20W solid state variable resistance / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

20W solid state variable resistance

1 2 next

Move Thread

LAN_403

Avi

Sat Dec 03 2011, 06:04AM

Registered Member #580 Joined: Mon Mar 12 2007, 03:17PM
Location: Melbourne, Australia
Posts: 410

Looking for a relatively simple technique to make a solid state 20w adjustable resistance. The resistance of the device must be known at all times and not change when the current changes. Any ideas on how to do this?
If i use a mosfet or a BJT with a fixed gate voltage/base current, the main current conduction resistance seems to change depending on the current draw, so that method is out.
Note, there may be external resistances (of unknown value and possibly changing) within the circuit so determining the resistance by current alone will not work either.
I have thought about using 8 resistors of values of 1x 2x 4x 8x etc, and a relay across each one. but i am having trouble finding 8 power resistors of doubling value (i only need to about 200 ohms or so)
Note the supply voltage will only be 24v or less, 24v across 200 ohms is 2.9W so 20W is not required for all resistances but it may make design simpler

Pinky's Brain

Sat Dec 03 2011, 10:42AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

A sense resistor in series with the MOSFET and an opamp circuit which keeps the voltage across the MOSFET AND the resistor equal to some multiple of the voltage across the resistor? (Plus some filtering to make sure it doesn't become an oscillator.) The amplification of the opamp circuit determines the effective resistance of the whole.

Dr. Dark Current

Sat Dec 03 2011, 11:10AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

An op amp with its inverting input connected to a shunt in the source of the output MOSFET, and its non-inverting input connected to a resistor divider from drain. The output of the op amp goes to the gate, of course. Can use single supply if you use an op amp which has its common input voltage range from ground (eg. LM358)

Steve Conner

Sat Dec 03 2011, 11:14AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

What Pinky's brain said. (edit-and dr kV) I've seen commercial "electronic loads" intended for testing power supplies, that offer constant voltage, constant current or constant resistance modes. The constant resistance mode works as described.

One issue is that the finite response time of the feedback circuit will make the "resistor" look like it has a little series inductance. And the MOSFET has a bit of capacitance too.

You can do the binary thing with 15 1 ohm resistors, just wire a single one, a string of 2, a string of 4 and a string of 8. If you want to continue, the next building block is a 15ohm with a 1 in series.

Pinky's Brain

Sat Dec 03 2011, 11:14AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Dr. Kilovolt : You're right ... I wasn't thinking straight, I was thinking the circuit would become slightly nonlinear if you just used a multiplication of the Rsense voltage for Vds.

So what you get is :
Vsystem = A*I*Rsense + I*Rsense => Rsystem = (A+1)*Rsense.

PS. still have to be careful about oscillation.

Steve Conner

Sat Dec 03 2011, 11:21AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Yep, you need a balance arrangement where the shunt voltage (negative) and the drain voltage (positive) are combined at one op-amp input. I guess the non-inverting one.

Pinky's Brain

Sat Dec 03 2011, 11:57AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

PS. if the discussion seems disjointed ... it's because of my unfortunate habit of editing straight after I post.

Dr. Dark Current

Sat Dec 03 2011, 11:58AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

Pinky's Brain: The circuit I described takes the total load voltage (transistor + shunt) and total load current (which is of course the same for the transistor and the shunt), so the V/I of the load should be constant.

Avi

Sun Dec 04 2011, 05:36PM

Registered Member #580 Joined: Mon Mar 12 2007, 03:17PM
Location: Melbourne, Australia
Posts: 410

Pinky's Brain wrote ...

A sense resistor in series with the MOSFET and an opamp circuit which keeps the voltage across the MOSFET

what? that sounds like a constant current regulator...

Dr. Kilovolt wrote ...

An op amp with its inverting input connected to a shunt in the source of the output MOSFET, and its non-inverting input connected to a resistor divider from drain. The output of the op amp goes to the gate, of course. Can use single supply if you use an op amp which has its common input voltage range from ground (eg. LM358)

do you mean like this?
I would like digital input or analog voltage (3.3v max is possible without too much mess, 12v max with an op amp) input to control the resistance if possible

Steve McConner wrote ...

One issue is that the finite response time of the feedback circuit will make the "resistor" look like it has a little series inductance. And the MOSFET has a bit of capacitance too.

Speed and response time wont be an issue, doesn't need to be fast in my application.

Pinky's Brain

Sun Dec 04 2011, 06:07PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

When he says speed and response time he doesn't so much mean how fast it responds to changes in resistance ... he means in how fast it responds to changes in current, ie. how much like an ideal resistance it behaves.

For digital control you can use a digital potentiometer to handle the divider on the opamp's positive input ... that will also limit your bandwidth a bit though.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob