mmWave Presence Detection - ESPHome style

Thanks for the code. Works neatly.
Could you please share if it is possible to adjust the sensor’s sensitivity - if so how?
Thanks

Where did you see it’ll be available in 2 weeks? I still see out of stock at least for the human sensor

34 in stock → https://www.digikey.com/en/products/detail/seeed-technology-co-ltd/101990883/15976335

I have ordered one myself. Curious to perform a direct comparison between the SeedStudio and DFRobot.

Open to comparitive test suggestions; response time, wall reflection vs wall penetration, etc…

3 Likes

If these things trigger with the wind, that seems even less helpful than PIR. What am I missing here?

I’m seeing this in the manual:

DP3: Someone close to the device /
someone moving away from the device /
someone moving without direction

Does it mean that I can effectively detect if someone is very close to the device with the human sensor too? I thought that was only possible with the 60GHz version

I cannot speak to the SeedStudio at this time. The DFRobot responds to the effects of the wind, not the wind itself. If a breeze moves an object in the detection area, it will trigger.

Oh, I misread then. That makes way more sense!

Is the advantage of things like the Aqura that it does a better job of filtering noise? As it and the LifeSmart one advertise as ‘human’ sensors.

I have neither the Aqara FP1 nor the LifeSmart.

My experience with the DFRobot is that it responds “instantaneously” ( eg. <0.5 sec ) and have been a reliable human sensor for months.

2 Likes

Then I’m still misunderstanding. If its triggering with the move of a piece of paper or plant, how is it good at detecting humans?

Apologize for clearly being obtuse here…

It detects movement. Regardless of what or whom makes it. It is not advertised as “ignores everything else.”

That’s not what this says. Is it just advertising BS?

The human presence sensor adopts millimeter wave Doppler radar, sensing human movement and static presence.It detect via human movement Doppler parameters and human physiological parameters synchronized sensing technology for wireless perception of human status in the area

You will have to ask them

But you’re my resident expert! :smiley:

If one ever makes it my way, I’ll be sure to test how see-thru my cats are :stuck_out_tongue:

1 Like

@crlogic Thank you for putting together the threads on these sensors, they’ve both been super helpful in getting the sensors to work - great job!

I’ve been playing with the DFRobot one the last few days and noticed that after spending some time in the room (30 min+ with minimal motion such as typing), it’ll take about 1.5 minutes to switch from on to off (irrespective of the delay period set). Working well otherwise.

Did you notice anything similar?

Care to run an experiment?

Place the sensor in a temporary location within arm’s reach. After the 30+ minute period mentioned, place the sensor face-down on the desk. Leave room and confirm on-to-off delay.

I’ve been there before and removing possible movement interference has been helpful in troubleshooting configuration vs movement.

There’s a review posted on this blog

The moes human radar (TS-601) has similar presence detection but it doesn’t have fall detection like TuYa MIR-HE200-TY_fall control via MQTT | Zigbee2MQTT

Testing the moes and it’s pretty good so far detecting sleep, sitting on couch



@crlogic did you also get the seeed fall sensor?

1 Like

SeedStudio should arrive today (crosses fingers). Plan to assemble it w/ a DFRobot on a single ESP32 for some direct comparisons. Probably take some time as I like to fiddle w/ options/parameters.

2 Likes

I have order 3 of them as the delivery was as expensive as only one sensor LOL.

I’ll share my experience with it when it arrives!

Nice.

I noticed that the SeedStudio User Manual places some emphasis on input supply voltage ripple (section 2.1).

As I have no way to test my ESP’s, I was going to follow the Fig 2 schematic and place two capacitors on the 5V output of my ESP32.

image