Hi Spec,
Thank you for the insight of my psyche, I didn't know that yet, keep up the good work!
Hi not1xor1 Your last scheme looks better with R7 1-Meg and C2 4N7.
But because of the amplification in Q6 it becomes difficult to get the whole stable again.
Because of this you need C7 1N and R14 1K.
I think it is useful to know, that for every ampare output current you need about 50uF over the output, as a rule of thumb.
The more phase margin the control loops have, the smaller this value can become. (and some other things)
With this kind of "standard" circuits as in this topic, use around 50uF/Ampere for good dynamic behavior.
Even if you have sufficient phase margin, the dynamic behaviour may not be sufficient. (not fast enough)
It becomes even more difficult when you hang the i loop in the circuit,
Fault peak current protectionI always use the BC337 as in your circuit Q9 to limit the peak current.
and then with an extra opamp make second loop for an adjustable current control.
R2 and Q9 then limit the current a little above the maximum current the power supply should supply.
Q9 is much faster in responding than an opamp loop.
So in case of a short circuit Q9 protects your power trasistoren and your D.U.T. better than the current loop.
But only if you make C7-1nF and especially the previous schematic C8 of 100uF that your surge protection with Q9 will not work!
The high value of R16, 240K can make a nice noise contribution.
C1 of 10pF is too small, I take depending on the total circuit setup here 10 to 100nF.
A larger value helps to reduce the noise contribution and parasitic capacities.
But an extra resistor is needed directly at the opamp input, this to protect the input during short circuits.
Then the energy in C1 is partly dumped in the opamp, one or two diodes also help to keep the opamp intact in these error conditions.
If the opamp itself already has diodes, take into account that you usually should not exceed the 10mA peak current in or out a opamp input!
Furthermore it is good to think about bias currents, the LM358 is in this respect an old opamp so that higher bias values are unfavourable for the noise contribution and offset drift specifications.
Most power supply circuits will exhibit undesirable behaviour in the transition area between CV and CC.
And then it also depends on the amplification of the circuit (R16 and R17) and the current on which the power supply is set.
That's the hardest part of designing power supplies.
Spyce helps, but you will also have to build a test circuit to run into things Spice can't do for you.
I hope this helps.
Kind regards,
Blackdog