Techman Posted March 6 Share Posted March 6 You definitely need a noise filter between the UPS and your powerline Powerline Noise Filter – Insteon Link to comment
IndyMike Posted March 6 Share Posted March 6 42 minutes ago, Merlin said: Yeah, the PLM is not under the UPS. It does plug into the same 4 way outlet that the UPS plugs into, and they share the same dedicated circuit. I can certainly put a noise filter in front of the UPS as a test to be sure it isn't a device in that rack (UPS or other gear) that is introducing noise on the line. The UPS has a 1500VA/1000W capacity, though its nominal draw is at or below 200-300W with the gear I have in there right now. If you can recommend a good filter, that would be helpful. I've seen a bunch of powerline noise filters on Amazon, but I don't know what type/model/frequency ranges I should be attenuating. I don't have any issues with having the PLM on the garage circuit, other than the fact that stuff does occasionally get plugged into that circuit like battery chargers, power tools, etc. But for testing it wouldn't really matter since nothing is plugged into the garage circuit right now aside from the portable dehumidifier that I ruled out as a noise source earlier. The two circuits are also on two different phases, so that could also be something to test I suppose. I don't know how well the Insteon devices manage bridging, but prior to a few weeks ago, I never noticed any issues reaching equipment so I assume it isn't an issue. I get that... I work in tech and have a basic understanding of electricity, electronics, and cable signaling. Good enough to do basic electrical repairs and troubleshoot networking installs, etc. I'm just less familiar with the Insteon equipment, and assume they have a very high MTBF so everything should work as expected. Yeah... since this is a sinewave UPS, it cleans and filters the power coming in and going out for the stability of the equipment running under its power. So it should strip out any signals coming into the UPS, but pass any signals originating from UPS supported devices back out to the circuit untouched. But that is why the PLM is not under UPS power. 1500VA is a large UPS. Unless you are willing to build your own, the only thing that I know of that is currently available is the XTB-F15+ from Jeff Volp. The filter is advertised for X10, but will work for Insteon as well. https://jvde.us/xtb-f10/ Insteon is very good at bridging the phases - assuming you have dual band modules. X10, on the other hand, will have great difficulties in crossing phases without a dedicated coupler or 220V resistive path (oven, dryer). Again, not sure if the UPS is absorbing Insteon signals to the point where they are at the noise floor, or generating noise in the Insteon band (or both). It's certainly worth testing by moving one of the two to another circuit. Link to comment
Merlin Posted March 6 Author Share Posted March 6 (edited) 1 hour ago, IndyMike said: 1500VA is a large UPS. Unless you are willing to build your own, the only thing that I know of that is currently available is the XTB-F15+ from Jeff Volp. The filter is advertised for X10, but will work for Insteon as well. https://jvde.us/xtb-f10/ Insteon is very good at bridging the phases - assuming you have dual band modules. X10, on the other hand, will have great difficulties in crossing phases without a dedicated coupler or 220V resistive path (oven, dryer). Again, not sure if the UPS is absorbing Insteon signals to the point where they are at the noise floor, or generating noise in the Insteon band (or both). It's certainly worth testing by moving one of the two to another circuit. Yeah, the UPS is sized to be future proofed against expansion, though it's only using about 1/4 of its capacity today. That said, the Insteon noise filter is only 10A, so I'd be more comfortable with the XTB-F15+ if I'm going to run a noise filter on the UPS for more than a short-duration test. They are comparable price-wise, so that isn't even a factor. I'll order one to be safe. If I find some other devices I need to filter, I'll look at the Insteon powerline noise filter and/or the Leviton 6288 DHC Noise Block Plug-In Module which I saw recommended in another post. Looking at Jeff's site, I'm curious if something like the XTB-ANR would be useful? Probably no way to know without putting a scope on my power lines to see what the signal to noise looks like at the distribution panel. It is too bad Jeff's XTBM is no longer being sold. I'd love to get an X10 signal meter to check each circuit to see what the noise levels and Insteon/X10 signals look like. All of the other X10 meters I've found (ESM1 and TesterLinc) appear be be long out of production. Edited March 7 by Merlin fixed typo Link to comment
