jeffpdavis1 Posted August 3, 2015 Posted August 3, 2015 I've had a program that has been working flawlessly for the last year that when enabled turns off all 3 kitchen lights by just switching one of them off after a certain time in the evening (e.g. as we're headed to bed). Now it no longer recognizes when I switch any one of them off. The event viewer (at any level) registers nothing. The light the switch controls of course turns off but the program is no longer triggered and none of the other lights in the "Kitchen Night" scene are turned off. This same behavior is happening in other programs of not recognizing previously working "If Control" conditions. I created a test program even simpler than this that also doesn't work. See second code segment. Echo Kitchen Off - [ID 00A8][Parent 008F] If Control 'Main Living Area / Kitchen Dining Lights' is switched Off Or Control 'Main Living Area / Kitchen Lights' is switched Off Or Control 'Main Living Area / Kitchen StoveTop Lights' is switched Off Then Stop program 'Echo Kitchen On' Stop program 'Kitchen 4 Nighttime?' Set Scene 'Kitchen Night' Off Wait 3 seconds Enable Program 'Kitchen 1 Motion Lights' Else - No Actions - (To add one, press 'Action') In the following program the West light is turned on at the switch but the North light remains off. Test Control - [ID 00C4][Parent 0001] If Control 'Main Living Area / GR W Ceiling' is not switched Off Then Set 'Main Living Area / GR N Ceiling' 50% Else - No Actions - (To add one, press 'Action') In addition, I thought that the Admin Console used to register the state of the switches correctly when a switch was turned on or off locally. But it doesn't any longer, except after an active Query. Could all this behavior be the result of yet another PLM going bad in less than 8 months. I sure would appreciate any help with this. I've looked for others having similar problems but it didn't seem to correlate with recent firmware releases. Thanks, Jeff
oberkc Posted August 3, 2015 Posted August 3, 2015 (edited) My experience is that programs, themselves, do not fail. I consider it more likely that the ISY/PLM is no longer seeing the communication from the switches. This seems to be supported by the fact that your event viewer is not seeing the commands. I would treat this as a communication problem. I also believe you can focus on two major possibilities: - something is messing with your powerlines or RF comms - link records in switch, PLM, or both become corrupted. For the first concern, ensure you have a good, clean circuit on which your PLM resides. No other electronic gadgets allowed, unless filtered (my rule). Be sure, also you have good communication between legs of your electrical system. For the second, my temptation would be to peform "show links table" option for one of your switches and "compare". If there is a mismatch, make a judgement about whether the mismatch is due to loss of links on PLM side or device side. It seems there is lots of talk around here lately regarding PLM failures and the loss of link records. I understand your PLM may actually show NO LINK RECORDS when these begin to fail. So...yes, this could be another bad PLM. It could also be one of those random, unexplainable things that happen from time to time. Restoring the PLM or the device can often solve these. I am not absolutely confident about this, but the fact that your device responds to queries suggests to me the link records are intact and you should focus, instead, on pure comms issues. I would not spend any more time trying to fix the program, itself. I don't believe this is the cause of your problems. Edited August 3, 2015 by oberkc
Techman Posted August 3, 2015 Posted August 3, 2015 in addition to what oberkc suggested, have you added new electronics, appliances or CFL bulbs, etc.about the time the programs quit working
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 Thanks for both of your replies. Yes the PLM is actually on its own circuit, shared only by a Monster power strip with three IOLincs in it. All 3 IOLincs have been working fine for a long time. I could plug the strip into a FilterLinc as another testing step but I think I'll leave it alone while I test other hypotheses. The link records for all three Kitchen lights were Compared and had mostly "Identical" with 2 or 3 "Identical/Ignore" which I have so far. Curiously the GR W switch which is in the same general area had one mismatch, that I've corrected with a Restore. Regarding the addition of any new electronics about the time the programs stopped working, yes I have. I can't pinpoint exactly when things stopped working but it may have been coincident with a couple of my motion sensor batteries dying (Grrrr, that's another story) and being replaced/deleted and added again. I wish there were a better option for those things. The biggest thing that changed was the addition of two Ubiquiti Unifi APs, one in the same Great Room / Kitchen area. These are both in the 2.4GHz range though and shouldn't (in theory) interfere. No other appliances or CFLs. I could I guess do without the APs but it sure evens out the wifi in and around the house quite nicely. There is another piece of information that I'm sure most will feel relevant. The UDI 994i Controller is plugged into a UPS that's plugged into a filter on a separate but physically near circuit to the PLM. I've struggled with ways to keep the Controller powered during outages. We have a generator that comes on 35 seconds after the outage is detected, but I have the controller send me an email when it detects the power change in either direction and used to need the controller to stay on for that. I'll look at those two programs again to make sure they would work in these conditions of losing and regaining power. I'll try filtering the powerstrip the IOLincs are in that's residing in the same circuit as the PLM. Would it be ok to have the Controller and PLM on the same circuit? I had it that way a year or so ago.
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 Also my switches are ToggleLincs that are not Dual Band, so RF shouldn't (should it?) affect it.
