| Posted: Saturday Apr 18th, 2015 06:13 am |
|
1st Post |
juwi_uk
Member
back to top
|
I love being woken up at 2:30 in the morning by one of my door contacts triggering and announcing itself as being open; it has just started doing it repeatedly!
It's a heavy duty sensor that supports wider gaps (up to 40mm) but I'm running it with just a 25mm gap. Its a normally open reed switch so the magnet keeps closed.
I've fired up Comfigurator and can see it triggering on/off in I/O monitor.
I unchecked the "Announce" checkbox and did a selective write to Comfort but it still kept announcing; so remind me what does that checkbox do?. I then set the "On" response to null and did a selective write and now the sound has stopped.
Sat hear now at 3am watching I/O Monitor; what else could I be doing at this time of the morning? 
I could see it triggering on/off a few times after that but now as I write this it seems to have stopped in the I/O window and I'm left with just the Flag updates from my UCM/HM every few seconds.
Any ideas; do you think it is the sensor or Comfort wrongly sensing. Surely there's not much to go wrong in the sensor apart from the magnet having enough Oomph to keep the contact closed and surely it shouldn't loose that over 2-3 years should it?
Confused and "Sleepless in Newbury"!
Julian
|
| Posted: Saturday Apr 18th, 2015 06:51 am |
|
2nd Post |
slychiu
Administrator
| Joined: | Saturday Apr 29th, 2006 |
| Location: | Singapore |
| Posts: | 5495 |
| Status: |
Offline
|
back to top
|
The announce box causes the zone announcement when the zone opens so unchecking it shiuld stop it
Your Zone Response may also have the announce zone response, so thaat could have also caused the announcement
It is unlikely that Comfort would give the wrong detection of zone
It is more likely that the magnet has weakened, as permanent magnets lose their magnetism over time. It may have been 40 mm but the strength may have reduced now
try opening the door slightly andn see at what distance the contact iss detected
|
| Posted: Monday Apr 27th, 2015 08:49 pm |
|
3rd Post |
palmlodge
Member
| Joined: | Thursday Dec 14th, 2006 |
| Location: | United Kingdom |
| Posts: | 507 |
| Status: |
Offline
|
back to top
|
I've had the same quite a few times. Always had to replace the door sensor.
I'm waiting for the next one to fail, but I've yet to find a source for resilient units. If you hear of anything, please do let us know.
|
| Posted: Tuesday Apr 28th, 2015 11:23 am |
|
4th Post |
mikeinnc
Member
| Joined: | Wednesday Mar 18th, 2015 |
| Location: | Perth, Australia |
| Posts: | 69 |
| Status: |
Offline
|
back to top
|
When I installed my contacts, back in 2001 when I was building my house, I replaced quite a few of the magnets that came with the sensors with rare-earth magnets I bought from a 'magnet supplier' (yes, there was such a business - supplying schools and labs etc!). I've never had a problem since...... 
I know quite a few of the websites, like DX.com, sell these today.
|
| Posted: Friday Aug 28th, 2015 07:23 am |
|
5th Post |
mikeinnc
Member
| Joined: | Wednesday Mar 18th, 2015 |
| Location: | Perth, Australia |
| Posts: | 69 |
| Status: |
Offline
|
back to top
|
Well, I may have spoken too soon. Just like juwi_uk, I've been woken at random - usually in the early hours of the morning - with a "window open" announcement. That wasn't too bad (I could have just turned the announce off!) but when it fired the alarm in the middle of the day yesterday, it was the last straw. The particular sensors are magnetic and there are two in series (two windows in the room). The wiring is now almost impossible to access as it was 'embedded' in the (double brick) building during construction. I have measured the loop resistance at the panel and it is 9.5 ohms when closed. The (top hung) windows are completely and firmly closed, and it takes a good 20mm of opening for the circuit to respond (it's those rare earth magnets!) As a reference, another loop with four sensors in series measures 3.6 ohms, and a single door sensor loop, 2.3 ohms. So, my question is, what is the tolerance for 'digital inputs i.e what range constitutes a "short circuit"? For now, I have swapped the loop with another to see if the s/c tolerance on the other panel input is better, but if this doesn't work, could I use EOL resistance (and change the loop type accordingly) to swamp the higher-than-expected loop resistance? Many thanks!
|
| Posted: Friday Aug 28th, 2015 01:10 pm |
|
6th Post |
slychiu
Administrator
| Joined: | Saturday Apr 29th, 2006 |
| Location: | Singapore |
| Posts: | 5495 |
| Status: |
Offline
|
back to top
|
9.5 ohms will not cause the zone to be detected as open, it needs more than a few hundred ohms. As this is an intermittent issue which just started to occur they may be otther issues which cause this, perhaps strong electrical interference
What is the sensitivity of the zone type, Increase to the maximum 800 ms so that noise may be ignored
|
| Posted: Saturday Aug 29th, 2015 12:28 pm |
|
7th Post |
Swiss-Toni
UCM Pi Users
back to top
|
Possibility that it could be damage to the wiring causing intermitten contact (worse case scenario), have you tired doing an insulation test on the wiring, or do buzz test from the end that is in the comfort panel with both the contact closed & open. This would also test the contact if it does not pass the test I would then be inclined to test the contact and cable run seperatley if it is possible.