Merlin Posted March 8 Author Share Posted March 8 Hey folks, I'm still waiting on the XTB-F15+ for the UPS (it's shipped but will be a few days). In the meantime, I have discovered that maybe my original assumption that RF devices were working as expected was false. When I tested previously with the Motion Sensor II, it was sending motion events to the EISY. However, it doesn't appear that those events are always happening, nor does it appear that RF devices can receive updates from the EISY... at least not often. When I did the restore devices command a few days back, it left 9 of my devices with the 1011 in front of their names in the admin console (6 dual band/wired devices and 3 RF/battery devices). I had some extra time today, so I was trying to get some of those 1011's cleared out. But it hasn't taken on any of them. I wasn't entirely sure how to put the Motion Sensor II in connect mode, but I did hit SET on it, which flashed its lights, but I don't know if that wakes it up or not. In any case, it doesn't accept query commands or write updates from the EISY. My 2441ZTH Wireless Thermostat is RF/battery powered, but I have it connected to a 5V power supply which means it should be awake all the time per the manual. The thermostat is able to send updates with current temperature and humidity to the EISY, but I cannot query it from the EISY or write updates to the thermostat from the EISY. I also have a Leak Sensor, from which my monitoring program still detects heartbeats. Though I don't see one for today, so it may not always be able to connect to the EISY. I haven't tried to wake it up yet to write updates. So that suggests that it may not just be the X10 events/powerline noise that is the problem. That points back to the PLM being a problem maybe? Though it could also be a side effect of the powerline issues as some of the Insteon dual band devices may be needed to relay RF signals back to the PLM and cannot do so since they cannot communicate with the PLM over the power lines. That said, none of them are really that far from the EISY/PLM so they shouldn't NEED to be relayed. Speaking of powerline noise. I was also thinking that while I'm awaiting the XTB-F15+ filter, I could potentially confirm or rule out the UPS as a noise source by just unplugging the UPS from the wall outlet and see if that eliminates the X10 signal or communication issues for the EISY/PLM. I'm going to try to test that tomorrow along with moving the PLM around. Thanks. Link to comment
hart2hart Posted March 8 Share Posted March 8 Depending on number of breakers in your box, you might start by turning off a few breakers at a time to see if remaining still work and narrow it down to the circuit making it easier to find the device. Link to comment
Merlin Posted March 11 Author Share Posted March 11 So, I'm still waiting on the filters, but I happened to check the logs today and all the X10 messages are gone. I ran a few queries and the system is acting normally again. I don't know how long it will stay this way but I have changed nothing yet. Link to comment
Merlin Posted March 15 Author Share Posted March 15 @Techman@IndyMike Hey folks. Another update. I'm no longer seeing the X10 messages in the event log. That said, I'm also no longer able to communicate with my Insteon network again, despite not seeing any obvious interference now. It was working fine for about 48 hours (no X10 messages and no issues communicating with the Insteon devices) but now its acting like it was before, just no X10 messages. To confirm, at this point I have made no physical changes, so the PLM and the UPS are still the only two devices on that circuit, and they are plugged into different ports on the same power receptacle. I finally received the 15A noise filter, so as a test, I installed the noise filter in front of the UPS power line just in case it was somehow generating noise I couldn't see. However, adding the filter did nothing to improve my ability to query devices in the Insteon network. As a test, I also tried unplugging the PLM and powering it back on on another circuit. No difference. In both case, the PLM still shows connected per PLM Info/Status. When I try to do a query, I just see this now... Thu 03/14/2024 11:26:26 PM : [INST-TX-I1 ] 02 62 4D 49 C4 0F 19 00 Thu 03/14/2024 11:26:26 PM : [INST-ACK ] 02 62 4D.49.C4 0F 19 00 06 LTSREQ (LIGHT) Thu 03/14/2024 11:26:34 PM : [INST-TX-I1 ] 02 62 4D 49 C4 0F 19 00 Thu 03/14/2024 11:26:34 PM : [INST-ACK ] 02 62 4D.49.C4 0F 19 00 06 LTSREQ (LIGHT) Thu 03/14/2024 11:26:43 PM : [INST-TX-I1 ] 02 62 4D 49 C4 0F 19 00 This example wasn't complete, but I've also seen them with 3 LTSREQ but still no update. But then I can also send a query and get only a single LTSREQ or none at all. I appear to see messages when lights turn on or the motion sensor triggers as I did before. I can generally tell when it isn't going to work since I get the "System Busy" dialog pop up on the screen showing a bar that crawls towards 100% when I do a query. When it works, it usually just updates immediately without the need for that dialog box. As an addition oddity, I'm now seeing weird things like this in the event logs... Thu 03/14/2024 10:59:51 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:51 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:51 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) Thu 03/14/2024 10:59:52 PM : [ Time] 23:00:00 4(0) This spammed the event log ~250 update lines for each of 10:59:51 and 10:59:52 and then stopped. Luckily it isn't doing that constantly. I just happened to notice it when I scrolled back a few pages when I was checking the event log (I have left it running all day to collect data). I see it happening for like 5 lines here and there, but it also goes for a while with no time update lines. So that seems odd. That is on the EISY side and not the Insteon side though. In any case, now that I'm not seeing the X10 messages I'm not entirely sure what I should be looking for. I know I can still power off each circuit one at a time and then run queries to see if that makes a difference. But I'm trying to avoid that given the impact on my family. Plus it seems non-deterministic since this does work fine occasionally and it will be hard to know if it flipping the breaker coincidentally happens when the network is working properly or not. Finding a time when everyone is gone and personal devices don't need to be running is a challenge. Thanks! Link to comment
