briankitt commented 4 months ago

I'm using zigbee-moisture-sensor-v2 with this sensor: https://www.amazon.com/dp/B0CLJ951YS

It works perfectly, well, almost.

The water sensor doesn't appear to broadcast often enough to satisfy the drivers minimum broadcast requirements.

All day long, the sensors go offline, then online, then offline, then online, etc.

It's done cyclic, so that indicates to me that the water sensors are working, just not 'often enough'.

Is there a way to tweak the driver to increase the timeout before it decides it's 'offline'?

I have 18 sensors, and get notified constantly all day (and all night) that they are going online and offline.

Mariano-Github commented 4 months ago

The driver configures the device with the default smartthings values:

300 sec for cluster 500, attribute 0002

21600 sec for battery, cluster 0001

The hub marks as offline when no message is received from those clusters for approximately 2:30 hours.

If it appears offline intermittently, it means that one of the clusters has not been configured correctly in the installation and does not send periodic messages or the messages do not reach the hub due to poor quality or signal strength.

These devices are usually placed in inaccessible places, under metal appliances, which attenuate the 2.4 GHz zigbee signal.

Try reinstalling the devices near the hub and placing repeaters, plugs, etc. for those that are further away from the hub.

You can use zwave sensors, which have lower frequencies, 800 MHz, and have less noise and have a longer range.

There are some tuya TS0207 models, which do not configure cluster 500 correctly and disconnect and reconnect when they receive the battery message, cluster 0001, which report every 6 hours

briankitt commented 4 months ago

It's not a communication issue, I can guarantee that.

1) I drop them in water, they immediately become 'active' in smart things,

and report wet.

2) I set 1/2 dozen around the hub (within inches) and they still cycled

offline/online/offline/online all day and night.

3) They cycle at regular intervals that you could set your clock to. To

me, it looks like they are reporting every 4 hours on the nose. It's very regular.

Is there any way I can change the driver or some setting to wait 5 hours before reporting it as 'offline'?

That's all I need to do to fix it.

On Mon, Apr 15, 2024 at 4:18 PM Mariano @.***> wrote:

The driver configures the device with the default smartthings values:

300 sec for cluster 500, attribute 0002

21609 sec for battery, cluster 0001

The hub marks as offline when no message is received from those clusters for approximately 2:30 hours.

If it appears offline intermittently, it means that one of the clusters has not been configured correctly in the installation and does not send periodic messages or the messages do not reach the hub due to poor quality or signal strength.

These devices are usually placed in inaccessible places, under metal appliances, which attenuate the 2.4 GHz zigbee signal.

Try reinstalling the devices near the hub and placing repeaters, plugs, etc. for those that are further away from the hub.

You can use zwave sensors, which have lower frequencies, 800 MHz, and have less noise and have a longer range.

There are some tuya TS0207 models, which do not configure cluster 500 correctly and disconnect and reconnect when they receive the battery message, cluster 0001, which report every 6 hours

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2057826450, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6AUM5GPZWZEUANBUA3Y5Q72VAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANJXHAZDMNBVGA . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

The time to display offline is strictly controlled by the hub, not by the driver.

The 4 hours interval fits with what I told you, it is offline after about 2.30 hours, 4 hours later the battery message arrives to hub and it goes online for another 2.30 hours and 4 hours later another battery message arrives again ...

It is a problem with some Tuya TS0207 models, they do not send periodic 300 sec reports from cluster 500.

I could to configure battery reports every 1 hour instead of 6 hours, if can identify with model and manufacturer, which tuya devices have the problem, but if I do it to all of devices indiscriminately it would mean excessive and unnecessary battery consumption on the devices that work correctly.

briankitt commented 4 months ago

What about giving us an option, where we can set that interval? We could go into settings, and set it on a device by device basis.

Here are screen prints of what I see: [image: image.png]

[image: image.png] [image: image.png] [image: image.png] [image: image.png]

On Mon, Apr 15, 2024 at 5:59 PM Mariano @.***> wrote:

The time to display offline is strictly controlled by the hub, not by the driver.

The 4 hours interval fits with what I told you, it is offline after about 2.30 hours, 4 hours later the battery message arrives to hub and it goes online for another 2.30 hours and 4 hours later another battery message arrives again ...

It is a problem with some Tuya TS0207 models, they do not send periodic 300 sec reports from cluster 500.

I could to configure battery reports every 1 hour instead of 6 hours, if can identify with model and manufacturer, which tuya devices have the problem, but if I do it to all of devices indiscriminately it would mean excessive and unnecessary battery consumption on the devices that work correctly.

