SCM20100E SoM doesent want to play nice with my linear regulator

Ive made 4 seperate boards with the same SoM with almost same power circuits powered by 12V
reg1
reg2
reg3

Board1 - the 5V regulator gets really hot 100+ C, but it works
b1

Board2 - same schematic as the above, but the regulator goes into thermal shutdown after some few minutes

Board3 - is a bit different, it has 8V, 5V and 3.3V - but the 5V and 3.3V never get the proper voltages once i solder the SoM on the board. they work as expected without the SCM20100E.

Ive also tried to replace the 1A 5V regulator with the biggest i could get for the package 1.5A and it doesn’t make a difference.

This same setup i had with G30 (112mA) and it worked ok vs. SCM20100E (205mA) which doesent, why is it getting so hot and not working on the 3rd board?

Sounds like there some something wrong with your design after the above power side of things.
For us to be able to offer any real help you need to show your full sch either publically or reach out to GHI or someone via PM to review your design. Eg are you sure your SCM20100 footprint is correct?

What sort of thermal layout do you have under the regulators on the PCB? Vout needs a good thermal area on the top of the PCB. The datasheet for the LM1117 has details of this.

I think Justin is right, check the rest of the design and layout.

The 3.3V regulator doesn’t heat up much, the 5V is getting at 100 C

What is the 5V feeding other than the 3.3V to the SOM?

How are you dealing with thermal from the 5V regulator?

Nothing, its just connected to mikrobus connector for future use with a click board

Have you tried to inject 5V from a bench power supply?

Yes, works ok, doesent heat up

Then regulator is too small or capacitors not sufficient causing it to oscillate.

How do the voltage rails look with a scope? Clean or noisy from oscillation?