|
| Posted: Saturday Aug 29th, 2015 12:36 pm |
|
8th Post |
Swiss-Toni
UCM Pi Users
back to top
|
mikeinnc wrote: Well, I may have spoken too soon. Just like juwi_uk, I've been woken at random - usually in the early hours of the morning - with a "window open" announcement. That wasn't too bad (I could have just turned the announce off!) but when it fired the alarm in the middle of the day yesterday, it was the last straw. The particular sensors are magnetic and there are two in series (two windows in the room). The wiring is now almost impossible to access as it was 'embedded' in the (double brick) building during construction. I have measured the loop resistance at the panel and it is 9.5 ohms when closed. The (top hung) windows are completely and firmly closed, and it takes a good 20mm of opening for the circuit to respond (it's those rare earth magnets!) As a reference, another loop with four sensors in series measures 3.6 ohms, and a single door sensor loop, 2.3 ohms. So, my question is, what is the tolerance for 'digital inputs i.e what range constitutes a "short circuit"? For now, I have swapped the loop with another to see if the s/c tolerance on the other panel input is better, but if this doesn't work, could I use EOL resistance (and change the loop type accordingly) to swamp the higher-than-expected loop resistance? Many thanks!
You say you have it on a digital input! magnetic contacts do not use a digital imprint as they are strictly an open/closed circuit,
You should have the zone setup as Door/window input in comfort comfigurator it should not make any difference that the contacts are in series......
If it is setup as a digital input and the cables are run within 100mm of electrical cables their is a possiability of an emf field causing it to trigger a responce (unlikely but a possiability)
Last edited on Saturday Aug 29th, 2015 12:42 pm by Swiss-Toni
|
| Posted: Friday Sep 4th, 2015 12:24 am |
|
9th Post |
Swiss-Toni
UCM Pi Users
back to top
|
I have noticed that my response output comands have been altered and responses that used to have names now have numbers, at first I didnt really pay attention to it until I found that some command have been altered and had to edit complete responses as the name said one thing but the comand line said another ie side lights of and comand would be output 1 off but when checked the actual comand would be Output 1 on. I think this maybe what has been happening for these random lights functions. (comfigurator V3 10.4.0)
|
| Posted: Friday Sep 4th, 2015 02:40 am |
|
10th Post |
tech07
Administrator
| Joined: | Thursday Jan 29th, 2009 |
| Location: | Singapore |
| Posts: | 106 |
| Status: |
Offline
|
back to top
|
If the response numbers have changed to numbers, this should have been caused by reading from Comfort using an older version of Comfigurator
Comfigurator now has 2 options when reading from Comfort- Keep names from configuration
- autogenerate names from Action code
Th 1st option may lead to errors if response numbers changed due to adding actions
|
| Posted: Saturday Sep 5th, 2015 11:49 am |
|
11th Post |
mikeinnc
Member
| Joined: | Wednesday Mar 18th, 2015 |
| Location: | Perth, Australia |
| Posts: | 69 |
| Status: |
Offline
|
back to top
|
Swiss-Toni wrote: mikeinnc wrote: Well, I may have spoken too soon.
You say you have it on a digital input! magnetic contacts do not use a digital imprint as they are strictly an open/closed circuit,
You should have the zone setup as Door/window input in comfort comfigurator it should not make any difference that the contacts are in series......
If it is setup as a digital input and the cables are run within 100mm of electrical cables their is a possibility of an emf field causing it to trigger a response (unlikely but a possibility)
Thanks for that useful info. I did have the sensors set-up as 'DoorWindow' and I have checked the wiring as best as possible. As houses in Perth have double brick walls and concrete sub-floors, and the wiring is in the wall cavity (and it is a 2 story house) wiring is almost impossible to get to after the house is completed! As I said, I've swapped the sensors to another input and I've had one false alarm since. I've now defined the sensors as a custom i/p type, and use a response to see if it is a real security breach or an intermittent glitch! The room also has a PIR so if that doesn't fire, it's a good indication that it is not a breach. Short term fix? Yes, but the best for now without MAJOR difficulties! One other point - most (but admittedly not all) false alarms have been in the early hours of the morning - 2am to 3am - so I doubt it is electrical interference as nothing is "on" at that time - no lights; no air-con, etc. I'd expect interference problems to show as units were switched on or off rather than a steady-state current...... As we are just coming out of winter, I did wonder if it was a temperature related problem as that is when it is coldest...well, 7 or 8 degrees is cold for us!!]
|
| Posted: Tuesday Sep 8th, 2015 10:43 pm |
|
12th Post |
Swiss-Toni
UCM Pi Users
back to top
|
Just had a thought, a while ago their was an issue with comfigurator setting the outputs in thermostats automatically and you had to manually remove the output response from "functions-thermostats" cant remember the thread now but if you check your file settings in comfigurator this maybe the reason for strange behaviors (it happened to me where it stopped me from using some outputs)
|
|