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2057950272, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6DYCHAFEK33SSL5MD3Y5RLTJAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANJXHE2TAMRXGI . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

These devices are normally asleep, only accepting configuration changes during installation or when they are awake.

If you tell me the model and manufacturer of the devices that disconnect, I can change the configuration on those devices in the installation, reinstallation

briankitt commented 4 months ago

Did you not get my prior response? I sent the make, manufacturer, and screen prints from ST? https://www.amazon.com/dp/B0CLJ951YS

On Wed, Apr 17, 2024 at 5:34 AM Mariano @.***> wrote:

These devices are normally asleep, only accepting configuration changes during installation or when they are awake.

If you tell me the model and manufacturer of the devices that disconnect, I can change the configuration on those devices in the installation, reinstallation

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2060948028, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6DS6TVIRVNZ4MAPZMTY5ZFZ5AVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANRQHE2DQMBSHA . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

I can not see the puctures posted

briankitt commented 4 months ago

And then the device is here on Amazon https://www.amazon.com/dp/B0CLJ951YS

Mariano-Github commented 4 months ago

Modified configuration for battery report interval 3600 sec.

To device configure with new values you need instal or reinstall devices.

To reinstall devices without uninstall it:

Install this driver version or wait to driver updated automatically
Put device in pairing mode
In app, add device with search nearby option
Device will re install and configured with lost any routine or scenes

───────────────────────────────────────────────────
 Name         Zigbee Moisture Sensor Mc
 Version      2024-04-19T20:08:04.635029836
───────────────────────────────────────────────────

To confirm that the battery is reporting every 1 hour you can see the battery time graphs in the app

briankitt commented 4 months ago

Well bummer. I uploaded the new driver to my SmartThings I deleted every water sensor I have (18 of them) I then paired all 18 back. I then verified that they were all using the new 2024-04-19 driver, which they all did automatically. Within 2 hours they were going offline due to no response.

On Fri, Apr 19, 2024 at 3:14 PM Mariano @.***> wrote:

Modified configuration for battery report interval 3600 sec.

To device configure with new values you need instal or reinstall devices.

To reinstall devices without uninstall it:

Install this driver version or wait to driver updated automatically

Put device in pairing mode

In app, add device with search nearby option

Device will re install and configured with lost any routine or scenes

─────────────────────────────────────────────────── Name Zigbee Moisture Sensor Mc Version 2024-04-19T20:08:04.635029836 ───────────────────────────────────────────────────

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2067223212, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6G74WF4DJ4SHC5VYATY6F3ITAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDANRXGIZDGMRRGI . You are receiving this because you authored the thread.Message ID: @.***>

briankitt commented 4 months ago

What's next? They are still cycling offline and online?

Mariano-Github commented 4 months ago

I have no idea what is happening to your devices.

These devices are cheap and work well with your platform, but on other platforms they can be a pain.

If you send me the logs with the device pairing CLI and I can see what messages the device sends when it is configured, then I could analyze what happens, but I can't guarantee anything with these devices.

briankitt commented 4 months ago

I don't know where these logs are, I looked for them, but can't find them. Can you point me in the right direction? Thanks

On Sat, Apr 27, 2024 at 2:38 AM Mariano @.***> wrote:

I have no idea what is happening to your devices.

These devices are cheap and work well with your platform, but on other platforms they can be a pain.

If you send me the logs with the device pairing CLI and I can see what messages the device sends when it is configured, then I could analyze what happens, but I can't guarantee anything with these devices.

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2080401935, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6DMEMMKSR3TTMLM4NLY7NIYHAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOBQGQYDCOJTGU . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

This is a tutorial to install cli in windows and capture logs

https://community.smartthings.com/t/edge-driver-or-device-handeler-for-a-tuya-zemismart-zigbee-plug-with-energy-mesurment-capabilities/245383/18

briankitt commented 4 months ago

This is VERY informative, and I've learned an incredible amount about my devices. 1 thing, for example, is that devices are reporting far more frequently then I guessed.
I apologize, I should have done this first, this would have given you so much more information about my problem.

First off, check out this screen print. Look at the devices that are offline. I intentionally captured my time, so you could compare my time (Central Daylight Time) to the Smart Things debug log time stamps.

Now, look at the log. Even though these devices are showing 'offline', they are reporting on a regular basis? So how does that happen? They are successfully reporting, yet, SmartThings shows they are offline? That seems odd. But clearly, the problem is not in the reporting of the status?

