ESP32-S3-PICO-1 Issues

[r758432]

Hi,

I recently purchased a pack of ESP32-S3-PICO-1 SIPs. However, I cannot establish USB connection with them. I have previously connected to other S3 modules, and I am wondering if there is some issue with the PICO that is preventing me from connecting to it. I have attached an image of how it is connected on my PCB, and can confirm that all voltages provided are as they should be. Have I made a mistake on my PCB, is there a known errata on the USB interface or is this likely a soldering issue?

Thanks

S3PICO.png (482.43 KiB) Viewed 4319 times

[ESP_Sprite]

Do you have access to the U0TxD pin? If so, do you see serial chatter on that when you apply power to the board? Also, can you post the schematics?

[r758432]

Here is the schematic. I can confirm that the power supply is stable.

I can't access the UART right now, but should be able to Monday.

I can't see anything on the USB monitor, even in the wrong boot mode or corrupted previous S3 boards would have some kind of ping when plugged in.

Could it be that VDD_SPI needs to be connected to something? Or does USB need to be enabled somehow?

Schematic.png (174.99 KiB) Viewed 4211 times

[r758432]

Well, this is embarrasing. I don't seem to have given the SIP a ground connection.

I'm very sorry for anyone's time I may have wasted, I'll update once I've added a ground wire

[r758432]

Final update:
It works, I simply forgot to ground the EP pad

[ESP_Sprite]

No worries, shit happens, glad you found the issue!

[second.string]

I'm having this exact same issue, and unfortunately my main IC pad is already confirmed connected to ground. I know the chip is running properly because I can flash both a bootloader and bootloader+basic application firmware over UART. All voltages look good, except I don't see any output on VDD_SPI (aka 0V). I left it floating, from what I can tell from the datasheet it will act as an output for external flash if necessary, set to 1V8 or 3V3 depending on the strapping pin configuration.

Everything with the ESP32 is running just fine except for the USB. There's nothing that happens when I plug it in. On mac, there is zero change under the USB tab in the System Information application, and there's no new serial port enumerated in `/dev/cu.usbserial-****` or `/dev/tty***`. On Windows, there is also nothing that appears in Device Info or when running `mode` in cmd.exe to see a list of COM ports.

Because the rest of the ESP is running just fine, this feels like a configuration issue somehow. I've used the WROOM package of the S3 before, and have successfully set it up to flash/log over USB. I can't remember if I had to do anything special there to get that working though.

Any thoughts on this? Happy to share my schematic and/or layout but it's basically the exact same as the one shared previously with respect to the USB pins.

[ESP_Sprite]

Did you potentially swap your D+ and D- pins maybe? In general you should see a 15K'ish pullup (within the ESP32S3) to +3.3V on D+ while D- should be floating,

[second.string]

I've checked 10 times that I didn't flip them but I still could be missing something obvious. I'm not seeing any issues with the schematic or layout (and I've done this basic USB C 2.0 sub circuit multiple times before), but I've attached both for others to take a look. I redacted a lot of the layout just to make it easier to focus on the USB section - I'm happy to share the rest of it if that would be helpful.

I'm not seeing any pull-up on either data line, they're both at 0V with the cable plugged in. I can try to pull off the two 0 ohm series resistors to see if the ESP will pull up D+ by itself when it's open circuit.

I know my USB lines are not impedance matched - that has never caused an issue before for these slow speed 2.0 signals, especially when the trace length is ~1.5cm.

My USB C cable works fine for programming other S3 (wroom) boards over USB, so I know the cable is good.

I added flux around the ESP and used a heat gun until the solder liquified, then let cool again, to try to re-seat any of the connections. No change. I also don't think I added that I assembled two separate boards and I'm seeing this same issue on both of them.

[second.string]

Popped off the 0ohms and I see both lines from the ESP-side are floating. D- around 0V6-0V7, D+ around 0V3 to 0V4. I also gave the USB connector another hit of flux and the soldering iron to double check that nothing dodgy was happening with pins there but that did not help.