LM317 Based Digitally Controlled Power Supply

[Kleinstein]

The LM317 when used as the regulator is not very accurate. This starts with the output pins / contacts and the heating of the reference and amplifier part. So a serate build regulator is often the more accurate way. Adding current limiting / regulator to a LM317 regualtor circuit is challanging, arguably more tricky than starting from scratch with OPs and power transistors.

Usually the regulation is done analog - a digital solution is rather demanding on the ADC / DAC side and still often slow.
When starting with PWM as a set signal one goes PWM -> fitler to get an analog set voltage and than goes with an analog regulator (e.g. classic PID style) from there.

[MrAl]

The LM317 when used as the regulator is not very accurate. This starts with the output pins / contacts and the heating of the reference and amplifier part. So a serate build regulator is often the more accurate way. Adding current limiting / regulator to a LM317 regualtor circuit is challanging, arguably more tricky than starting from scratch with OPs and power transistors.

Usually the regulation is done analog - a digital solution is rather demanding on the ADC / DAC side and still often slow.
When starting with PWM as a set signal one goes PWM -> fitler to get an analog set voltage and than goes with an analog regulator (e.g. classic PID style) from there.

Hi,

Yes, they do not mention that directly in the data sheet so it's easy to miss.  If you look at the spec's carefully and surmise yourself that the temperature is going to rise with significant load in most cases, then you might realize that the on-chip voltage reference temperature is also going to rise and thus the output voltage will change more significanly than the voltage reference.

[Zero999]

The good thing about the LM317 is it has over-temperature and safe operating area protection, which you don't get with a plain transistor.

The circuit I posted above will should work with the LM317. Note R11 is only there to keep the simulator happy.

EDIT: Added -5V supply.

[Kleinstein]

The circuit with the LM317 as a power transistor replacement would need a negative supply for the OP-amp. Otherwise it would not work below some 1.2 V and could not limit the current with these low voltages (e.g. with a short).

[Zero999]

The circuit with the LM317 as a power transistor replacement would need a negative supply for the OP-amp. Otherwise it would not work below some 1.2 V and could not limit the current with these low voltages (e.g. with a short).

Yes, you're right. I somehow messed up by copying the wrong schematic. I even said that myself.

I'll edit it.