+ AngraMelo
The LM723 is a great chip and still going strong after all these years.
By the way, neat schematic: nice and easy to follow.
Here are a few points about the PSU design and the early current limiting fault:
According to your schematic, the PSU should limit at 0.65/(0.1/4) = 26 amps. Is that your intention?
There is an issue with the four current sensing resistors. Sufficient to say at this stage that connecting four low value, high current resistors in parallel is problematic, but that is unlikely to be the cause of the early current limiting problem.
There are no base emitter resistors on any of the five transistors this means that they increase current fast but decrease current slow. It would be wise to connect a 560R across the E/B of Q1 and 220R across the E/B of Q2 to Q5.
PSUs, of all types are notorious for oscillating and the other members have touched on this point. As you probably know, there are two control loops on an LM723 PSU; voltage and current. I would suggest, at this point of the investigation, to remove both connections to the current limit transistor so that any issues with the current limit loop are eliminated.
The next thing to do is to disconnect the LM723 and transistors from the 28V raw supply line. Then draw 10A from the raw 28V supply line and monitor the voltage directly across the reservoir capacitors with a scope. Then if you could post the exact absolute voltage levels at the peaks and troughs of the ripple, that would assist in the investigation. Is it correct that the mains supply in your local is 220V RMS, 60Hz?
You have a reservoir capacitance of 2 x 22mF = 44mF. Can you state what tolerance the reservoir capacitors are? -40% +100% would be typical. If you have the type number for the reservoir capacitors that would be ideal. Can you also post the bridge rectifier part number?
You say that under maximum PSU output current (7A) the base current into Q1 is 150mA, but I calculate that for 10A PSU output Q1 base current should not exceed 25mA. The gross over current is odd. Two reasons for this fault condition spring to mind: either the circuit is oscillating, or the five output transistors have much reduced current gain.
As you most probably know, the physical layout of power circuits is critical to maximize frequency stability and accuracy. In the case of your power supply, star point techniques are essential. And on a similar theme, adequate decoupling is also essential, as has already been pointed out.
One last question: is it true that this PSU has always had the problem with early current limiting?
I do have an idea what the problem may be, but I will need the requested data above to progress any further.