Some months ago there was a topic about banding on the IMX462, see: viewtopic.php?p=2182756&hilit=exposure#p2182756. I have the same inno-maker IMX462 camera here connected to a RPI3 and I also see the same issue.
Unfortunately the topic is closed now, so I can't just continue the investigation there but instead have to open this new thread.
Anyway, as claimed previously the HGC mode does help. However I do want to have a fix for the source of the problem which seems to be an improper analog 3V3. I noticed the StarlightEye had a similar issue in the past (https://github.com/will127534/StarlightEye/issues/10), but they claimed it is fixed in their latest iteration. So when I look at the schematic the StarlightEye is nowadays taking a dedicated 5V DC input, and they make their own analog 3V3 using an LDO on the camera board itself. Back to the IMX462, it is taking power directly from the CSI2 connector, and thus does not use the 3V3A that's available on the RPI3 but just the regular 3V3.
So as I understand a dedicated 3V3A might help us fix the issue. I can read schematics and perform mods, but I'm not a hardware engineer pur sang so I might need some guidance here. First of all, I have some a cheap scope, but what kind of noise do we think to measure with such thing?
Secondly, I external 3V3 circuit could probable also help us fix the issue, say for example using following module https://www.diyinhk.com/shop/audio-kits ... 15ax2.html ?
Unfortunately the topic is closed now, so I can't just continue the investigation there but instead have to open this new thread.
Anyway, as claimed previously the HGC mode does help. However I do want to have a fix for the source of the problem which seems to be an improper analog 3V3. I noticed the StarlightEye had a similar issue in the past (https://github.com/will127534/StarlightEye/issues/10), but they claimed it is fixed in their latest iteration. So when I look at the schematic the StarlightEye is nowadays taking a dedicated 5V DC input, and they make their own analog 3V3 using an LDO on the camera board itself. Back to the IMX462, it is taking power directly from the CSI2 connector, and thus does not use the 3V3A that's available on the RPI3 but just the regular 3V3.
So as I understand a dedicated 3V3A might help us fix the issue. I can read schematics and perform mods, but I'm not a hardware engineer pur sang so I might need some guidance here. First of all, I have some a cheap scope, but what kind of noise do we think to measure with such thing?
Secondly, I external 3V3 circuit could probable also help us fix the issue, say for example using following module https://www.diyinhk.com/shop/audio-kits ... 15ax2.html ?
Statistics: Posted by witte — Thu Oct 17, 2024 9:00 pm — Replies 1 — Views 23