Smart Implant + sensor DS18B20 BUG

[fastvd 7]

I describe the problem. There is a lot of HC3 in many of my clients. In all these objects, I read the temperature of the warm floor by Smart Implant
FGBS-222 + sensor DS18B20. Everything here is ok, except that sometimes the light turns off or just briefly turns off and then turns on - the temperature sensor shows 0 degrees! Just a glitch in the system! At first I tried to do Soft reconfiguration - nothing helped, then Complete reconfiguration - it doesn't help either!

I found a solution using the sorting method: either completely remove the device from the system and add it again, or simply turn it off for a few minutes until the system identifies it as unavailable and then turn it on! The first option is very inconvenient, because you need to completely reconfigure, all the IDs of the child devices change, which leads to changes in all scenes, etc. The second method with the shutdown is simpler, but also not comfortable. as the system must work!

Therefore, I single it out as EXACTLY A BUG\PROBLEM in the fibaro system itself or in the Smart Implant device!

Anyone else having problems? Can someone find a 3rd solution to the problem?

Is there a way to programmatically reset the device?

Please login or register to see this attachment.

Edited December 11, 2022 by fastvd

[Fabir 73]

@fastvdHi, I'm also using quite a few smart implants with up to 4 sensors DS18B20 on each SI. I haven't seen the same issue.

 

In your descreption, what do you mean by the light turns off? There is a power outage for the house? If this is the case I recomend a USV device.

 

[fastvd 7]

Yes, no light = no electricity! I'm from Ukraine and we currently have a big shortage of electricity due to the war! The device you recommend is all ok, but it is not a solution to the problem, but simply a reduction in the number of such triggers! And let us solve the problem itself! Maybe it didn't pop up for you, because you don't have many such devices! My friends and I put HC3 in a total of about 10 controllers and about 30 implant + temperature sensor connections! Everyone has different power supplies, different conditions of use, but everyone has this problem from time to time!

In my home, where almost all implants are on guaranteed power supply, this can happen once a year, and where there is no guaranteed power supply, it can happen once a month! And it shouldn't be like that, it's definitely a system glitch!

[petergebruers 1,268]

As you say, this is probably an issue users simply haven't noticed. It could be a HC firmware bug, Smart Implant firmware or hardware problem. I also would like to offer you another possible explanation.

 

Do you use genuine DS18B20 or clones?

 

You can tell with an Arduino sketch, and also this: If you bought a "waterproof" version (with cable and metal sleeve) then it almost 100% certainly is a clone.

 

The main reason to talk about this is because we know about at least these issues when using clones:

 

• Some seemed reasonably accurate, but there were outliers, even at room temperature. Some seemed to be out of spec (2 °C) near freezing point.
• Some seemed "fixed" to doing 12 bit conversions only.
• Some did not work 2-Wire aka phantom powered aka parasite powered.

 

 

1 hour ago, fastvd said:

in a total of about 10 controllers and about 30 implant + temperature sensor connections!

Do you see any pattern? For example... clones versus genuine sensors?

 

I do not own a smart implant, I do own a bunch of DS18B20 on other Fibaro and non Fibaro devices. I never had a DS18B20 "drop off" like that, I never had to restart a device.

Edited December 12, 2022 by petergebruers

[fastvd 7]

oh well...i did my test too!
And I found the reason!
and it is not in the fact that some of the sensors are original or not!
Let me remind you: I have about 10 objects that I service, and without exception, all of them have a pair of implant + temperature sensors or, in most cases, an implant + several temperature sensors.
I bought them from about 3-4 suppliers, all of them were launched into the system the first time, whether it was 1 implant + 1 sensor or 1 implant + 6 sensors - everything always started well!

and I remind you: the problem was both with 1 implant + 1 sensor and with 1 implant + 3-4-5-6 sensors... I did not find that template.
How I conducted my test and made a conclusion for MYSELF! So:

1) I have about 10 objects, near them 40 implants, with 1-2-3-4-5-6 sensors.
2) What I thought before the test was a problem due to surges in the 220 V network

3) that's why I ordered a new implant, added HC3 to my home, bought 10 new 100% original sensors not in hermetic flasks, where 100% says dallas
4) I had a pair of implants + 1 sensor right on the street at home, and it was not on the guaranteed power supply and it had problems from time to time

5) first I removed my new implant with the first sensor from aliexpress that I got my hands on...
6) I will say in advance - I didn’t torture him, I didn’t abuse him, I overloaded him 150 times at the time of loading, and the moment of almost full loading, at the moment when I forcibly set the polling command with HC3, during the reboot I turned it on 5-7 times and turned off 12v - there were no problems!

7) so I went outside and removed my home 100% sensor that had problems from time to time!
8) BUT...I wanted to completely make fun of the system and just vandalized online and took nippers and bit off 3 wires from the sensor that connect it to the implant! JUST OBVIOUSLY!

9) I removed the old link implant+100% working sensor from the system, connected my cut sensor, which 100% had problems and started doing the same tests as in point 6

10) yes, as in point 6 - my 100% problem sensor behaved ABSOLUTELY normal!!
11) at this moment, the thought comes to me to check, and what is now showing my street implant, on which I SUDDENLY cut off the sensor while online!!! And what do you think!!!??? I finally found my 0 !!!

12) I realized that the culprit is not the sensor, but the implant+sensor connection itself...
13) I had an implant + sensor on a simple coil clamped in WAGO connectors, which are often used in electronics.
14) Then I take similar actions on my stand, open the contact on one of the three cables of the sensor - I have 0 !!!
15) I connect online on the go - I have 0
16) I overload the implant with connected cables - everything is ok.
17) Therefore, for myself, I concluded that the culprit is not FIBARO, not the IMPLANT, not the SENSOR itself, but precisely the quality of this connection!

 

PS:
1) For the following week, I will try to correct my theory of bad eating at 2 objects where there were constant problems. All I plan to do is to solder all the contacts!
2) Since this process is not fast, in about a month I will already have +- a working, practical result

3) I forgot to say that during the test I also changed different power supplies! including those that I use constantly in objects! The problem is not in this or in the guaranteed supply!

 

 

Edited February 4, 2023 by fastvd

[inastalz 0]

fastvd@

 

So are you only using solid soldered power to the sensors? If the temperature sensor shows 0°, the only way to wake up the sensor is to completely reconfigure the device?

[Nomader 1]

this just happened to me as well on a power outage.. after the power was on again i had "0" in the sensor value. tried healing, re initializing (i am using openhab) but nothing worked.. only removing the implant from power for a few minutes and reconnecting fixed it.. very inconvenient

[Tech-00 0]

I had the same issue with sensor value of "0" that seems to have started after power outage. Managed to get it working by removing power a few times. Was ok until next power outage a few days later. After this I was not able to get it working by removing power. After a few months I reconnected the wires from sensors to the implant and then I got sensor readings back! A bit strange that it first occured after a power outage when the fault seems to have been bad sensor connections.