b_p_smith
join:2002-02-13
 ·RipNet
| Magnetic reed switches - Can they "reverse" themselves?
Long story short: I have a sensor on my home alarm system for the sump pump. Simple reed switch with a float with a magnet in it. Water rises, float rises, switch closes, panel goes into alarm.
Well, last week (of course at 1:30 in the morning), the system went into trouble. After much investigation, one of two things have happened:
1) The panel software has suddenly decided that this zone is Normally Closed instead of Normally Open, or
2) The reed switch itself has magically transformed from NO to NC.
Bottom line, if the magnet is up, the alarm is clear. If the magnet is down, the alarm is raised. For now, since I'm not worried about the sump level, I have the assembly mounted upside-down.
So my question is: I know that these switches can "stick" one way or the other and stop working. But is it possible for such a switch to actually *reverse* its logic?
Brad.
--
RipNet (wireless) -} WinXP -} WinProxy 5 -} LinkSys BEFSR81 -} about 8 machines running Win2003 & XP |
|
Kringle
Dr.D
Premium
join:2004-02-27
Pierrefonds, QC | I've never heard of that happening. I would say that it's more likely that there's a programming glitch in your panel. If it is the reed switch, I certainly want to hear about it.  |
|
b_p_smith
join:2002-02-13
·RipNet
| Luckily, my alarm dude can dial into my system from his home and download/upload all the settings. The system is over 10 years old and has been flawless up to now. A sudden "logic error" with no reasonable explanation would be a bit worrying, to say the least, since you never know what else may have "glitched". I did a full power-down (with battery disconnected) reset to see if that had any effect, but no go.
I guess when I finally get it figured out, I'll post back.
Brad. |
|
Kringle
Dr.D
Premium
join:2004-02-27
Pierrefonds, QC | Good luck with that. |
|
jjoshua
Premium
join:2001-06-01
Scotch Plains, NJ | reply to b_p_smith
Maybe a hardware error on the panel?
The place where the wires connect to the panel may have something which switches between "normally open" and "normally closed" mode. |
|
b_p_smith
join:2002-02-13
·RipNet
| Checked that. It would appear it's all software driven. I even have the original installer's manual (I make sure I grab all that stuff), and there's nothing there. No scorch marks on the boards either, which is always a good thing. 
As long as one of the fire sensors don't "reverse" and kick in the sirens in the middle of the night, it's all good for now. |
|
Kringle
Dr.D
Premium
join:2004-02-27
Pierrefonds, QC | If it's all software controlled, it could be that the flash memory is going bad or it's really just another one of those irreproducible glitches. |
|
TheMG
join:2007-09-04
Edmonton, AB | reply to b_p_smith
If you have a multimeter or continuity tester you could test the switch to see if it's doing what it should. |
|
tschmidt
Premium,MVM
join:2000-11-12
Milford, NH
·Verizon Online DSL
| said by TheMG  :could test the switch to see if it's doing what it should. That is a good first step. That will quickly identify if problem is the switch itself or panel logic.
I'm betting on it being a problem with the panel, either hardware or software.
/tom |
|
SmokChsr
Who let the magic smoke out?
Premium
join:2006-03-17
Saint Augustine, FL
·AT&T Southeast
| reply to b_p_smith
One question that comes to my mind is, was the float switch ever tested when installed? My logic is perhaps the switch was a bit "inexperienced", and didn't realize it should be open or closed as the case may be in said position. Thus it may have been in the wrong state when the panel was programed. As the switch became more accustomed to its job description it decided it was in the wrong state for the given condition and changed, with out letting the alarm panel know that it had finally figured out it was wrong before. |
|
b_p_smith
join:2002-02-13
·RipNet
| The switch closes when the magnet is near, and opens when it's moved away. If I had access to a new switch of the same type, I'd compare them, but I don't.
The switch has been in operation for 10 years, and not only was tested on install, has actually triggered for real a few times when the levels got high.
At this point, I'm in a holding pattern until my alarm dude can log into it and take a peek. My guess is if there was a logic glitch (bad flash, etc), there will be other things wrong that aren't yet obvious.
