FGBS-001 Universal Binary Sensor - Freezing/stops reading temps

[kelvincoffey 4]

I have a FGBS-001 that I have been using (or attempting to...) for 1.5 years.

 

The unit has 3 temp probes connected, and source is a 200W switch mode power supply (which also supplies the RGBW dimmer with 40W of lights and capacity to add more RGBW's...)

 

I am seeing the temperature flat-line from the sensor after a few hours, or days, or weeks, depending on luck.  The only way to reset the temp feed is to power cycle the UBS.  I can still communicate with the UBS via the HC2, changing config and the like, so it seems to be the digital controller in the UBS its freaking out seperate to the zwave controller. (this is an assumption, and I have no idea how the internals of the UBS works....)

 

I have gone from a connection screw block from device to probes, to soldered connections (thinking dirty data from the connection block) and saw no improvement.  I have done the factory resets etc....

 

Anyone else had these issues? Any resolutions? Any ideas?   

 

I have reached out to Fibaro email support and they proved to be consistent with my previous experience of their ability to support their products.

[Hologram 0]

Second this.......

I'm running a VeraEdge and have several universal sensors that I try to use.

I have the exact same experience as you, or actually worse...

One or more temp sensors will stop working after a random time. The binary sensor itself seems to work fine, but not even powercycle will recover the lost temp sensor.

The only thing that helps is to completely reconfigure the device.. Thus resetting my sensor id's etc... 

A major inconvenience you may say..

I also assumed bad connections first, but now I have given that up and just reconfigure whenever an "important" sensor dies......

[kelvincoffey 4]

Mine is connected to bathroom floor heater.

 

so I get up in morning and can't walk on the floor without shoes as it's so hot it burns my feet!

But according to Fibaro it's a cool 16 degrees... since 1.04am

[Hologram 0]

I got one that that measures the outside temperature. The other three sensors that are connected are really just measuring the ambient temperature in the cabinet at the moment (as I cant bother using them for anything else).

As of now one of the "ambient" sensors are dead, and has been so for a couple of months.

 

A few weeks ago we finally finished installing a water based heating system, so I have two 1000L tanks in the basement that I monitor using two UBS's.

Four sensors in the main tank, three in the secondary.

I then control pumps based on the temperature readings to balance them.

By now I have already reconfigured twice, and now the bottom sensor in the main tank is dead again...

One interesting thing to note though, the second time things went to *abuse* there was like 4-5 sensors that died almost simultaneously!

Across two different UBS's!

So that would indicate that the issue can trigger due to some outside factor, noise, supply power irregularities etc...

Both the UBS's are connected to the same 9V power supply. I might try to add a capacitor there to filter the voltage..

 

 

[Jamie mccrostie 82]

4 hours ago, Hologram said:

I got one that that measures the outside temperature. The other three sensors that are connected are really just measuring the ambient temperature in the cabinet at the moment (as I cant bother using them for anything else).

As of now one of the "ambient" sensors are dead, and has been so for a couple of months.

 

A few weeks ago we finally finished installing a water based heating system, so I have two 1000L tanks in the basement that I monitor using two UBS's.

Four sensors in the main tank, three in the secondary.

I then control pumps based on the temperature readings to balance them.

By now I have already reconfigured twice, and now the bottom sensor in the main tank is dead again...

One interesting thing to note though, the second time things went to *abuse* there was like 4-5 sensors that died almost simultaneously!

Across two different UBS's!

So that would indicate that the issue can trigger due to some outside factor, noise, supply power irregularities etc...

Both the UBS's are connected to the same 9V power supply. I might try to add a capacitor there to filter the voltage..

 

 

It could tbe power supply, that might be a good place to start.

[Hologram 0]

As a test I now replaced the fairly old 9V supply with an adjustable laptop supply I had laying around from work.

Should be a fairly decent and reliable piece of equipment so if that's not sufficient then I don't know what is..

I'll give it a few weeks and report back.

[lyodf 3]

Hi all,

it is really strange what you are reporting on the operation of your UBS modules. Personally i never had any problems with these modules. I have 2 FGBS-001 with 4 temperature sensors each, and with them i read the different temperatures of the rooms for my heating system.
For their connection i used a standard three-conductor telephone cable and all junctions were welded with hot pond. As a module power supply i use a stabilized 12V line provided by the alarm control panel, so absolutely no swinging or sine waveforms.

 

I would like to pay close attention to these two latest factors: junction type and power supply.

Enrico

[Hologram 0]

A short update here:

So I replaced the power supply on may 30th, and since then it has really been running fairly well.

I did not reconfigure them at that point, so one sensor was already dead, but the remaining six has worked flawlessly for roughly one and a half week.

(average mtbf before this was half a week or so I'd estimate..)

 

Yesterday however I had an electrician in, and he had to take down the main power for a minute or so.

This then killed three sensors.... two on one UBS, and one on the other....

So all in all I'd say that I find it beyond all doubt that it is related to the power, but ofc it's incredibly annoying that even a decent power supply is unable to take the UBS's through a powerloss without errors...

 

It does also seem like the behavior varies from controller to controller, like I said initially I have to reconfigure the UBS's, basically an exclude-include process, to get the dead sensors back up.

According to @kelvincoffey, with the HC2, you "only" need to powercycle them..?

[kelvincoffey 4]

@Hologram  in 2 years, none of mine have died. a power cycle fixes them.

 

When you say it "killed" your sensors. What do you mean? Throw it in the bin and replace? or reprogram,?

 

My Sensor are in my screed bed under tiles. So if they die, I am in a lot of trouble....

Edited June 11, 2017 by kelvincoffey

[Hologram 0]

Sorry, that was a very imprecise statement indeed... They are not broken, they just no longer update...

So in terms of practical use they are dead.. I guess dormant would be a better word.

 

After a reconfiguration of the UBS (exclude-include) they are back in business, power-cycle is not enough in my case..

[Half pipe 0]

On 5/29/2017 at 8:27 PM, Hologram said:

Second this.......

I'm running a VeraEdge and have several universal sensors that I try to use.

I have the exact same experience as you, or actually worse...

One or more temp sensors will stop working after a random time. The binary sensor itself seems to work fine, but not even powercycle will recover the lost temp sensor.

The only thing that helps is to completely reconfigure the device.. Thus resetting my sensor id's etc... 

A major inconvenience you may say..

I also assumed bad connections first, but now I have given that up and just reconfigure whenever an "important" sensor dies......

Hologram, I am trying to do the same - control heating pumps based on temperatures in a buffer tank. Do you mind sharing your experiences? Anyhow, I use athom/homey platform. At this point I can't even include the binary sensor. I already killed one, so it's the second attempt :(. Does anybody know how to test it with a multimeter?

[Kifo 0]

I have the same issue with 3 different universal sensors. The time to failure varies between ~1 week and 15 months. Sometimes only one temp sensor freezes, sometimes all at once. The only way to get them back is to exclude and include them in my z-wave network (with all the related hassle regarding IDs etc.).

I already started replacing some of those sensors with more reliable ones from another company, at least in some critical applications.

 

My guess for the reason: Electromagnetic pulses, either over the power supply or "over the air" through the sensor wires.

[TomaszK 0]

Guys,

 

I have been having the same problem for a couple of years already. I tried every advice I could find online (shorter wires, better power supply, I even go more UBS devices to have my temperature sensors connected as single probe per device)

Nothing works, every now and then I have a problem with "hanging" temperature on one of the sensors.

 

My configuration is:

• UBS with a single temp probe for a water container
• UBS with a single temp probe for a heating boiler
• 4 x UBS with two temp probes each for room temperatures
• all controlled by VeraLite controller

Every now and then a probe fails and stops reporting a temperature - could be even just a single one from the pair of two connected to the same UBS.

The solution for me was always to click on "reconfigure device" in Vera and click the button on the UBS multiple times to confirm the reconfiguration. This process deletes the temperature sensor devices from Vera and adds new ones with new IDs.

 

This works if I am on site and can reconfigure quickly. But very often I am 150 km from the system and I am not able to physically press the button on the UBS.

 

Has any one succeeded to reconfigure the hanging UBS remotely?

 

I am assuming that this must be possible and here is the reason why: recently I updated the software on my VeraLite. Before the update I had a couple of UBS sensors hanging. After the update (and resetting the whole Vera system) all the devices were live and reporting the temperature.

This happened without me touching any buttons on the UBS itself - so I am assuming the same thing must be possible to do remotely.

 

Anybody has an idea how to achieve it?

 

This would be very helpful as sometimes my water tank temp sensor hangs and this is a very important temperature for me. The water is automatically heated up by a solar installation and if it's not used it can actually over heat and start boiling or even explode. So I built a safety system using these Fibraro UBS temperature sensors and Fibaro switches that control the water pumps (basically if the water gets too hot it is spread out to an idle heating boiler or even to radiators in the house). I did not know then that these sensors are so unreliable.

 

Tomasz

On 12/13/2018 at 6:31 PM, Kifo said:

I already started replacing some of those sensors with more reliable ones from another company, at least in some critical applications.

 

@Kifo

Could you share what other temperature sensors are you using? Would they work with Vera controller over Z-wave standard?

[petergebruers 1,261]

Could be Vera, UBS firmware or DS18B20. If you own a Arduino or compatible device, you can run a sketch to find out if your sensors are genuine