IWR6843AOPEVM: IWR6843AOPEVM reduce power consumption for Overhead 3D people tracking

Hi,

We are looking into your query. Please allow us a couple of days to respond.

Regards,

Sharan

Hi, thank you for letting me know.

Hi,

Have you considered using the IWRL6432AOPEVM? It is much lower power consumption by default.

We can reduce the power of the IWR6843 somewhat, but it will always be in the 100's of mW.

Best,

Nate

Thank you for your response,

The IWRL6432AOPEVM radar does not seem compatible with the 3D People Overhead firmware. Additionally, we require high reliability for short-range measurements (<6 meters) with a radar positioned at a 45° angle. I suspect this radar might be somewhat less reliable in this regard.

Reducing power consumption seems more appealing, provided it does not significantly impact the radar's frame rate, which should be at least 2-3 frames per second.

Hi,

That's fine, then you can use the IWR6843 overhead firmware. I don't think the 6432 will be less reliable, though it will have reduced angular resolution. What is your power consumption requirement here? That will help us understand what device is best. I don't see the IWR6843 getting below say 100 mW of power. Any lower than that and you will need the IWRL6432.

Best,

Nate

Hi,

Thank you for your response.
Based on my measurements, when the radar is active, it consumes approximately 2.8W. When it is powered but the "sensor start" command has not been issued, it consumes 1.6W. The objective is to minimize the overall power consumption of the radar as much as possible. Let me explain:

Since it doesn’t seem possible for the radar to alternate between "sensorStop" and "sensorStart" modes, it is necessary to restart it by cutting its power supply. The event loop looks like this:

1. Radar startup: 1.6W for 160ms
2. Sending radar configuration: 1.6W for 170ms
3. Sending the "StartSensor" command: 2.8W for 110ms
4. First data acquisition: 2.8W for about 100ms
5. Second data acquisition: 2.8W for about 100ms

This results in an overall radar power consumption of approximately 2.17W over 640ms.

When the radar's antennas are active, its power consumption can be reduced by adjusting its configuration. However, for now, we intend to keep this configuration largely unchanged. This means that we are constrained to a consumption of at least 2.8W for 310ms (or maybe less by cutting CLI UART after it start ?). To reduce the overall power consumption, the remaining focus is on the phase that consumes 1.6W.

We are investigating whether it is possible to lower this power consumption (e.g., to below 1W) and/or reduce its duration (for instance, maybe by hardcoding certain radar parameters directly into the firmware).

Regards,

Valentin

Hi,

It will not be feasible to cut power to the device because the tracker needs data over multiple frames together to work correctly. If you want to reduce power consumption on the IWR6843, see this app note.

https://ti.com/lit/an/swra689/swra689.pdf?ts=1734386248989

Best,

Nate

Thank you for this documentation,

It seems to me that the tracker only needs two point clouds to create a track. This way of using the radar would solely be to exit a standby mode. If the radar detects a track after two measurements, it would then be continuously powered; otherwise, the power cycle described earlier would restart. Is this a feasible solution?

Is it possible to save time on the radar startup time?