UPDATE:
HACKING OF THE LD2450 RADAR
From the specification I was able to obtain, it appears that Hi-Link’s radar is a close clone or a modified S5KM312CL.
I managed to get access to some of the documentation, which indicates that by setting a specific state on one of the circuit pins, this device can be switched into a mode where it sends additional data. I marked the pin in the photos. Using makeshift methods overnight, I connected to the mentioned pin and managed to read, or at least receive, FFT data: range, doppler, peak. I’m still struggling with the format of this data and the packet frames.
It’s possible that I’m mistaken and the data I’m receiving is something different. Especially since the expected baud rate is 115200, but the data resembles what I want to achieve only at a baud rate of 921600.
It’s also possible that the data format for Hi-Link is different, so I ordered RD-03D which are officially built on S5KM312CL and should be compatible with the documentation (time will tell).
ORIGINAL POST TEXT:
After an extensive search, I’ve discovered that many sensors, including HiLink and AiThinker, seem to be clones of the S5KM312CL, which itself might be a clone of another sensor – it’s a complex maze.
Up to this point, my research indicates that Iclegend offers the most extensive configuration options. Notably, their 1T1R/1T2R sensors bear a striking resemblance to those from HiLink.
A key discovery I’ve made is that setting certain pins to a high state for a few nanoseconds after startup or reboot switches the UART to raw mode, unleashing a wealth of data.
The resources available are comprehensive, including source codes, datasheets, advanced calibration/configuration tools, and exhaustive documentation. However, purchasing these tools is a major challenge. My attempts to use Iclegend’s tools on Hi-Link sensors have been unsuccessful, likely due to some customized internal configurations that prevent compatibility with the original tools.
Another obstacle is the cost and availability of Iclegend’s development board, priced steeply at $250 and seemingly impossible to purchase. Access to most tools and documentation is restricted, requiring proof of purchase.
Additionally, I’ve learned that this radar system can operate in a master/slave mode, akin to the LD2461.
In summary, there’s a wealth of information and potential here. The optimal route seems to be purchasing the 1T2R sensor from Iclegend, complete with their support, and possibly designing custom PCBs for managing raw data transmission. The less desirable option involves hacking the LD2450 and modifying it to enable raw data mode through specific pin settings.
But their tools and firmware options are crazy in possibilites.
few links: