The layout traces need to be minimized. Another problem can be if the battery impedance is too high, a boost converter can't start up. Trouble shoot the circuit with a lab supply first.

The layout you show does not correspond to the schematic I posted above for the LED constant current design.

 

The layout traces need to be minimized. Another problem can be if the battery impedance is too high, a boost converter can't start up. Trouble shoot the circuit with a lab supply first.

The layout you show does not correspond to the schematic I posted above for the LED constant current design.

Is it safe to test it without load?
There have been changes in the plan, because there are 10+ circuits now, and some are triggered with transistors and some with drain controller.

 

Is it safe to test it without load?

The constant current design needs the load to work. The constant voltage booster does not need a load.

 

The constant current design needs the load to work. The constant voltage booster does not need a load.

Finally,I got it working with the new PCB layout.
I measure approximately 23.6-24.5 V at the output.
Thank you for helping us!

A few more questions:
Is it correct that 20mA @ 24V , I need minimum 3.2V @0.15A + conversion lost + inefficiency?
For extra safety, if I place a resistor/transistor at the input with the battery to limit the maximum inn current, will it affect the IC regulator or is it OK as long as I don't load the IC regulator with more power than it can deliver?

 

Boost converters need high peak currents to start up. It may not work with current limiting on the input.

 

A few more questions:
Is it correct that 20mA @ 24V , I need minimum 3.2V @0.15A + conversion lost + inefficiency?

That's about right value.