Constant current load Help needed

[LukeB]

Hello helpers.

I have been having problems with a simple constant current load I have been trying to make based of a video I saw on greatscott's youtube channel. I have made a version of the schematic for myself which is linked as pdf. It for some reason will not lower the current below 1.8 amps. Does anyone know why this would be.

BTW I now realise that the approach scott used was rather terrible. The mosfet ends up in that highly resistive part of its curve and heats up a lot. I will now find a better way of making a constant current load but still want to figure out what went wrong with this design.

Thanks

[LukeB]

VCC is 12v by the way and the battery being tested is a 12ish volt car battery (and others in the future).

[xavier60]

You will need to measure some voltages to find the problem. If the circuit is oscillating, nothing will make sense.
Do you have an oscilloscope?
Or try some various value resistors in series with the Gate, close to the Gate.
It might also help to slow down the op-amp with a small capacitor between the output and inverting input. A resistor will need to be put in the feedback path.
The MOSFET's dissipation power will always simply be D-S voltage multiplied by current. 

[LukeB]

Yeah I do have an oscilloscope. Without the capacitors it will oscillate but once they are in there there are no oscillating voltages that I can find.

[LukeB]

Why does there need to be a resistor in the feedback path?

[xavier60]

A resistor would have been needed in the feedback path if a compensation capacitor were to the added to the op-amp to slow down its response speed. There is no need to do this if the circuit is really  stable.  You should check all parts of the circuit with your CRO for oscillations.
The best way to find the problem is by first understanding how the op-amp compares the Set Voltage to the Feedback voltage and varies its output voltage to a value that cause the MOSFET to pass a current that causes a voltage drop across R1 that exactly matches the Set Voltage.

[xavier60]

The problem is likely being caused by the xx741 type op-amps having an input common mode range that doesn't go down to the negative rail voltage.
Do you have another type of op-amp to try?

[The Soulman]

How much current do you need?
For the same purpose  I've build a dummy load with LM317's, actually one as voltage
pre-regulator and one as current sink to share the dissipation a bit.
It's good between 6 and 20 Volt at 1 Ampere (fixed) continues.
In the same box I also have a set of  three 12 Ohm power resistors that I can wire in for additional Voltage handling (additional 12 Volt per resistor).
The original idea was to also have a 100mA sink, but never included it and also never needed it.

Obviously it is possible to make the current fully adjustable with a LM317, and depending on your requirement you may get away with just one and a few passive components, much simpler than what you are attempting now.

[LukeB]

I am hoping to really only get like 1 or 2 amps out of it. Just for testing battery capacities. I am interested in what you said about the 741 series. What does it mean that it cant go down to the negative rail. Is this the reason why I am getting a voltage that I cant go below?? Is it like a minimum output voltage?? As you can tell I have a lot to learn. I have a lm311 but I can probably get another opamp instead. The LM317s seem good like you guys have said. What makes them better than my one??

[LukeB]

Sorry just realised that the lm317 isnt an opamp but i still like that idea. Was thinking of a different opamp (LM358).

[xavier60]

Sorry just realised that the lm317 isnt an opamp but i still like that idea. Was thinking of a different opamp (LM358).

The LM311 is a comparator.
The data sheet for the LM358 op-amp states "• Input Common-Mode Voltage Range Includes Ground"
You can learn things from data sheets.
It is a dual op-amp, so you will need to choose one half and use the appropriate pins.
With the different op-amp, the oscillation problem could come back, so check with your oscilloscope again.
The circuit really should have compensation added to slow it down and improve its stability.

[MiDi]

A picture of the circuit board could help...

What kind of current sense resistor do you use?
It could lead to oscillation when you use a wirewound one - the inductance of it should be kept low.

You should throw out C2, it shorts the battery when connecting...

A resistance between OP out and Gate should be introduced, op amps are quite sensitive to capacitive loads (it drives the effective gate capacitance Ciss) - try a 1k.


Make shure to set it to <=0A when you apply the battery, else the mosfet is fully on and shorts the battery when connecting.
Alternative is to apply battery and then power the circuit.

[MarkF]

You may like what I did.
A LM358 would work here as well.

   

[xavier60]

You may like what I did.
A LM358 would work here as well.

   

That's a good example of op-amp compensation and MOSFET Gate resistor.
The component value are not critical.

[LukeB]

There's a picture of the PCB. All of this advice is really helpful. I will try to add in the mosfet gate resistor and see if that makes the diff.

[LukeB]

I put in a resistor there and it did nothing. The voltage across the resistor remains at 1.8v even if the set voltage is at 0v.

[LukeB]

Ohh I think I am beginning to understand. If the common mode voltage range is 13v ish then that means that an input of within 2v of either end of the supply voltage will have funny results. Is that correct or am I just confusing myself?

That could explain why I have problems trying to get the voltage to go to 0v. It hovers at 1.8 no matter what the inputs. That 1.8 seems to correspond with the 2v ish range thingo.

[xavier60]

Ohh I think I am beginning to understand. If the common mode voltage range is 13v ish then that means that an input of within 2v of either end of the supply voltage will have funny results. Is that correct or am I just confusing myself?

That could explain why I have problems trying to get the voltage to go to 0v. It hovers at 1.8 no matter what the inputs. That 1.8 seems to correspond with the 2v ish range thingo.

That's right. Now you need to find an op-amp that has a common mode range that includes the negative rail like the LM358, but in a single channel type if you want to use the same PCB layout. You could wire in an LM358 to prove that it's a fix for the range problem.
Someone here might already know of a suitable op-amp.

[LukeB]

Oh ok thats all good, I will just breadboard a new one. Probably should have done that last time hehe. Oh well it was only a home etched one anyway so pretty cheap.