LT3013 SHDN does not work

There might be something coupling on your traces. Is your PCB your own design?

Hi, it does work in simulation but I have the issue in real world.
Is there any idea what could be the reason ?

Yes, it's my own design

Can you try using higher resistor values for R27 and R39? More than 100kOhm range. See if this fixes anything.

I tried but nothing has changed.

Let me try my hardware in the lab this week and I will let you know the results.

I tried the SHDN function using demonstration circuit DC813A in the lab and SHDN seems to turn the output off o.k.  I modified the circuit for 20.5VOUT according to your schematic.  I show my scope capture, schematic and setup in the images below.  Is your load pulling up the OUT of your LT3013?  Can you get demonstration circuit DC813A and try it?

Hi, I disconnected the load and tried again however nothing's changed. I checked the voltage on SHDN pin through a scope and it seems stable around 0V. Do you think any magnetic coupling and induced voltages on the SHDN pin can cause this issue ?

I was caught by this same issue with a LT3012. What I didn't realise was that, like the LT3013, the 'B version of the LT3012 does NOT have the shutdown function. Like the LT3012, Analog treats the LT3013 & LT3013B as completely different devices, with separate data sheets. You will find no mention of the LT3012B in the LT3012 data sheet, and vice versa.

This is a trap for young (& old) players.

If SHDN doesn't work, first check that you haven't loaded the 'B version of the chip.