LeeG Posted August 3, 2015 Posted August 3, 2015 (edited) When multiple devices fail to show state changes it is usually either a comm issue or the PLM has lost link records. It only takes a few minutes to run File | Restore Modem (PLM) which rewrites the PLM link records. Edited August 3, 2015 by LeeG
oberkc Posted August 3, 2015 Posted August 3, 2015 (edited) Your understanding that single-band devices should not be affected by RF issues is the same as mine. That "monster" power strip concerns me. Is this a surge suppressor or some type of power conditioner? I think it a possibility that this could be a contributor to some comm issues. You may want to consider replacing it with a simple power strip without any electronics or suppression capability. I would also tend to filter UPS in general, regardless of circuit, as a preventative measure. Once filtered, I would keep the power supply of the ISY on the UPS rather than move it to the same circuit as the PLM. Edited August 3, 2015 by oberkc
Techman Posted August 3, 2015 Posted August 3, 2015 It's OK to plug the ISY into the UPS and the UPS should be plugged into a filterlinc. Don't use a filterlnc with the IOLincs as it will prevent their (single band) powerline signal from getting to the ISY. Try unplugging the Monster Power strip to see if it helps. Some power strips can create noise on the powerline. Is your PLM a dual band model? Changing the batteries in the MS or adding a AP shouldn't create the problem you're having. As side note you could always update your single band switches to dual band which should help improve reliability.
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 When multiple devices fail to show state changes it is usually either a comm issue or the PLM has lost link records. It only takes a few minutes to run File | Restore Modem (PLM) which rewrites the PLM link records. I was going to do this until I read the warning about ensuring all battery devices are in programming mode. For all three devices, the links compared identically. It seems like the process would take settings from the ISY and write them to the PLM regardless of whether the Motion sensors and door sensors are in programming mode, right? If that's the case then I'll try that right away.
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 Your understanding that single-band devices should not be affected by RF issues is the same as mine. That "monster" power strip concerns me. Is this a surge suppressor or some type of power conditioner? I think it a possibility that this could be a contributor to some comm issues. You may want to consider replacing it with a simple power strip without any electronics or suppression capability. I would also tend to filter UPS in general, regardless of circuit, as a preventative measure. Once filtered, I would keep the power supply of the ISY on the UPS rather than move it to the same circuit as the PLM. I bought another one that is simpler than the Monster one. That's probably a good idea to switch it out.
LeeG Posted August 3, 2015 Posted August 3, 2015 A File | Restore Modem (PLM) does not update devices unless a new PLM is being used where the Restore Modem (PLM) is changing the PLM address in all the devices. A simple File | Restore Modem (PLM) rebuilds the PLM link data base only.
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 It's OK to plug the ISY into the UPS and the UPS should be plugged into a filterlinc. Don't use a filterlnc with the IOLincs as it will prevent their (single band) powerline signal from getting to the ISY. Try unplugging the Monster Power strip to see if it helps. Some power strips can create noise on the powerline. Is your PLM a dual band model? Changing the batteries in the MS or adding a AP shouldn't create the problem you're having. As side note you could always update your single band switches to dual band which should help improve reliability. Yes, the PLM is the 2413S dual band one, and as you might guess I started out swapping out old toggle switches with a few ToggleLinc. Now I've got 20 ToggleLincs and upgrading to the other kind is not really a job I would relish but wish I'd had the foresight I have now. Oh well.
Techman Posted August 3, 2015 Posted August 3, 2015 Yes, the PLM is the 2413S dual band one, and as you might guess I started out swapping out old toggle switches with a few ToggleLinc. Now I've got 20 ToggleLincs and upgrading to the other kind is not really a job I would relish but wish I'd had the foresight I have now. Oh well. Make sure that you don't have a computer or any electronic devices plugged in close to where the PLM is as they can interfere with power line signals coming and going to the PLM. Did unplugging the Monster power strip make a difference?
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 FIXED. No unplugging the Monster power strip did not make a difference but I am going to replace it anyway. File | Restore PLM was successful and all the programs work as they did previously. Thanks to all for the helpful suggestions. Jeff
LeeG Posted August 3, 2015 Posted August 3, 2015 Unfortunately that likely means the current PLM will fail again. What does Tools | Diagnostics | PLM Info/Status show for PLM firmware? xx.yy.zz. v?? / Connected
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 25.29.36 v9B / Connected Does the firmware tell you anything? I was thinking that this doesn't answer the question of cause. There were a couple of GFCI outlets that were thrown inside because of landscapers using their portable concrete mixer. Neither of those outside circuits were shared by the ISY or PLM but still it happened a couple of times. I noticed nothing in the ISY system when it happened. Any other guesses as the cause? I know they're supposed to last 2 years but the last two have been under a year or at least closer to one year than two.
Brian H Posted August 3, 2015 Posted August 3, 2015 There should be a white sticker on the back of the 2413S. It should give you a hardware revision. I have a 2413S with 9B firmware and it is a hardware 1.B. I also don't think it is two years old but I would have to look for the invoice or look at my purchasing records on the Smarthome sales site.
LeeG Posted August 3, 2015 Posted August 3, 2015 (edited) A PLM with v9B firmware is old enough to have the capacitor problem. They over heat and cause the power supply voltage to drop losing some/all of the PLM link records. Once the link records have been lost it is common for the problem to surface again. The PLM hardware v2.x firmware v.9E has corrected capacitors. Edited August 3, 2015 by LeeG
jeffpdavis1 Posted August 3, 2015 Author Posted August 3, 2015 The one that's operational right now says V1.B on the back of it and v9B in the PLM Info/Status box. My spare that I got in Feb 2015 says V2.0 on the back. How do I tell what the version is that compares with "v.9E" or is the v2.0 enough to confirm it has the corrected capacitors? Perhaps I should call it quits on the current one, replace it with the v2.0 I have, get Smarthome to send in a replacement under warranty and then keep the new one as the next one in rotation.
LeeG Posted August 3, 2015 Posted August 3, 2015 PLM with hardware v2.x (from white sticker) has capacitor fix.
Brian H Posted August 3, 2015 Posted August 3, 2015 I have seen posts that 1C had the new value capacitors. 2.0 definitely would have them.
Recommended Posts