Another thought; would it be a good idea to use a differential amplifier with a gain of let's say 10, and divide the output down with a pot before it is fed into the CL pin, and adjust the current limit that way instead of adjusting the gain?

That seems to work well in my simulation. Just keep in mind that it's just a simulation.

 

After looking at Ron H's schematic i have been playing around with different resistor and capacitor values around the differential amp, the attached schematic seems to work really well (10K and 22pF).

However when i change to 100K resistors or 100pF capacitors it oscillates like crazy again. Are the caps somehow "matched" to the oscillating frequency, or is there some ratio between the resistors and caps?
Unfortunately 22pF and 100pF are the smallest caps i have at home right now so i cant test any other values.

 

After looking at Ron H's schematic i have been playing around with different resistor and capacitor values around the differential amp, the attached schematic seems to work really well (10K and 22pF).

However when i change to 100K resistors or 100pF capacitors it oscillates like crazy again. Are the caps somehow "matched" to the oscillating frequency, or is there some ratio between the resistors and caps?
Unfortunately 22pF and 100pF are the smallest caps i have at home right now so i cant test any other values.

My circuit is a very simplistic model of an LM723. The model posted by t06afre may be more realistic, but if I get around to running a sim with it, it will be later today.
Resistors and capacitors fomr time constants. The time constant changes as the values change. In feedback circuits, time constants change the frequency response. Some combinations may lead to instability, or better stability. If the entire circuit is accurately modeled, then, theoretically, the stability can be predicted and controlled. In this circuit, we don't know enough about the guts of the 723 to be able to fully analyze the circuit. The accuracy of the TLC272 model is also questionable, even though it came from the manufacturer.
The bottom line is there is no substitute for hardware testing. Simulations are OK only up to a point.

 

Some thing to consider. The TLC2272 has +/- 8 volt as max supply voltage. And I also do not know how well the TLC2272 work as single supply amplifier. And it also (like 99.999% of all opamps) do not like voltage input higher then the supply voltage. So here is something that might not be good at all. What is your supply voltage to the TLC2272 and in which range is your output from the lm723

 

I'm currently powering both the opamp and the regulator from +15v so it's pretty close to the maximum. The LM723 seems to go up to 12 volt with this configuration.

This is actually something i didn't think about since i have only been testing at 15V, i will need another opamp if i'm to use a 30V input for the regulator as i intended. I'll have to look around if i can find a suitable one.

 

I'm currently powering both the opamp and the regulator from +15v so it's pretty close to the maximum. The LM723 seems to go up to 12 volt with this configuration.

This is actually something i didn't think about since i have only been testing at 15V, i will need another opamp if i'm to use a 30V input for the regulator as i intended. I'll have to look around if i can find a suitable one.

LM358 or LM324 looks good in simulation. You may have to fiddle with the compensation cap values.

 

Here is some simulation results. This will give you an idea on what is going on during current limiting. Of course I am in the hands of the simulator here. To save some time I used pseudo components in the current sense part of the schematic. The simulation shows going in and out of current limiting. By changing the gain factor from the differential measurement of the current sense resistor

 

Wow, neat simulation interesting stuff.
I haven't had time to fiddle with this for a couple of days but tomorrow i will hopefully get back into it and see if i can get any results with the LM324.
Thank you for taking the time!

 

After looking at Ron H's schematic i have been playing around with different resistor and capacitor values around the differential amp, the attached schematic seems to work really well (10K and 22pF).

However when i change to 100K resistors or 100pF capacitors it oscillates like crazy again. Are the caps somehow "matched" to the oscillating frequency, or is there some ratio between the resistors and caps?
Unfortunately 22pF and 100pF are the smallest caps i have at home right now so i cant test any other values.

As Ron H noted resistors and capacitors have a time-constant, which is simply the product of R and C. RC circuits form networks that have a change is impedance with frequency. The frequency at which the impedance changes is determined by the time-constant. Thus 10k and 22pF have a time-constant of .22μs. To get the same time-constant with a 100k resistor would require a capacitor of 2.2pF (a very small value for a discrete capacitor)

 

Thanks for the info, that is surely something i have to read up on and try to better understand, at the moment i'm just testing different component values and measuring with the scope at various points to see what happens.

After a couple of hours testing various configurations today i finally got it stable with gain, i have attached a schematic of my current setup.
With the pot R32 i can adjust the current from 650mA down to about 65mA, and the waveform looks very clean, there is a little bit of spiking just as it enters and leaves current limiting when i turn the pot back and forth fast, but it seems to settle very quickly. (btw the schematic shows a sense resistor of 0.1Ω but i still used 1Ω in these tests).

The next step would be to try and see if it stays stable with a 0.1Ω sense resistor and a gain of 100 instead of 10 (to get down to the same ~65mA lowest limit).

 

Well, no it doesn't. I tried a 0.1Ω sense resistor and R31 in the schematic above is changed to 1K to give a gain of ~100.

Attached is a screen capture from the scope, ch1 shows the regulators output and ch2 shows the output of the amplifier (CL pin on 723).
When varying the current limit pot the amplitude varies for the output, but it still has these huge dips down to 0 volts all the time when it is in current limit mode.

Could this be a result of the high gain? Seeing as the output voltage drops pretty fast down to zero, but it takes a long time for the op-amp's output to come down below the 0.65 threshold...

 

@Ron H
However, my point is that in the basic app note circuit, the emitter of the current limit transistor rides on the positive output terminal--my crainial simulator cannot seem to determine if this is a common base or common emitter configuration as I can see it either way--if common base, it has low gain--if common emitter, high gain--and unlike the posted circuit, Vce varies little as it starves the base drive of the series pass transistor, thus making it all the more confusing.

Coming in late to the party, some thoughts:

I see it as a common emitter circuit, but it actually doesn't matter. Either way, the collector heads toward the emitter when the voltage across the shunt is large enough.

Also, I think you have touched on the cause of the original problem. In the original schematic the CS pin is tied to GND rather than to the output. This is a significant modification of the original circuit. It makes life a bit easier for the external opamp because now its output never has to go very far above ground, rather than operating one diode drop above the system output. This is why an LM324/358 works in this application. But it changes greatly the Vce operating point for the transistor, and we don't know how good a transistor it is. You can't assume that it is a 2N2222 and can handle 30 volts across it. Also, if the 723 output has much capacitance on it and a short is applied rapidly, the 723 output transistor base could be snapped to GND by the external opamp and the current sense transistor, driving the base-emitter junction into reverse conduction with no current limiting. I vote for connecting CS to the output, and then figuring out if the external opamp circuit needs adjustment.

ak