottom.

I've heard Jfets can act as resistors but in my simulations the source voltage just seems to copy whatever the voltage is on the gate. What am I doing wrong? Can you not use a jfet as a end to end resistor?

It should be noted that I plan on using a digital feedback loop to negate any non-linearity and that I will use a SIC Jfet for the high voltages.

To clarify, I effectively need a drop in resistor replacement that I can control with voltage

  • 2
    \$\begingroup\$ "...in order to get a desired voltage" - if the goal is to get a particular voltage, why do you need a voltage controlled resistance? \$\endgroup\$ – Bruce Abbott Feb 22 '16 at 3:20
  • 1
    \$\begingroup\$ FETs (or BJTs, for that matter) do not act as resistors, i.e. Is is not going to be proportional to Vsd, although the name "transistor" is combined of the words "transfer" and "resistor". A transistor is much more like a controlled current source than a resistor. FET can be used as a voltage-controlled resistor for small-signal applications, but that is not what you need. It sounds like you're trying to do something very simple if a very roundabout way. \$\endgroup\$ – biggvsdiccvs Feb 22 '16 at 3:44
  • \$\begingroup\$ I've heard Jfets can act as resistors but in my simulations the source voltage just seems to copy whatever the voltage is on the gate. What am I doing wrong? Actually, nothing. With the load tied from source to ground, The FET is a source follower. Look it up... Questions: 1. What's your source voltage, your load resistance, and what kind of voltage variation do you need across the load? 2. Does your load have to be high-side driven or can it be low-side driven? 3. How about adding a schematic to your post so we'll all be on the same page. 4. are we talking MOSFETs or JFETs? \$\endgroup\$ – EM Fields Feb 22 '16 at 10:15
  • \$\begingroup\$ @BruceAbbott: If the goal is to drop a particular voltage across, or pass a particular current through the load, using a fixed voltage source, then a variable resistance in series with the load and the source will allow that to happen and a voltage variable resistance will allow it to be done automagically, with feedback, which the OP plans to use. \$\endgroup\$ – EM Fields Feb 22 '16 at 10:24
  • \$\begingroup\$ Sorry I should have been a little more clear, effectively what I need is a drop in resistor replacement, not necessarily a voltage control. \$\endgroup\$ – coinmaster Feb 22 '16 at 13:30

A FET (MOSFET or JFET) only approximates a fixed resistance when the drain-source voltage is relatively low (say less than a few V); when VDS is high, it approximates a constant current source. The value of the resistance or of the current depends on the gate-source voltage.


The Voltage controlled resistance characteristic only holds true at low voltages like jp314 stated .So for high voltages you would be looking at lots of devices in series ,this is not looking practical .An alternative approach is to use LDRs .The old big ones like ORP12 which is being remanufactured will take 100V or more while keeping a proper linear VI curve.These days you can use Leds to drive them .Galvanic isolation could be a real benefit if you are dealing with high V.These LDRs are slow to respond however .Another way is to get 2 resisters whose series sum represents R max and one of the resisters represents Rmin .For example 2 1K resisters with a semiconductor switch like a mosfet across one of them is a set up for Rmax =2K Rmin=1K .Now you just use PWM to drive the switch and filter out the PWM as needed.

  • \$\begingroup\$ I thought SIC Jfets were different because they run on ultra high voltages? This is the one I was looking at infineon.com/dgdl/… I looked into LDRs but they don't have a very wide range of on resistances and they are supposed to be very slow to respond. The resistor thing you talked about seems to only be of 2 selectable resistance values unless I'm misunderstanding? \$\endgroup\$ – coinmaster Feb 22 '16 at 13:32
  • \$\begingroup\$ @coinmaster .The resister thing allows any resistance to be selected between Your fixed values Rmin and Rmax.If the duty cycle in the example I gave was 50% the effective resistance would be 1K5 .Your SiC fet will hack High voltages but it still only behaves like a resister at relatively low volts. \$\endgroup\$ – Autistic Feb 22 '16 at 19:20
  • \$\begingroup\$ Hmmm, that resistor thing sounds pretty good, do you mind drawing a schematic for me? \$\endgroup\$ – coinmaster Feb 22 '16 at 22:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.