Variable constant current load

[tubos]

I'm building this based on schematic i found at eevblog.See circuit below.

It all works but when I increase the potvalue until about 100mA
A high frequency component signal becomes apparent at the output
from my 2nd opamp.

Adding caps on my batterysupply didnt help.
But adding a 10 uF on the Drain of the mosfet solved the problem.

Can someone explain why this happens?

What else can I improve on this circuit?



View attachment 65953

 

[MrAl]

Hi,

R1 can work with the input capacitance of the MOSFET to form an oscillator by introducing a phase shift. There are different ways to compensate for something like this. To understand it better you might do a root locus on the circuit (second op amp section and everything after that).

For one, R1 seems a little large. Lowering that value a lot should help. Also, a small parallel cap (parallel to R1) could help too. Slowing down the MOSFET helps the op amp keep up also.
Bipolar transistors in these apps usually are more stable.

In any case, it has to be tested over the expected load types that it will be working with.

 

[KeepItSimpleStupid]

That's not a constant current source. Far from it. The "constant" is always changing because it depends on the battery voltage. You need to use a reference IC.

The TI datasheet kinda sucks. https://www.electro-tech-online.com/custompdfs/2012/07/lm324.pdf They said it's good for 0V input. That almost never works. I suspect at least a few mV is needed. The OfFset voltage is going to play a part in the actual current value.

It does depend on how critical the current source is.

But, it needs a reference. Whether it be an IC or a buffered Zener diode.

The first OP amp needs to be driven by a LOW Z source and it needs to be stable. Vos is critical in this circuit. Vos is a temperature dependent parameter. The circuit will probably have trouble at VERY LOW currents and you may need an RC filter somewhere. The 1N4148 might be better if it was replaced by a Schotkey diode.

 

[MrAl]

Hi,

Sometimes they draw a battery instead of a power supply. Yes a reference diode would help any of these circuits.

Also, the inputs can go all the way to zero volts because the inputs are PNP bases.

Also, the input of the first op amp should be high enough impedance so it doesnt have to be driven with a low Z source.

This circuit obviously is not made for low output currents.