Above is a typical oscilloscope response from the current limiting in action. The voltage is set at 4V and the current limit is set at 10mA. The load is a 10ohm resistor. No adjustments have been made yet! As soon as the load is applied the voltage is brought down to 80 mV (should be 100mV). I assume that the slope is due to the discharging of the 220uF output capacitor taking 5-10mS to discharge. According to the normal discharge calculation it should take around 8mS so it seems to fit. Using a 100ohm load this discharge time is longer (See below!) – sort of proving it to be the capacitor and the load in conjunction that determines how quickly that slope will reach the correct voltage.
Here the load is 100ohm and the voltage set for 6.6V. Current limit is again set for 10mA. The sequence of events we see here is first of all the voltage from the DAC (also the input to CA3140) is brought down very quickly when the current limiter is activated. That is the cyan curve that starts out at around 800mV. The reaction is very quick and the output voltage (yellow curve) immediately begins to fall – but will be slower this time as we only have a 100ohm load to discharge the output capacitor. During the 60mS that we can see the voltage has not time to reach its intended voltage of around 100mV.
As the DPSU can only supply a voltage and has no means to force down the voltage that is kept up by the output capacitor I might include a MOSFET that can short out the output capacitor and thereby make the output voltage go down quicker! I will do some tests on this. It might be controlled in parallell with the first MOSFET via a capacitor for a very brief discharge. I will test this idea out and be back if it works well! 😉 But it might be too coarse as I have no means to limit the discharge to a proper level … I also do not want to introduce any “ringing” … Well we will see! Everything is a compromise!
At these very small voltages there is some non-linear response and I plan to handle that in two ways: By adding a constant current source at the output I will avoid the very lowest voltages. At 1mA the sense voltage is just 0.1mV over the 0.1ohm sense resistor. Having 20-40mA constant load will ease this a little bit. Every step will still be just 1mA=0.1mV but al least we have left the GND proximity. The rest will be handled in the Arduino software. For example: Having a constant current of 20mA and wanting the output current to be limited to 5mA means I will have to set it for 25mA! I hope that is clear? Although the sense voltage is amplified around 12.5 times still the voltages will be very low and the LM358 Op-amp does not work very well so close to GND.
Now it the time for more varied measurements and more testing with more varied loads: capacitors, motors, relays and naturally with more voltages!
See you soon again!
/Sfenn
Ar2uino Scribblings 