Prev: Advanced Circuits 4PCB
Next: MOSFET transient dissipation
From: Joerg on 2 Dec 2009 19:26
Steve wrote:
> I've put together a voltage controlled current source for use at high
> currents (5 Amps) based on the desgin at
> http://www.daycounter.com/Circuits/Current-Servo/Current-Servo.phtml
> It uses a single supply op amp (OPA340 or OPA703) driving a low gate
> threshold logic level MOSFET (RFP30N06LE). Load connects between
> drain and power source. Current sense resistor from MOSFET source to
> ground with that junction feeding back through 10 K to inverting
> input. Voltage control applied to non-inverting input.
>
> Because of the high current and the desire to minimize power
> dissipation in the sense resistor and voltage drop across it, I am
> using a 0.1 ohm sense resistor. This is resulting in the current not
> following the V=(current / 0.1 ohms) equation. I observe 0.725 A
> (expect 1 A) at 0.1 control voltage and 3.5 A (expect 5 A) at 0.5 V.
> These large errors are even present at low currents. Things are
> better for higher valued sense resistors. For example, output current
> is only 5% low when a 1 ohm current sense resistor is used. Is some
> resistance not being compensated for or ?
>

Check for oscillation first, as Jim wrote. With the cap from OUT to IN-
it shouldn't happen but depends on your load. Which opamp did you use?
Does its input CM range include the negative rail? What's the offset spec?

Also, mind that cell phone signals and such can recitify in many opamp
inputs, so if you have long wires there you might at least want to
consider a cap from IN+ to ground. Hoping it's a ground plane ...

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
From: ChrisQ on 3 Dec 2009 04:23
Steve wrote:
> I've put together a voltage controlled current source for use at high
> currents (5 Amps) based on the desgin at
> http://www.daycounter.com/Circuits/Current-Servo/Current-Servo.phtml
> It uses a single supply op amp (OPA340 or OPA703) driving a low gate
> threshold logic level MOSFET (RFP30N06LE). Load connects between
> drain and power source. Current sense resistor from MOSFET source to
> ground with that junction feeding back through 10 K to inverting
> input. Voltage control applied to non-inverting input.
>
> Because of the high current and the desire to minimize power
> dissipation in the sense resistor and voltage drop across it, I am
> using a 0.1 ohm sense resistor. This is resulting in the current not
> following the V=(current / 0.1 ohms) equation. I observe 0.725 A
> (expect 1 A) at 0.1 control voltage and 3.5 A (expect 5 A) at 0.5 V.
> These large errors are even present at low currents. Things are
> better for higher valued sense resistors. For example, output current
> is only 5% low when a 1 ohm current sense resistor is used. Is some
> resistance not being compensated for or ?
>

Schematic looks ok and it's not unstable, so have you run all the
critical voltage sensing wiring to a star point ?...

Regards,

Chris
From: Robert Baer on 3 Dec 2009 06:29
Steve wrote:
> I've put together a voltage controlled current source for use at high
> currents (5 Amps) based on the desgin at
> http://www.daycounter.com/Circuits/Current-Servo/Current-Servo.phtml
> It uses a single supply op amp (OPA340 or OPA703) driving a low gate
> threshold logic level MOSFET (RFP30N06LE). Load connects between
> drain and power source. Current sense resistor from MOSFET source to
> ground with that junction feeding back through 10 K to inverting
> input. Voltage control applied to non-inverting input.
>
> Because of the high current and the desire to minimize power
> dissipation in the sense resistor and voltage drop across it, I am
> using a 0.1 ohm sense resistor. This is resulting in the current not
> following the V=(current / 0.1 ohms) equation. I observe 0.725 A
> (expect 1 A) at 0.1 control voltage and 3.5 A (expect 5 A) at 0.5 V.
> These large errors are even present at low currents. Things are
> better for higher valued sense resistors. For example, output current
> is only 5% low when a 1 ohm current sense resistor is used. Is some
> resistance not being compensated for or ?
>
Are you using Kelvin sensing or just the grunty single wire sensing?
From: Robert Baer on 3 Dec 2009 06:31
Jim Thompson wrote:
> On Wed, 02 Dec 2009 18:17:06 -0500, Steve <steve(a)nomail.net> wrote:
>
>> I've put together a voltage controlled current source for use at high
>> currents (5 Amps) based on the desgin at
>> http://www.daycounter.com/Circuits/Current-Servo/Current-Servo.phtml
>> It uses a single supply op amp (OPA340 or OPA703) driving a low gate
>> threshold logic level MOSFET (RFP30N06LE). Load connects between
>> drain and power source. Current sense resistor from MOSFET source to
>> ground with that junction feeding back through 10 K to inverting
>> input. Voltage control applied to non-inverting input.
>>
>> Because of the high current and the desire to minimize power
>> dissipation in the sense resistor and voltage drop across it, I am
>> using a 0.1 ohm sense resistor. This is resulting in the current not
>> following the V=(current / 0.1 ohms) equation. I observe 0.725 A
>> (expect 1 A) at 0.1 control voltage and 3.5 A (expect 5 A) at 0.5 V.
>> These large errors are even present at low currents. Things are
>> better for higher valued sense resistors. For example, output current
>> is only 5% low when a 1 ohm current sense resistor is used. Is some
>> resistance not being compensated for or ?
>
> Either your 0.1 Ohm resistor is high, or more likely, trace resistance
> hasn't been accounted for.
>
> Maybe arrange PCB traces for Kelvin sensing?
>
> ...Jim Thompson
....THAT would be the first thing to do.
Convert from a grunty single wire feedback th true Kelvin sensing.
 | 
Pages: 1
Prev: Advanced Circuits 4PCB
Next: MOSFET transient dissipation