[SOLVED] Current Source with Op-Amp

[albus]

So I'm trying to simulate a basic current source using an opamp and a mosfet. I'm using a powermos in order to have 300 mA current. You can see the schematics below.




My question here is the inverting input of the op-amp should be the same with the reference voltage however this is not the case in the simulation. Am I missing some basic knowledge here? I wasn't sure if this was a design problem or software problem so I opened the thread here.

 

[FvM]

I suggest to try with +10V OP supply voltage.

With the current values, the OP simply isn't able to turn the MOSFET on. Or use an OP with output rail-to-rail capability.

 

[albus]

Thank you, that did the trick. I didn't quite understand why Mosfet wasn't in saturation region though because Vgs was still a lot bigger than threshold voltage. My goal is though still to operate with 5v voltage supplies.

 

[FvM]

MOSFET datasheets clarify that threshold voltage is specified for rather low Id current < 1 mA. There should be MOSFETs with lower threshold voltage and OPs with higher positive output voltage. A pull-up resistor might be sufficient to achieve it with the presently used OP.

 

[albus]

I've searched for op-amps with rail to rail capabilities and power mosfets with less threshold voltages and I believe I found them. Thank you again, it works exactly the way I want now.

 

[crutschow]

A pull-up resistor on the output of a non-rail-to-rail op amp will generally allow the output voltage to rise to the rail (if the load is a high impedance such as a MOSFET gate).

- - - Updated - - -

.................. I didn't quite understand why Mosfet wasn't in saturation region though because Vgs was still a lot bigger than threshold voltage. My goal is though still to operate with 5v voltage supplies.

From the data sheet the maximum threshold voltage for the IRF222 is 4V so the 3.144V you show in simulation is not enough to turn on the transistor at all for a worst-case tolerance device.

 

[albus]

I've understood that later I haven't been thinking about the worst case scenario.

I've got another question now.This circuit should have a big output impedance however I believe I'm doing some mistake while trying to measure it. I know how to measure a load with one pin connected to the ground, but in my case the load is floating. How should I measure the output impedance ? When I've connected an AC voltage source to the output and measured it, the output resistance turned out to be 5 ohms which I don't think is the correct answer. When I used a coupling capacitor to connect the voltage source and removed the floating load all I've seen was the impedance of the capacitor. Any thoughts?

 

[albus]

Does nobody have an idea?

 

[FvM]

The problem description in post #7 isn't clear. I don't understand how you want to measure the current source impedance without a load resistor respectively a bias current.

Varying the output voltage and observing the current change would be a straightforward way.

 

[albus]

R1 resistor in my circuit is the load resistor. It doesn't matter if it is 0 or 400 ohms, the bias current over the mosfet doesn't change.

Drain node is the output node isn't it ? I'm trying to measure the output impedance simply by adding an AC voltage to that node.



Edit: Added the configuration to measure the output impedance.

 

[FvM]

I'm trying to measure the output impedance simply by adding an AC voltage to that node.

How, exactly? Impedance is V/I, so which current are you observing? To determine the output impedance, you'd need to measure the transistor AC current or the AC voltage across the low side shunt.

 

[albus]

On the last configuration at the right hand side, there is an ac voltage source. I'm measuring the voltage at the drain node and dividing it by the current that comes out of the AC voltage source. Does it make any sense?

 

[FvM]

On the last configuration at the right hand side, there is an ac voltage source.

Obviously. I was referring to this schematic in my latest post.

This way, you are mainly measuring Rload. It must be increased to infinity to see the actual output resistance, or you need to substract admittance numbers.

It would be easier and possibly more accurate to calulate the expectable Rout. It's a function of transistor transconductance, output impedance, OP gain and shunt resistor value.

 

[albus]

Sadly, I can not increase Rload to infinity, because then the circuit won't act as if it was a current source due to mosfet going into triode region. I did calculate the expectable Rout, however my assignment was to simulate and find it in pspice. Thank you for the help, i will try to find some other way to get around this problem.

 

[FvM]

It's trivial to get the output impedance in a simulation, because you can probe any branch current of your choice.

I thought you are talking about real measurement. I mentioned options to measure the impedance, however.

 

[crutschow]

To measure the DC output impedance simply change the load resistance (say from maximum, where the circuit still is operating, to zero ohms) and measure the load voltage change and current change (which should be small) for those two loads. The output impedance is the voltage change over the current change.

 

[JZJIANG]

The polarity of amplifier is wrong. Try to swap inverting and non-inverting inputs of your opamp. The gate of the power transistor must follow the change of its source to form a negative feedback, however, in your design, the power transistor and the opamp form a positive feedback, that's why the negative input goes to 0V (i.e. -VCC).

 

[FvM]

The polarity of amplifier is wrong. Try to swap inverting and non-inverting inputs of your opamp.

That's nonsense.

 

[JZJIANG]

Why? I think the circuit has stability issue.

 

[FvM]

I wonder if you are talking about a different circuit. I'm referring to the circuit in post #1 and #10. It has overall negative loop gain, which is the necessary condition for stability.