Sensor contact lost

[nking]

I get this error "Lost Sensor Contact!!!"about once a day and wonder what the Cumulus logic is for deciding that there is an error (I've done a quick search without success). I think I read at sometime that the error warning can be turned off but I may have imagined this. However, depending on the logic - what would be a nice solution would be to have a user variable to state how long/number of "No Contacts" before reporting it as an error. I would add that this would be a nice to have feature if it doesn't involve much effort.

[steve]

I get this error "Lost Sensor Contact!!!"about once a day and wonder what the Cumulus logic is for deciding that there is an error

It's part of the data; the console sets a flag.

I think I read at sometime that the error warning can be turned off but I may have imagined this.

I think you imagined it

what would be a nice solution would be to have a user variable to state how long/number of "No Contacts" before reporting it as an error. I would add that this would be a nice to have feature if it doesn't involve much effort.

That would be fairly easy to do; the "number" version, that is. You would have to take into account the fact that Cumulus reads the data every 10 seconds but the data is only updated every 48 seconds. How many consecutive error messages do you typically get (in the diags file) currently?

[AllyCat]

Hi Neil,

I believe that Cumulus simply repeats the "error" message from the station. However, AFAIK, the station only reports this after 6 - 8 transmission messages are "lost", so it should only occur if the system is really "struggling".

Therefore, an extra delay in Cumulus may not help (much); the proper solution is to try to improve the wireless connection (maybe easier said than done).

Cheers, Alan.

[Charlie]

the proper solution is to try to improve the wireless connection (maybe easier said than done).

If it's a recent phenomenon, the solution is most likely to put in new batteries.
Do NOT use rechargable ni-cad or ni-mh.

If it's been going on almost since the first installation, then you are at the hairy edge of range and need to address that as Alan suggests. A search of the forum will bring up lots of advice.

Changing the number of times you get the message before reporting it is treating symptoms not the disease. If the station is set up properly you should never see this message until several years after the first install.

[nking]

Thanks for the replies.

Steve - here is a filtered list of the lost contacts

Line 3081: 25/08/2012 03:29:37.280 : Lost sensor contact!!!
Line 3082: 25/08/2012 03:29:47.288 : Lost sensor contact!!!
Line 3083: 25/08/2012 03:29:57.296 : Lost sensor contact!!!
Line 3567: 26/08/2012 18:24:16.782 : Lost sensor contact!!!
Line 3950: 27/08/2012 21:39:23.874 : Lost sensor contact!!!
Line 4363: 29/08/2012 05:00:42.087 : Lost sensor contact!!!
Line 4809: 30/08/2012 16:06:43.050 : Lost sensor contact!!!
Line 4873: 30/08/2012 21:20:17.750 : Lost sensor contact!!!
Line 6204: 02/09/2012 11:16:54.767 : Lost sensor contact!!!
Line 6497: 03/09/2012 08:35:14.940 : Lost sensor contact!!!
Line 6539: 03/09/2012 12:03:14.702 : Lost sensor contact!!!
Line 6908: 04/09/2012 17:03:57.538 : Lost sensor contact!!!
Line 7460: 05/09/2012 04:57:49.764 : Lost sensor contact!!!
Line 7545: 05/09/2012 06:55:25.428 : Lost sensor contact!!!

It's not quite daily! If I am interpreting the data correctly it seems that the error light comes on as soon as one lost contact is registered.

Alan/Charlie

The site has not changed and the batteries are new. It is happening more since using the newer kit but as you will note it really is a minor issue. It's just another FO's anomaly and just like the "spike removal" Steve has built, I wondered if a similar approach could be employed for lost sensor contacts. Steve is very good at "making a silk purse out of a sows ear"

[steve]

It's interesting that, with the exception of the 25th, it's only reported once. You would imagine that the console would only set/unset the flag at 48-second intervals, so you would expect Cumulus to report it four or five times at least, for each occurrence. So perhaps these are actually spurious - some kind of data corruption?

[nking]

Hi Steve,

Do you need any other data from me to assist or is this one of those things to just live with. I think you may have missed the duplicates for 30/8, 3 and 5/9, not that this is a significant number. I will try another combination of console and transmitter to see if I still get the Lost Sensors, this will probably have to wait a few days.

[steve]

I think you may have missed the duplicates for 30/8, 3 and 5/9, not that this is a significant number.

Those were separate occurrences. My point is that as Cumulus reads the data every 10 seconds, and the console updates it every 48 seconds, then every time the console signals 'contact lost', you would expect to see at least 4 notifications in Cumulus, 10 seconds apart.

[nking]

I think you may have missed the duplicates for 30/8, 3 and 5/9, not that this is a significant number.

Those were separate occurrences. My point is that as Cumulus reads the data every 10 seconds, and the console updates it every 48 seconds, then every time the console signals 'contact lost', you would expect to see at least 4 notifications in Cumulus, 10 seconds apart.

I see what you mean - yes it's odd. Would this also imply that Cumulus takes it as a lost sensor after only one occurrence?

Probably not relevant but the new FO's seem to have different communication protocol to the old FO's as I have mention on the forum https://cumulus.hosiene.co.uk/viewtopic.php?f=13&t=7741 , I wonder if there is any connection? (no pun intended) .

[steve]

It's probably worth you turning on the debug log (delete any existing file) until you next get a lost contact error, then we can look at the raw data and see what it looks like. Might be informative.