mmWave Presence Detection - ESPHome style

Hm, strange. Then there is probably something wrong with my soldering. I have problem with reliability for more than 3 m.

Did you try it with a bigger difference from the electric equipment?

Edit: My uart gives some funny log. Is than normal?

[17:27:59][D][uart_debug:158]: <<< “$JYBSS,1, , , *\r\n”

That’s normal. The first number says whether there is presence or not. 1=occupied

Ah thank you! I fear I fried one of my boards. It only detects movement if it is 5 cm away and that one second delayed. The uart settings have no effect on this one board. Anyone else having this problem?

I’ve also got the DFRobot one setup - thanks for the config!

I’m wondering in which orientation the beam is - looking at the wiki I can’t tell from their pictures.

Initially I thought that having it with the pads 3 wide 2 high would have the 140 degree beam horizontal, but in testing it feels like it’s the other way around as it wasn’t quite making it to the edges of the room. Certainly many of the pictures above have it mounted 3 high 2 wide.

Any ideas?

Good question. Would need a decent sized room to really test I would think.

Other mmWave datasheets have radiation emission charts. Sadly not the DFRobot. No harm though if you can spend some time tuning placement. This is where a very short cooldown latency is helpful.

Flagging this related device in case there is interest over here. Cheers.

1 Like

From testing it looks like having it mounted 3 tall x 2 wide (like in the image below) would give a 100x40° beam.


1 Like

Hello All
In past I wast trying to set up RCWL-0516 chip. Didn’t get very far as other commitments kicked in and I didn’t have time to progress this.

What is the differance between the two?

Also, instead of the USB cable, can you have a rechargable battery?

Thank you

This seems to reflect what I’ve tested also… Note I’m assuming 100 x 40, means 100 horizontal, 40 vertical. What’s bugging me is why… :wink:

The DFRobot spec says 90mA so think you may struggle to get a meaningful battery life out of it, while running all the time. I’m currently measuring the temp of the enclosure around 10 degrees above ambient so think it is consuming quite a bit of energy.

Seeing as the DFRobot is out of stock in most places, I’ve seen this 60 Ghz alternative, do we think this could substitute as a replacement? This would be my first time making a DIY sensor, so want to make sure before I purchase anything.

I was curious about this one as well and have looked at the datasheet which appears decently thorough; including radiation dispersion patterns.

One important difference with this sensor is that it requires local software installation for configuration using their software and no communication protocol is explained in the datasheet.

1 Like

Here’s a video on using the dfrobot sensor


Inception… :wink:


Why do so many youtube videos have no useful information in either the thumbnail or the title. This drives me mad !

(Rhetorical question - I know the answer. But it still annoys me to no end.)

Too keep this comment somewhat on topic, I’ve ordered my first sensor through Pimoroni. :slight_smile:
It was in stock there and should arrive in a few days.
Thanks to everyone who took the time to research & document this particular sensor !

1 Like

The latest project has been posted…

Hi, great guide and and awesome that you share this with us.
I am looking to make this but found that DFrobot also has a smaler sensor mmWave Radar Sensor Arduino-Human Presence Detection Wiki - DFRobot
You think this will work also, similar spec but on the 5.8GHz band.
And cheeper.
Want to do around 10pcs so finding a similar but cheaper sensor would be awsome.

Also, thought about something else… Sliding doors usually use this kind of sensors.

Would that not be possible to connect to an esp also?

I’m sure you could but 12-24V, you’d have to power it from some other source, I don’t know of any boards that will output more than 5V.
Plus wouldn’t the range on these be quite low ?

Yeah power you need to work around… And it states that if you mount it 2m up and point it down, the range is 2m… I guess the real range is similar to other sensors if you not point it to the floor.