Hey everyone, I’m really stuck here and hoping someone can point me in the right direction.
We’ve got an Arducam UC-717 Rev.D1 camera module that we’re trying to integrate into a Xilinx ZynqMP SoC design. The problem is, I cannot find any documentation whatsoever on what imaging sensor this board actually uses. Some ex-colleagues tried getting it to work before they left but they never managed to produce a single camera image — and apparently they didn’t follow any instructions (or there weren’t any).
So what I really need to know is:
What imaging sensor is on this board, and where can I get the datasheet?
How many differential (MIPI) lanes does it use for data transmission?
What lane rate should I be configuring?
Without these I’m dead in the water on the FPGA side. I’ve attached photos of the module (front and back) — hopefully someone can identify it.
Thanks for the detailed post and the photos — those are helpful.
Before I can confirm the sensor and point you to the right docs, could you snap a couple more shots? Specifically, I need a clearer view of both the front and back of the camera module itself (not just the carrier board). Sometimes the sensor marking or the ribbon connector layout gives us the clue we need to nail down which variant you have.
Also, just to confirm — you mentioned this was purchased from Arducam directly, correct? And the platform is a Xilinx ZynqMP, so this isn’t going through a Raspberry Pi ISP pipeline, which means you’ll likely be doing some custom driver work regardless.
Once I can ID the sensor, I’ll get you the relevant documentation and the MIPI lane details you need.
Thanks for the quick response. Here are the pictures again — I double-checked and these are the best shots I have of the module, front and back. Let me know if they’re clear enough or if I need to pull the module and take closer ones.
And yes, correct on both counts — purchased from Arducam, and we’re on a ZynqMP. We already have a MIPI CSI-2 RX subsystem in our FPGA design, so we just need the sensor specifics to configure the PHY and the capture pipeline properly.
Okay, thanks — those photos are enough to work with. I can confirm your UC-717 Rev.D1 is using the Sony IMX296 sensor. That’s a 1.58MP global shutter sensor.
Now, fair warning — that doc is written with a Raspberry Pi target in mind, so it won’t be a drop-in solution for your ZynqMP. But it should give you the key details you need. From the IMX296 side, here’s what you’re looking at:
MIPI lanes: 2-lane MIPI CSI-2
Lane rate: The sensor supports up to around 594 Mbps per lane (check the IMX296 datasheet for the exact table — lane rate depends on the frame rate and resolution mode you select)
Datasheet: The full Sony IMX296 datasheet should be referenced for the register map and timing details. You’ll need that to configure the sensor over I2C from your ZynqMP.
For your FPGA side, you’ll need to configure your MIPI D-PHY and CSI-2 receiver for a 2-lane input at the appropriate lane rate, and then handle the RAW data output format that the IMX296 produces. The docs I linked should help you get the driver model sorted out — you’ll just need to adapt it from the Raspberry Pi V4L2 framework to whatever your ZynqMP pipeline expects.
Let me know if you hit any specific snags once you start digging into the IMX296 datasheet.
