Issues with DY-SQ100B Water Leakage Detector

[pyrodex]

Referencing this posting in the group from last year:

https://groups.google.com/g/rtl_433/c/Is2bSWStKlc

I purchased a few of these from AliExpress and am not having the same luck as previously reported.

Here is my configuration:

# Ansible Managed file, please do not edit locally!
output kv
hop_interval  60
frequency     433.92M
#frequency     315.00M
frequency     345M
#frequency     868.00M
frequency     915M
convert     si
report_meta time:iso:usec:tz
report_meta level
report_meta noise
report_meta stats
report_meta protocol

decoder {
    name=LeakDetector,
    modulation=OOK_PWM,
    short=316,
    long=968,
    reset=916,
    gap=0,
    tolerance=261,
    bits=33,
    unique,
    get=Location:@0:{16}:[62080:HotWaterHeater 5853:KevinSink 7133:KitchenSink],
    get=Message:@16:{8}:[250:Alarm],
    get=Battery:@24:{4}:[14:Ok 12:Low]
}

I am testing with a unit that via rtl_433 reports the following:

time      : 2023-08-20T12:04:19.304477-0400        Protocol  : 30
model     : Generic-Remote                         House Code: 62080
Command   : 205          Tri-State : 110XX000101Z
Modulation: ASK          Freq      : 433.9 MHz
RSSI      : -0.1 dB      SNR       : 13.3 dB       Noise     : -13.5 dB

Would appreciate the assistance on getting this resolved.

Thanks!

[zuckschwerdt]

If you can: grab a sample file, check there is a good signal and upload here as zip. See https://triq.org/rtl_433/ANALYZE.html

[pyrodex]

These help? @zuckschwerdt

https://www.icloud.com/iclouddrive/008kV4vl1UfQMNEUCJP2dNeYA#g003%5F433.92M%5F250k

https://www.icloud.com/iclouddrive/06euv-z3y0neBntG4gCcGuiew#g002%5F433.92M%5F250k

[pyrodex]

If you can: grab a sample file, check there is a good signal and upload here as zip. See https://triq.org/rtl_433/ANALYZE.html

Here you go.

Archive.zip

[zuckschwerdt]

Wow, that is some ugly and unstable signal. I assume there is a constant alarm going (the signal at 433.82 M) and the two additional FSK signals araound 433.95 M are TPMS or a weather station.

Timings look like a classic EV1527 24(25 with sync) pulses PWM at 500 or 1500 µs in 2000 µs intervals.
E.g. rtl_433 -R 0 -X 'n=waterleak,m=OOK_PWM,s=500,l=1500,r=2000,bits>=24' /Users/zany/Desktop/Archive/g003_433.92M_250k.cu8
the code here is {25}8780328 or just 878032 if we assume classic trailing sync.

Are you sure there is not just an alarm but also a heartbeat signal? Did you see those? And a battery ok bit would also be very uncommon.

[gdt]

@pyrodex What's the status of this issue? Are you still trying to work on a decoder?

[pyrodex]

I’ve been trying but for side tracked by another project of course… On Oct 15, 2023, at 6:52 AM, Greg Troxel ***@***.***> wrote:  @pyrodex<https://github.com/pyrodex> What's the status of this issue? Are you still trying to work on a decoder? — Reply to this email directly, view it on GitHub<#2606 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAA2RH4N4BV3GQBTNNIEIJTX7O56BANCNFSM6AAAAAA3XLYVTM>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[klohner]

I took a look at this signal and got the same result @zuckschwerdt saw. This does not look like the same kind of signal as seen in the "DY-SQ100B Water Leakage Sensor" referred to in [this message thread].(https://groups.google.com/g/rtl_433/c/Is2bSWStKlc)

You can use a flex decoder like this on your signal:

rtl_433 -R0 -X "n=WaterLeak,m=OOK_PCM,s=512,l=512,g=5000,r=15000,bits=106,symbol_zero={4}e,symbol_one={4}8,unique,repeats>=5,get=CODE:@1:{24}:%06x" -r g002_433.92M_250k.cu8

model     : WaterLeak    count     : 7             num_rows  : 10            len       : 25            data      : 8780328       DATA      : 0f0065

You might need to tweak the gap (g=) value, and maybe tweak or remove the ,unique,repeats>=5 if this isn't sensitive enough in your environment.

The CODE "get" value assumes an initial sync bit instead of a trailing bit in case that seems more useful for you.

Here's a version formatted for rtl_433.conf:

decoder {
    name=MyWaterLeak,
    modulation=OOK_PCM,
    short=512,
    long=512,
    gap=5000,
    reset=15000,
    bits=106,
    symbol_zero={4}e,
    symbol_one={4}8,
    unique,
    repeats>=5,
    get=CODE:@1:{24}:%06x
}

[gdt]

Is someone going to make a PR to put that flex decoder in the sources? What do we need to do to close this issue?