Other Parts Discussed in Thread: LM317,
Probably something obvious that I'm not seeing.
I want to use the LM317 as a current source to drive 10mA through a set of resistors and hence set a specific voltage at each resistor junction. My goal is to set the resistor junction voltages relative to the most negative point (i.e. the bottom end of R3). If I configure the device thus:
it works fine. The top end of R3 is 1.5V more positive than the lower end and the top end of R2 is 2.75V more positive. But, if I configure the device thus:
I get very different results. The current is no longer 10mA - more like 37mA - and hence the voltages are way off.
Why is this? In both cases, the LM317 is seeing exactly the same source voltage - the LM317 doesn't "know" whether one end is a circuit ground or not.
I thought that it was a quirk of the simulator but testing a real circuit gives the same result.