Valve failure

[frobnitz]

Hi. This morning before heading to work I noticed a pool of water in my driveway. One of the large area sprinkler zones was running, but should have shut off about 3 hours before. I unplugged the 24 VAC transformer that feeds the rain8 relays and the water stopped, and I figured I would look at it some more when I got home.

When I did get home, I plugged the transformer in, and the sprinkler started up again. I looked at the zone status page and noticed that the status indicated "valve failure". I set the zone for a manual 2 min cycle, and sure enough, it turned off fine at the end of the cycle.

I do have the safeties set on the rain8 for 1 hr on this zone, so I am investigating why they failed to stop it, but I am guess the relay stuck somehow. But I have some ideas for some changes to MCS Sprinklers to make my normally ultra reliable system even more so.

1) If a "valve failure" state is detected, can HS email me an alert to notify me of the fault? I don't normally check zone status because the system is so reliable. But if MCS detects a fault, it would be great if it could tell me about it.

2) Can there be at least an option if a valve failure is detected to try cycle the zone for 1 or 2 minutes to clear the fault? And also to notify there was a failure and it was cleared? That would be automatically try what most people would do anyway.

Any possibility to see these added in a future release?

Thx
mike

[Michael McSharry]

It makes sense to elevate the Zone Failure to an email notification. It may take some thought as to when not to do it to prevent floods of email.

What Zone Failure means is that mcsSprinklers tried 4 times to command the valve to the desired position and the Rain8 never returned status or the status that was returned was the wrong one.

An idea that may be worth exploring is to continously try to turn a failed valve off if there is no auto program active. I believe when the auto program is active it will command any valve to the correct position that is found to be reporting the incorrect status.

There is already the option for the Rain8Net to poll every minute for status and in those cases it will try to set the valve to a correct position if status is not reporting it as such. I would need to look if it excludes the failed valves from the retry, but I do not think so.

[BraveSirRobbin]

I'm curious to know the reason for failure also (i.e. was it the Rain-8 getting 'stuck' in the on state)?

[Michael McSharry]

There are EPROM timers available in the Rain8 and their purpose is to turn a valve OFF after the programmed timeout duration. If you have not set these up then you should. These are the primary reason to use a Rain8-type device rather than a basic relay type output. You want protection logic as close as possible to what you are trying to protect. For example, any issue associated with the abilty of the host (PC, Security panel, etc) to communicate with the valve/relay would prevent its abilty to turn off a relay that was set in the ON state. With the Rain8 the communication path is a printed circuit trace to the triac and the microcontroller logic is pretty welll proven so likelyhood of a failure in the protection capabilty is small. Of course EMI can do funny things to electronics so there are no 100% guarantees.

When I did get home, I plugged the transformer in, and the sprinkler started up again. I looked at the zone status page and noticed that the status indicated "valve failure". I set the zone for a manual 2 min cycle, and sure enough, it turned off fine at the end of the cycle.

If you are using Rain8Net then the master Rain8Net module can obtain its power from the RS-232 cable so when you unplug the 24VAC it deactivated the valve, but may not have affected the Rain8Net processor. This processor continued to maintain the valve ON command so when 24VAC was restored the valve again opened.

Whatever was causing the valve failure detection by mcsSprinklers could have been reset by the 24VAC power cycle so the Rain8 was now able to accept new commands.

It would be useful to look at the Run Log to confirm the mcsSprinkler control actions during the time of the failure. If you have debug enabled then the low level communication history is also available. If you did not restart mcsSprinklers then the Serial IO page (and IOWindow.txt file) will have the low level information.

Any additional information you can provide would be useful to try to characterize the failure mode so appropriate recovery logic could be designed.

[BraveSirRobbin]

That was the point I should have made clearer in my earlier post (if the unit failed to turn off WITH the internal run away timers enabled).

[frobnitz]

I had the internal runaway timers set. They did not close the valve as per the setting (I have drop on 120 min failsafes, and sprinklers like this one on 1 hr failsafes).

I am using the Rain8net devices, and yes, I did not pull power to the main module. But the light on the module in question (I have a master and 2 expanders), did go out when I pulled the 24VAC power.

Unfortunately, I upgraded MCsprinklers to the current rev after the failure to see if there was notification of a failure as a new option (I was running about 6 months behind current rev). So I don't have the low level data present.

The other zones ran fine with that one stuck valve open for the sprinkler program that morning. So comms were running just fine, even to other zones on that same rain8net.

Thx
Mike