BlueSCSI 2026.04.27-rel on PICO 1 Crashes when not plugged into a computer.

[Snial]

This is for a BlueSCSI v2 - 2023.10a.

It's a DIY build using my own PICO 1 Rev B0 from 2020, which is soldered in place with headers. I think my soldering is decent. It passed the initial tests with the 5x blink pattern and no SD card and then not getting a 3x blink pattern, though I believe it also didn't show an ACT activity (apart from during the first ms or so) then the green light on the PICO went on solidly.

I found that if I power it by USB plugged into my Mac (a 2012 Mac mini) it will boot OK and I think it's in EnableUSBMassStorage=enabled mode judging from the log. The Green LED flickers then. However, if I power it via the Berg connector or via a USB power supply (i.e. one that can supply 2.1A at 5V) it will show a steady green LED after a few seconds, which is the PICO crashing. Then I see the kind of log detailed below.

Has anyone seen this before?

Platform: BlueSCSI (Pico1)
FW Version: 2026.04.27-rel Apr 28 2026 00:33:46
Flash chip size: 2048 kB
I2C Supported
=== Bootloader Log ===
[0ms] Bootloader version: Apr 28 2026 00:33:33 BlueSCSI
[0ms] Looking for firmware: BlueSCSI_Pico_DaynaPORT*.bin
[182ms] Bootloader continuing to main firmware
=== End Bootloader Log ===
Reclock RP2040 & Pico 1/1W based boards to standardized speed
Initial Clock set to 125MHz
Reclocking the MCU to 203MHz
Setting the SDIO clock to 41MHz
After reclocking, system reports clock set to 203MHz
 
=== SD Card Info ===
SD card detected, exFAT volume size: 60874 MB
SD Name: SD64G, MID: 0x03 (SanDisk), OID: "SD" (0x53 0x44)
[29ms] -- Debug = Yes
[29ms] 
[29ms] === Global Config ===
[29ms] Reading configuration from bluescsi.ini
[31ms] 
[31ms] === Finding images ===
[79ms] === Finding images in / ===
[80ms] DBG -- Using block size, 512 from filename: HD30_512 20MB.hda
[81ms] == Opening /HD30_512 20MB.hda for ID:3 LUN:0
[81ms] ---- INFO: Image file is fragmented (2 fragments), FastSeek enabled for best performance.
[81ms] ---- Configuring as disk drive
[82ms] DBG ---- Geometry set from defaults: SectorsPerTrack=63 HeadsPerCylinder=255 total sectors 41056 (not divisible)
[82ms] DBG ---- Read prefetch enabled: 8192 bytes
[85ms] 
[85ms] == Platform supports ROM drive up to 800 kB
[85ms] DBG ---- ROM drive image not detected
[85ms]  
[85ms] === Configured SCSI Devices ===
[85ms] ID: 3, BlockSize: 512, Type: Fixed, Quirks: Apple, Size: 20528kB
[87ms] 
[96ms] DBG Waiting for USB enumeration to enter Card Reader mode.
[1196ms] DBG Waiting for USB enumeration timed out after 1000ms.
[1196ms] DBG -- Try increasing 'USBMassStorageWaitPeriod' in the bluescsi.ini
[1196ms] Clock set to: 203MHz
[1196ms] Initialization complete!
[1198ms] --------------
[1198ms] CRASH!
[1198ms] Platform: BlueSCSI
[1198ms] FW Version: 2026.04.27-rel Apr 28 2026 00:33:46
[1198ms] scsiDev.cdb: 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 
[1198ms] scsiDev.phase: -1
[1198ms] SP: 0x2003F778
[1198ms] PC: 0x20011040
[1198ms] LR: 0x2000EFC1
[1198ms] R0: 0x00000001
[1198ms] R1: 0x00000000
[1198ms] R2: 0x00000000
[1198ms] R3: 0x20035074
[1198ms] STACK 0x2003F778:    0x00000001 0x00000000 0x00000000 0x20035074
[1198ms] STACK 0x2003F788:    0x00000000 0x2000EFC1 0x20011040 0x21000000
[1198ms] STACK 0x2003F798:    0x00000000 0x2003F7AC 0xFFFFFFFF 0x20010D3D
[1198ms] STACK 0x2003F7A8:    0x100532FC 0x40A8E200 0x00000000 0x408F4000
[1198ms] STACK 0x2003F7B8:    0x2003D0B6 0x20005DD9 0xFFFFFFFF 0x20033F50
[1198ms] STACK 0x2003F7C8:    0xFFFFFFFF 0x00000000 0x20041F01 0x18000000
[1198ms] STACK 0x2003F7D8:    0x00000000 0x200011D5 0x2003D0BA 0x10020168
[1198ms] STACK 0x2003F7E8:    0x20041F01 0x18000000 0x00000000 0x20000E93

[erichelgeson]

Decoding the stack it's clear it's a B0 double issue. I'm going to see if I can scan/catch this in our CI job so those don't sneak in accidentally. If you're comfortable building you can for a test comment out the line that logs the voltage. If not I'll see if I can get a PR with the scanner + fix tomorrow. Thanks!

[Snial]

Thanks very much for the quick reply! I can pull the project and build it, then try and find the line of code that logs the voltage. I guess you're saying that it's trying to log the voltage and that causes it to crash? I was wondering at one point if it was an issue to do with the CPU frequency, if B0's couldn't be reliably overclocked to 203MHz.

[Snial]

Hi Eric, just to check on my 'deadline' for compiling the project. When are you next likely able to look at it, so I can aim to build it at my end in time (I'll probably need to up date the SDK and compiler tools including gcc-arm-none-eabi).

[erichelgeson]

There could additionally be a 203MHz OC issue on the B0 (I have none to test) - though that can easily be worked around by setting a SpeedGrade in the ini file. Also there's no deadline for you, I was just saying if you wanted to try, if not thats 100% fine too :)

[Snial]

@erichelgeson . Thanks for the reply. I was in the process of installing the compiler and sdks (I have cloned BlueSCSI too). So, if I stuck:

[whichSection?]
SpeedGrade=125000000

Should that work?

OK, I'm having a bit of difficulty installing everything. I have the normal PICO sdk updated. I did brew install --cask gcc-arm-embedded, but it was gcc 15 and the directions on BlueSCSI seem to say you need gcc 14. Can't figure out how to brew install ARM embedded v14.