Hello Hackaday!

I am trying to get adjustable current limiting via DAC setpoints working, but am coming up against some problems: while the limiting does work, it is very imprecise, non linear / non deterministic (the same analog input value does not guarantee the same output), and tends to drift (+ / - tens of mA).

For those who have done similar... is this just a limitation of the LM317? Am I expecting too much with my approach here? Or am I completely missing something obvious?

Any thoughts / suggestions / comments would be appreciated.

My latest schematic can be seen below:

As you can see, the current limiting is done after the first regulator, and consists of a shunt resistor (I am using 1ohm). The DAC value is added to the voltage drop from the shunt resistor, and applied against the Vadj pin. This design does work (for large changes to the set point), but fine changes seem to be lost in the other inaccuracies.

Thanks for any comments!

Cheers