3 second delay (only) for the same signal/sensor?

[padrino]

Hi,

I am using rflink adapter for iobroker.

With a window sensor like this

51wsgimzXYL._AC_AA140_.jpg (2.02 KiB) Viewed 30123 times

,

which is recognized as "EV1527", I track opening/closing of a door.

This seemed to work quite reliable, but sometimes rflink just missed a state (change).

After a little investigation the cause was found - my wife.
She just rushes too fast out of the door, so the time between the opening and the closing signal (which both are the same) is too short.
Tried to messure it, there needs to be a delay of about 3 seconds to work.

I noticed with my different temperature sensor, that they sometimes get logged, even if they submit their data within a second or so.

So I thought, maybe its related to this "EV1527" kind of sensor.
As I have more of them (on windows), I tried to get signals from the door and another, very shortky after each other.
And to my suprise (again within about a second) this seems to be no problem for rflink, it receives and acts on both signals.

So, my theory, the delay only happens with the exact same sensor and not just if a signal comes from the same type of sensor.
Could this be?
Is this maybe some safty mechanism to prevent rflink from a false "double recognition"?

If this is known and wanted, is there a way arround it, maybe have this delay settable?

Would be great, if someone enlightend me. =)
Thanks.

CU,
padrino

[Stuntteam]

There is a mechanism to prevent reporting of double states. However, this is only for the same sensor and the same state.
Indeed as you noticed this means that 2 different sensors can send very shortly after eachother as far as it concerns RFlink.
It also means that, again for RFlink, the same sensor can send an on and off signal very shortly after eachother as the signals/states are different.

The problem, at least for the sensors I have tested, is within the sensor..
The sensor transmits fairly long and when you trigger a state change when the sensor is still transmitting,
it does not abort the transmit sequence to send the new state and also does not send the state change after completing the transmission.

[padrino]

Thanks for your reply.
Well, in my (probably special =)) case, the sensor only has one signal/state to transmit, ie. "true".
I just managed to position the magnet in a way that the sensor gets triggered on removal and returning of the magnet (rest of the "AI" is done internally by a toggle of the state).
Would it be possible to have a way to individually configure the "protection delay"?
Don't know if I am right, that is 3 seconds at the moment?
I think in my case, 2 seconds would make it work.
Not sure if 2 seconds would be "save enough" for the public, though...