I have purchase a few of the Linptech HPS's, for the life of me I can't get them to work. I have tried the built-in driver and @kkossev 's awesome driver.
Aside from random (and sometimes significant) difficulties getting them to pair, they are erratic as heck.
Primary use case is to replace an aeotec PIR in the bathroom, to get presence behind the shower screen. It is mounted on the wall above the doorway. Bathroom is quite small.
Regardless of the sensitivity settings, it will sometimes trigger motion, but forget holding presence unless I move closer than about 1-2 feet from the sensor, and even then it won't 'hold' for more than maybe 5-10 seconds. The LED seems to flash randomly. There are also false positives. I have tried two separate sensors in case the first one was a dud.
Summary: it honestly feels like these devices are nothing but a random timer firing for motion triggers and LED activity. Neither settings nor actual motion/presence seem to have much impact. I have re-paired more times than I care to admit. My ZB mesh is very stable aside from these devices.
Below are screenshots of device data & current states, and logs from the sensor while I walked in and stood 2-3 feet in front of the sensor for a few minutes. I've read numerous posts where various folks claim they work well (but still need to rely on a PIR), but I'm at a loss...is there something I'm doing wrong here?
The logs that you submitted show very erratic behaviour of your mmWave sensor...
There should not be 'Motion is active 0s' logs sent four time in one second.
There should not be 'motionDetectionDisatance is 3.75 meters' logs sent so frequently!
I can think of two possible reasons for this misbehavior :
Problem with the power supply - the device resets itself due to power glitches.
Heavy interference from something in this small room ...
Is it likely to have any of these possible interference sources/reasons in your use case?
I would first try to set up the mmWave sensor (just for tests) in a better environment - test it in a large and rarely visited room, where :
There are no fans and air conditioners
There are no curtains or plans moving
There are no ventilation or water ducts
No pets and robots allowed! : )
There is no movement on the other side of the wall where the sensor is installed
The room is large enough so that the reflected microwave signals from the walls are weak enough
No other mmWave sensors around
The mmWave sensor and its power supply cable are not close to any other electronics that can generate electromagnetic noise
I would also increase the fadingTime parameter - 10 seconds is too low / too optimistic for a reliable detection of 'no presence' state in a problematic environment :
Try to make the mmWave sensor work OK in a near-ideal environment first - you will have a good starting point for adjusting the sensitivity settings when you move the sensor to its final destination.
First, check or change the power supply - these 'motionDetectionDisatance is 3.75 meters' logs should be seen only when you refresh the device (I suppose you are not refreshing it from a RM5 rule), or when the device is rebooted.
They do work quite well, majority of us who have purchased them are very happy with them. Some of us like to add a secondary traditional motion sensor if we have one that triggers slightly faster than the Linptech, so we get a quicker turn on event when there is motion in the room. But it's not a requirement at all to add a secondary sensor.
@kkossev 's comments and suggestions are spot on (as usual), follow his lead and you will get this figured out.
I have five of them working well at the moment. I do have an issue where they sometimes will not be responsive after rebooting the hub, and I have to unplug/plug them to have them reconnect. As such, I avoid rebooting unless I have to.
Others have said they stop responding on their own every couple weeks, but that is an issue I have not seen with mine. Everyone seems to have different experiences with these, so environment can be a factor, and things like ceiling fans and heat sources can trip them so it takes some adjustment to get them dialed in for direction, range and sensitivity.
I had one that started seeing a ghost at 3 meters, and it was aimed down into the room from about seven feet. I started to wonder if I had a mouse living under the floor or something. I dropped that sensor down three feet to look straight out level into the room, and that ghost (or mouse detection) went away. It was probably getting some kind of reflection, but it was strange how it started doing it after being fine for a year in that position.
I used the built-in driver until I was directed to kkossev's driver. Now I also use them as illuminance sensors for my indoor lighting app with that driver, and they work very well as illuminance sensors.
its worked reliably and without issues powering an aeotec multisensor 6 for at least 2-3 years now. But just maybe the power input circuit of the aeotec has a larger capacitor than the linptech which could buffer any fluctuations from the PSU? don't know if thats grasping for straws but i'll try to find another USB source to see if it makes a difference.
How long is the USB power cable between the adapter and the Linptech sensor? If you have on hand some ferrite beans that can be installed at both ends of the cable, you can try if these will help. Or, try to kee the cable as short as possible and away from any other cables.
Do you have any inductive loads (such as fans) around? If yes - is there a fan speed controller in use? These are often sources of very strong electromagnetic interference.
I’ve also found that the beam is much narrower on these than my pir sensors and have to aim them down when mounting them up higher. The ones that work best for me are mounted around chest height.
Linptech supplied USB cable is rather long… and thin! If you don’t have clip-on ferrite ring cores on hand, try to find a shorter cable. As shorter and as thicker - as better.
mine's mounted above the doorway, but i angled it down using the supplied mounting bracket thingy so hopefully that should negate beam width as an issue.
chest height would be nice, but then i'd have WAF issues due to a highly visible/ugly (and long) USB power cable snaking down around my bathroom wall cabinet.
Hi @kkossev right again! all it took was a simple change in USB cable route so it didn't cross over an LED light strip(!). This made a HUGE difference. Like night & day. Touch wood its now solid as a rock and it actually does what its supposed to, more detailed testing and tweaking aside.
key points for me:
the significance of the repeating motionDetectionDistance log entries which indicated the device is rebooting.
I honestly didn't think an LED light strip would act like such an antenna for EMI from its controller - which is over a metre away. LED strip is Mercator Ikuu brand.
How ridiculously sensitive the Linptech HPS sensor must be, responding so drastically to this remote source of EMI.
per @danabw comments: i now know what the fuss is about...now i finally have experienced a working HPS i've gotten a taste of a much more reliable lighting automation experience, in addition to simplified lighting logic & rules. at least in this room, i've even ditched the PIR (for now?) as initial tests showed it was unnecessary with a now-reliable-HPS. woo...freakin'...hoo!
mine work fine and as others occasionaly drop of the net so i use a driver that i can tell when it happens and have them on a usb switch that when it does occur manually intervene and reboot them .. probably happens 2 times a year or so..
but its on a separate electrical circuit to the USB charger/Linptech. EDIT: with no fan speed controller..