Here is 1 hour worth of log. Let me know if you need more. Zigbee.log

Mariano-Github commented 4 months ago

I have looked at the entire log, taking note of the most relevant of the 18 sensors:

starts at 18:17:10 and ends at 01:31:11
At 18:37:52 "Washing Maching" sensor was removed
At 18:55:12 It is reinstalled and configured.
Battery configuration 3600 sec. Device response: Ok
IASzone configuration, cluster 0500 (dry or wey) 2100 sec. Device response: Ok
The battery and IASZone attributes are read and the capabilities are initialized with the values received: 93% and Dry.
Only one sensor (Master small sink), during the log, spontaneously reports cluster 0500, (dry) only once at 00:31:11.
No other sensor spontaneously reports either cluster 0500, (dry or wet) or cluster 0001, (Batt).
All the logs you see are the activation of the Health Check function to read cluster 0500, attribute 0002, because the messages are not received.
the device does not respond to readings because it is asleep.
Result, they are all offline except the "Master small Sink" when the report was received at 00:31:11.

The problem is not very clear:

The device responds that it accepts the configuration, but no periodic reports are received from the device, neither battery nor IASZone.
When a spontaneous reporter message is received from a device, the Rssi intensity: -86 dbm is quite low.
It could be that the messages are issued and do not reach the Hub, although it is strange that almost none arrive.
It could be that the devices do not send messages correctly, but it is rare for them all to fail.
You could try placing two sensors near the hub to see if they stay online and send periodic reports.
I can't think of anything else to try.

I attach a photo of the summary of the logs PXL_20240502_095110587

briankitt commented 4 months ago

I intentionally excluded that one, and then re-included it, for purposes of logging.

Would more logging help? The one that says 'laundry water inlet' is less than 5 feet from the hub? If that's too far, the hub has a problem. And that one is currently offline.

I could do a whole day of logging if you want.

On Thu, May 2, 2024 at 4:52 AM Mariano @.***> wrote:

I have looked at the entire log, taking note of the most relevant of the 18 sensors:

-

starts at 18:17:10 and ends at 01:31:11

At 18:37:52 "Washing Maching" sensor was removed

At 18:55:12 It is reinstalled and configured.

Battery configuration 3600 sec. Device response: Ok

IASzone configuration, cluster 0500 (dry or wey) 2100 sec. Device response: Ok

The battery and IASZone attributes are read and the capabilities are initialized with the values received: 93% and Dry.

Only one sensor (Master small sink), during the log, spontaneously reports cluster 0500, (dry) only once at 00:31:11.

No other sensor spontaneously reports either cluster 0500, (dry or wet) or cluster 0001, (Batt).

All the logs you see are the activation of the Health Check function to read cluster 0500, attribute 0002, because the messages are not received.

the device does not respond to readings because it is asleep.

Result, they are all offline except the "Master small Sink" when the report was received at 00:31:11.

The problem is not very clear:

The device responds that it accepts the configuration, but no periodic reports are received from the device, neither battery nor IASZone.

When a spontaneous reporter message is received from a device, the Rssi intensity: -86 dbm is quite low.

It could be that the messages are issued and do not reach the Hub, although it is strange that almost none arrive.

It could be that the devices do not send messages correctly, but it is rare for them all to fail.

You could try placing two sensors near the hub to see if they stay online and send periodic reports.

I can't think of anything else to try.

I attach a photo of the summary of the logs PXL_20240501_170240915.jpg (view on web) https://github.com/Mariano-Github/Edge-Drivers-Beta/assets/74271621/40d408ef-8d9d-4787-a319-2285ccd7cff6

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2090055059, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6CTP3PMBJUEFJDB5C3ZAIEGLAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJQGA2TKMBVHE . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

What channel do you have the router's WiFi on and what zigbee channel does the hub have? they could be interfering

briankitt commented 4 months ago

I have a number of electronic devices, kind of all together. I have a Netgear ORBI 2.4 GHZ at 11, 5 GHZ at 40 I have a Century link router, that broadcasts 2.4 GHZ at 1, and 5 GHZ at 161 I have an AccuRite weather station that is broadcasting, but I don't know the channel it uses. I'm researching, I just don't have time to research now as I am at work. I have an X-Sense smoke alarm, but don't know how to find the channel it uses. I'm researching I have 3 Google Nest Displays, but 99% sure, they only use WiFi, no additional radio. I have a SmartThings hub, but I need to find out what channel it uses. I just don't know where to look, I'm researching.

On Thu, May 2, 2024 at 8:17 AM Mariano @.***> wrote:

What channel do you have the router's WiFi on and what zigbee channel does the hub have? they could be interfering

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2090478636, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6DEVY6JZGN5Q64NLC3ZAI4GVAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJQGQ3TQNRTGY . You are receiving this because you authored the thread.Message ID: @.***>

briankitt commented 4 months ago

It appears the zigbee channel is not changeable and cannot be displayed, at least, that's what I'm finding?