Techman Posted March 15 Share Posted March 15 @Merlin Take a look at your error log (you can ignore all the 170001 entries) I don't know what those entries are in your event log, I never seen them before. Keep the UPS plugged into the filter as a UPS is notorious for interferring with Insteon signals. If you click on DIAGNOSTICS, PLM INFO, does it show the PLM as connected? Link to comment
Merlin Posted March 15 Author Share Posted March 15 (edited) 4 hours ago, Techman said: @Merlin Take a look at your error log (you can ignore all the 170001 entries) I don't know what those entries are in your event log, I never seen them before. Keep the UPS plugged into the filter as a UPS is notorious for interferring with Insteon signals. If you click on DIAGNOSTICS, PLM INFO, does it show the PLM as connected? @Techman I grabbed the error log and removed all the lines with -170001. Here is all that is left after removing the 170001 lines starting with the most recent reboot a couple of weeks ago. It looks like it's just the startup sequence / file loading. This is under IoX v.5.8.0. Wed 2024/03/06 04:55:12 AM 0 -5 Start Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/TMP Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/LOG Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/WEB Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CODE Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/D2D Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/MAIL Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/USER Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/USER/WEB Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/NET Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/SEP Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/OADR Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/BILLING Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/DEF Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/GLOBAL Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/GLOBAL/i1 Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/GLOBAL/i1/nls Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/GLOBAL/i1/editor Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/GLOBAL/i1/nodedef Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/f1 Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/f1/i1 Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/f1/i1/nls Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/f1/i1/editor Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/f1/i1/nodedef Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/DEF/f1/i1/emap Wed 2024/03/06 04:55:12 AM 0 -110026 ./FILES/CONF/DEF/f10 Wed 2024/03/06 04:55:28 AM 0 -110022 ./FILES/CONF/INSTENG.OPT Wed 2024/03/06 04:55:28 AM 0 -110012 ./FILES/CONF/INSTENG.OPT Sun 2024/03/10 03:00:12 AM 0 -110026 ./FILES/CONF/BILLING/2024 So I assume this means it isn't seeing any errors. I have left the noise filter on the UPS, but it hasn't made any difference, and the the UPS wasn't filtered when it was working properly. Either way it shouldn't really matter. But yes, the PLM is still showing as connected under PLM Info/Status ("51.11.90 v9E / Connected"). Another note, while I do get [RF] update events from my motion sensor when I trigger it, I am still unable to query it directly, despite being powered directly via USB. Edited March 15 by Merlin Link to comment
ELA Posted March 16 Share Posted March 16 (edited) Hello Merlin, Do you have any plug-in modules such as a Lamplinc? Did not see any in your list. Here is one test that might help you progress using a plug-in module ( maybe consider getting one if you do not have one). Create an isolated test network using your new filter. With the filter plugged into a wall outlet plug a power strip( ideally one not advertised as a suppression device) into its filtered output. This somewhat isolates everything plugged into that power strip from the rest of your home power. Now plug your PLM and the one pluggable device ( I will refer to it as "lamplinc") into that power strip. Now see if you can communicate between the PLM and that one device only. If you cannot then you may need a new PLM. If you can communicate between the two use the level 3 viewer and confirm that the "hops left" are 2 or 3. Next put the PLM back into its normal location ( no more filter used). Put the Lamplinc next to it in the same outlet. Use the level 3 viewer to see what the hops remaining are. If that does not work try the same at another outlet location. If it does work then you can start moving the Lamplinc to various other locations and use the hops remaining as an indicator of the quality of the communications at each new location. This could help you zero in on a problem area of the homes power distribution. Watch as the number of hops left decreases greatly. Edited March 20 by ELA Link to comment
Recommended Posts