I saw this PIC development board, which has the feature of getting two outputs from the LM317. Interesting, I though. But in the LM317 datasheet the adjustment pin current is a maximum of 100uA, which wouldn't be enough to power anything. So is this design legitimate? I suppose the 3.3V output could be taking current from the actual 5V output? But wouldn't any change in current draw from the 3.3V output change the voltage drop across R4? Also, unless my math is wrong, the voltage at the adjustment pin has to be 7.2V in order for the R5/(R3+R5) voltage divider to output 3.3V.
Using the resistors R3 R4 and R5 with the values shown, gives a nice 5.0v on the outout of the LM317.
But the so called 3.3v point according to my calcs has about 1.72v.
My first thought is that the 3.3v connection is a really bad idea.
My second thought is that who ever drew the schematic was having a bad day and his brain was away with the fairies.
My third though is that it is a deliberate trap to catch out plagiarists, and the real equipment was not actually built like that.
Having thought about this for a while, I was going to comment that the schematic was awful because it did not explain the purpose of the 3.3v jumper, and that this was a prime example of bad documentation.
But wait...
I did a search of the Olimex website and found the development board in question.
There I found an isolated copy of the schematic, and a reasonably comprehensive user manual which incorporates a copy of the schematic.
I just use a couple of 100mA regulators, one 5V, one 3.3V, and select the output of either with a jumper across two of three pins - so the centre pin is out, one outer is 5V and the other outer is 3.3V.
I just use a couple of 100mA regulators, one 5V, one 3.3V, and select the output of either with a jumper across two of three pins - so the centre pin is out, one outer is 5V and the other outer is 3.3V.