Arducam 16MP quad array only works once then needs power cycle

Raspberry Pi Cameras
1

Hi everyone,

Hoping someone can help with this one. I’ve got the 16MP quad array HAT and it only works once, then I have to power cycle the whole thing to get it to respond again. It captures one frame fine, then the next attempt just hangs or fails until I cut power and bring it back up.

I found this thread which sounds like the exact same problem: 16 MP quad array only works once then needs to be powercycled

Here’s a test capture and the log from when it fails:

arducam-test.jpg

arducam-bug-report.txt

Any ideas?

2

Thanks for the detailed report. I went through the bug report log and the test image. Before we dig into the timeout, can you tell me how you’re currently powering the HAT - are you feeding it through the GPIO header or a separate supply? And what host board is it connected to?

3

Right now the HAT is powered through the GPIO pins. It’s connected to a Raspberry Pi 4. One other thing I should mention - the motor driver isn’t working either. Not sure if it’s related but I wanted to flag it.

4

Got it, thanks. Based on what I’m seeing in the logs, this points to insufficient power to the HAT. The quad array pulls a lot of current, and after the first successful capture the GPIO supply can’t keep up - that’s why it works once and then needs a power cycle.

Can you try powering the HAT with a separate Type-C connection instead of going through GPIO? That gives it a cleaner, higher-current supply.

I’ve noted the motor driver issue as well - let’s get the power cycle problem sorted first, then we can dig into the motor driver.

5

Okay so I tried powering the HAT with the extra Type-C connection instead of GPIO, and unfortunately nothing changed. It still works exactly once.

The preview came up the first time, and on the second attempt I get:

Error: Device timeout detected, attempting a restart!!!

Here’s a photo of how it’s wired now:

image0.jpeg

Just to recap your earlier reply:

  • Insufficient power suspected.
  • Motor driver is also not working and will be checked after the powercycling issue is solved.

So if the Type-C supply didn’t change the behavior, where does that leave us? The timeout happens every time on the second preview.

6

Thanks for the photo and the update. That’s useful info - if the extra Type-C supply made no difference at all, then this isn’t just the GPIO power path being weak. The ‘Device timeout detected, attempting a restart!!!’ on the second preview suggests the sensor array is dropping off the bus after the first capture rather than simply browning out on power.

So to summarize where we are right now: we suspected insufficient power, the Type-C supply didn’t resolve it, and we’re now looking at a device-level timeout on the second read. The motor driver issue is still on our list and we’ll pick it up once the power cycle/timeout problem is sorted. We’re going to dig deeper into the timeout path next.