On Thu, May 2, 2024 at 8:17 AM Mariano @.***> wrote:

What channel do you have the router's WiFi on and what zigbee channel does the hub have? they could be interfering

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2090478636, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6DEVY6JZGN5Q64NLC3ZAI4GVAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJQGQ3TQNRTGY . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

Zigbee channel can see and change in my.smartthings.com advanced users.

Do not use channel 25

Screenshot_20240502-200641

Wifi 2.4 ghz channel can see and change in your router settings

Mariano-Github commented 4 months ago

This is a publish of interference zigbee and wifi channels

https://www.metageek.com/training/resources/zigbee-wifi-coexistence.html

The fact that the router and the hub are very close can increase interference

briankitt commented 4 months ago

Oh, got it. sorry. Per your suggestion, I changed it to 25. I'll update you after 24 hours, that should be enough time for everything to adjust. I also changed my WiFi for my Orbi down to 2, since 11 can interfere with 25 based on the document.

Mariano-Github commented 4 months ago

Some olders devices can have problems with channel 25, but new devices i think work fine

briankitt commented 4 months ago

Over the last few days, I've tried a number of channels. Each time, I remove and re-add all 18 devices. Each time, they all go offline within a few hours.

Any ideas of next steps? I know you have put forth a great effort to try to understand what is going on. I hate to get rid of these. But I really want water sensors to watch for leaks.

On Fri, May 3, 2024 at 5:38 AM Mariano @.***> wrote:

Some olders devices can have problems with channel 25, but new devices i think work fine

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2092744365, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6ELSCQLTLLFWYB5E2LZANSJHAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJSG42DIMZWGU . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github commented 4 months ago

I don't understand anything, but everything indicates that this device is not working well.

Try a different zigbee detector to see if it is a problem with your zigbee network or not.

These are more expensive, but they do work well.

https://www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://www.amazon.com/-/es/Aeotec-SmartThings-alimentado-concentrador-inteligente/dp/B095TR9NYR&ved=2ahUKEwjNtuXhr_mFAxWExgIHHZkfBpwQFnoECBIQAQ&usg=AOvVaw0hwjdbyhc9pUR01NGk_dLK

briankitt commented 4 months ago

Wow, 18 at nearly 40$ a crack, we are looking at a purchase of around 700$

I'm going to try a different hub.

Mariano-Github commented 4 months ago

I meant buying 1 to test and see if it is a problem with the zigbee network or the devices.

Then you return it and with the conclusions you draw you make a decision.

There may be more models, you can ask in the community, but the TS0207 cause many problems to many users

briankitt commented 4 months ago

Thanks! Go ahead and close this ticket, we beat this one to death. I can't thank you enough for all of your hard work and advice on this.

On Mon, May 6, 2024 at 1:25 PM Mariano @.***> wrote:

I meant buying 1 to test and see if it is a problem with the zigbee network or the devices.

Then you return it and with the conclusions you draw you make a decision.

There may be more models, you can ask in the community, but the TS0207 cause many problems to many users

— Reply to this email directly, view it on GitHub https://github.com/Mariano-Github/Edge-Drivers-Beta/issues/116#issuecomment-2096650363, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAXDH6A6II3PKNI3IZ3CIC3ZA7DKVAVCNFSM6AAAAABGH3NCZKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAOJWGY2TAMZWGM . You are receiving this because you authored the thread.Message ID: @.***>

Mariano-Github / Edge-Drivers-Beta

zigbee-moisture-sensor-v2 - need to change timing for heartbeat #116

1) I drop them in water, they immediately become 'active' in smart things,

2) I set 1/2 dozen around the hub (within inches) and they still cycled

3) They cycle at regular intervals that you could set your clock to. To

starts at 18:17:10 and ends at 01:31:11

At 18:37:52 "Washing Maching" sensor was removed

At 18:55:12 It is reinstalled and configured.

Battery configuration 3600 sec. Device response: Ok

IASzone configuration, cluster 0500 (dry or wey) 2100 sec. Device response: Ok

The battery and IASZone attributes are read and the capabilities are initialized with the values received: 93% and Dry.

Only one sensor (Master small sink), during the log, spontaneously reports cluster 0500, (dry) only once at 00:31:11.

No other sensor spontaneously reports either cluster 0500, (dry or wet) or cluster 0001, (Batt).

All the logs you see are the activation of the Health Check function to read cluster 0500, attribute 0002, because the messages are not received.

the device does not respond to readings because it is asleep.