hardware-tiny PCB - ESP32 LED doesn't turn on, no 3.3V output from U1 (2 boards tested)

[Samos68-bit]

Hi, I built the hardware-tiny PCB (ordered fully assembled from JLCPCB, PCBA Economic) and I'm stuck on a power issue.
Setup:

5 PCB ordered, fully assembled (PCBA)
ESP32-WROOM-32 (38 pin, CP2102)
Tested with real 12V from CN-REMO (LG indoor unit), and confirmed with bench 12V supply

Main symptom: the ESP32's power LED never turns on when plugged into the PCB and powered via J1 (12V).
The ESP32 itself works perfectly fine when powered via USB-C directly (confirmed Wi-Fi connection, logs, everything normal). So the issue only happens when relying on the PCB's onboard regulator.
Measurements (same result on 2 different boards tested):

J1 input: ~12.7V confirmed with multimeter (between red and black wires)
Output on ESP32 socket pins (3V3 vs GND), WITHOUT the ESP32 plugged in: 0V
With ESP32 plugged in: LED stays off, no boot in logs

Question:

Is 0V between the 3V3/GND pin sockets (no ESP32, 12V applied on J1) actually abnormal? Or am I possibly measuring the wrong pins? Could someone confirm which exact pins on J2/J3 correspond to 3V3 and GND so I can double check?
Also: I had to substitute C1 (originally EEEFC1C100AR) with an alternative (RVT1E100M0405, same 10uF, 25V instead of 16V, same footprint) due to JLCPCB stock shortage at order time. Could this substitution be related to the issue?
Any help would be greatly appreciated, thanks!

[kchen]

It's expected for the power LED to turn on only when connected to USB. The capacitor substitution doesn't sound related to me, but I'm not really familiar with different capacitor options (and haven't reviewed the details). One other difference I noticed in your report is that you mentioned you're using an ESP32-WROOM-32. I used an ESP32-WROOM-32E since that's what the the project suggests -- don't know if that makes a difference (although I guess since you're not getting 3.3 V at all, that sounds like it shouldn't matter).

[Samos68-bit]

Thanks for the quick reply! That's good to know about the LED — that clarifies things a bit.
However, the real issue is more fundamental: I also measured the voltage directly on the ESP32 socket pins (3V3 to GND, on J2/J3 — without the ESP32 plugged in at all), with 12V applied on J1. I'm getting 0V, when I'd expect ~3.3V coming out of the U1 regulator regardless of whether the ESP32 is plugged in or not.
So it seems like the regulator itself isn't outputting anything, not just an LED behavior difference. I tested this on 2 different boards from my batch of 5, with the same result on both.
Regarding the WROOM-32 vs WROOM-32E: good point, I'll keep that in mind, but since the problem happens even without the ESP32 inserted, I don't think that's the root cause here.
Is there a way to verify if U1 (TPS82130, MicroSiP package) was correctly soldered? Since it's a BGA-like package I can't visually inspect the solder joints. Could there be a known failure mode for this specific regulator, or a common assembly mistake with this footprint?