Brad. |
|
SparkChaser
See the Light
Premium
join:2000-06-06
Downingtown, PA
clubs:
 ·Verizon FIOS
 ·Comcast
| said by b_p_smith :The switch closes when the magnet is near, and opens when it's moved away. Is this information from a measurement with ohm meter with the switch out of the circuit? If so do you get close to zero and infinity?
Ron |
|
b_p_smith
join:2002-02-13
·RipNet
| Yeah, I tested with the switch cut out of the circuit entirely. The line has supervisor resistors, so before I cut the switch out the proper test would have been voltages, and they were within spec for the open & closed state. It's just that the state itself is backwards. |
|
SparkChaser
See the Light
Premium
join:2000-06-06
Downingtown, PA
clubs:
·Verizon FIOS
·Comcast
edit:
October 1st, @08:55AM
| said by b_p_smith :Yeah, I tested with the switch cut out of the circuit entirely. So, from what your saying the switch is operating normally, open until a magnet is near it. Which leaves a weird hardware or software problem with the control unit. Since, it is just a voltage going in, there could be a switch (HW or SW) that tells it how to interpret to the signal.
"when you have eliminated the impossible, whatever remains, however improbable, must be the truth" - Sherlock Holmes 
--
"Facts do not cease to exist because they are ignored." - Aldous Huxley
|
|
b_p_smith
join:2002-02-13
·RipNet
| Right. In this case, the interpretation would be in software.
The main point of my original question was to decide if, originally, "normal" for the reed switch was NC, and it went open when the magnet was near; and somehow it became NO and closed when the magnet was near. Hardware-wise, that's the only logical explanation.
Since I think we've pretty much eliminated switch reversal as a likely cause, that points to software. But it'll be a while until I can prove that.
Absolute worst case, I'll buy a replacement float switch (I'm sure I can get them at Home Depot or similar) that's NC and just replace it. That'll tide me over until the panel decides to un-glitch itself (probably at 1:30AM again with my luck). |
|
Splitpair
Premium
join:2000-07-29
Cow Towne
 ·T-Mobile US
·T-Mobile US
·AT&T Southeast
| said by b_p_smith :The main point of my original question was to decide if, originally, "normal" for the reed switch was NC, and it went open when the magnet was near; and somehow it became NO and closed when the magnet was near. Reed switches do not reverse as if even the smallest margin of them did alarm sytems worldwide would become compromised and that defect would become well known and well litigated.
Wayne
--
If you cannot fix it with a buttset and some beanies you ain't a technician.
|
|
Splitpair
Premium
join:2000-07-29
Cow Towne
·T-Mobile US
·T-Mobile US
·AT&T Southeast
| reply to b_p_smith
BTW your friend should be able to upload the panel's current programming and see if the zone has defaulted to something other then what if should have been programmed as.
If the panel did a default that needs to be investigated further.
Wayne
--
If you cannot fix it with a buttset and some beanies you ain't a technician.
|
|
b_p_smith
join:2002-02-13
·RipNet
| Indeed. Especially if it was a default for just *one* memory slot. The system includes all sorts of text labels and whatnot, and in my checking none of them are compromised. If I could see things like scrambled labels or something, it'd be so much more obvious.
But yeah, he should have the original profile saved, and he should be able to compare. I'm going to try to get him to poke at it over the weekend.
Brad. |
|
Splitpair
Premium
join:2000-07-29
Cow Towne
·T-Mobile US
·T-Mobile US
·AT&T Southeast
| said by b_p_smith :But yeah, he should have the original profile saved, and he should be able to compare. I'm going to try to get him to poke at it over the weekend. Brad. If possible ask him if he can upload the current programming in the panel then do a “stare and compare” of that programming against what he has saved. If he is running Compass thats real easy to do.
FWIW with most alarm panels it is a value per address that controls zone types.
Wayne
--
If you cannot fix it with a buttset and some beanies you ain't a technician.
|
|
b_p_smith
join:2002-02-13 | Thanks for the tip. I'm not 100% sure what he's running, but we'll work through it. |
|
