LM317 - how it works ? strange REF feedback.

[boo9]

I have a broken PWM control board for inductive loads,  24A, input 10-40V DC (not stabilized)

There is a LM317 to supply DC to IC (IC markings are erased but I suspect SG2524) about which I have a question.

I dont understand LM317 configuration here, if someone please explain that to me.

How does the feedback voltage from OUT gets to REF pin of the LM317, there is a resistor divider, OK I follow that,  but then there is C2 cap that follows, separating the resistor divider from the rest of the cirucit, Can feedback loop work correctly with C2 placed there ? There is no current flowing from OUT throu resistor deviders, only during charge/discharge. Assuming C2 is charged to some V, can LM317 operate correctly and stabilize the V to the board ?

From what I manged to probe, the bottom connector of C2 goes to 12-13 pin of the IC (right row, two middle pins), the SG2524 has collector of two output transistors there.

cheers

[Zero999]

A capacitor is often connected between Vref and 0V to improve the ripple rejection. There's an example on the datasheet.

[boo9]

The C2 cap is positively not on REF/Ground pins on this board.

[Circlotron]

Maybe it is a design mistake that nobody picked up because the board sort of worked - until now - and this is the cause of the failure?

[boo9]

@Circlotron - you are correct, I just saw same design different board and there is a ground path from C2/R47k point.