1 wire net dies "Problem solved"

[tomas_hora]

@zorro_66 Any possibility to change the UTP to STP/FTP from the cabinet at least to the nearest hub?

Btw. nice installation :) Would be great if you could share with us a few lines describing the project and a few picture.

[zorro_66]

I use no hub I use the daisy chain
there can be problems
I can certainly tell you a little about my system and put some more

[zorro_66]

Now it has happened again
did not help with shielded cable
I'll test to update to the latest UniPi-Mervis-OS 2016-12-06

[tomas_hora]

@zorro_66 Please test and report, any possibility to use shielded cable between the sensors?

[wizro_pl]

@zorro_66 did you connect shield to the ground ? it is necessary.

[zorro_66]

Yes as in the picture

[zorro_66]

Now it has worked in two days
My second Unipi do not have these problems
where I have 20 pieces of sensors connected in a star network with no shielded cable.
Does it help to update the software in Rpi.
There seems to be some software errors.
Just got home 2 1-Wire hub from Unipi.
Can i connect with a standard RJ45 patch cable

[zorro_66]

Now, I have rebuilt my entire 1-wire network and used Unipi hub instead
so now we will see if the network stops die

[zorro_66]

My 1-wire is incredibly unstable
network dies at irregular intervals, showing unreasonable temperatures.
I have rebuilt my net for spec from Unipi wiki shielded cat5 cable and hubs.
But no difference 1-wire network die anyways.
So it can not be anything wrong with the network itself, but the fault must be in Unipi alt. software.
Grateful for help .

[tomas_hora]

@zorro_66 Can you try enabling the CRC check at the sensors? Just check the box in the sesors in the Scratchpad group. You might need to make the sensor as a custom and backup your project before playing with it

[zorro_66]

I do not understand what you mean now

[tomas_hora]

@zorro_66

[zorro_66]

aha tanks:thumbsup_tone3:

[zorro_66]

Problem solved, I finally found a bad solder
at the VCC pin caused power loss in my 1-wire network

[tomas_hora]

We are sorry for that @zorro_66!

But it is really hard to find out this kind of issue during our test but we will update our testing procedures... Thanks for the info.