SynchroLinc - Insteon #2423A5

[robl]

I'm trying to use / figure out the SynchroLinc and have run into some confusing behavior. Any advice or suggestions would be appreciated!

 

First some context:

• I have 3 devices working well (I think) for a clothes washer, a clothes dryer, and a dishwasher.
• In these 3 cases, I simply want to be notified sometime after it finishes, so I don't remember 2 hours later. Or a day later.
• I've found the timeout values are too short for all 3 devices, so I've had to trigger an insteon program that waits 30 seconds, and if on never occurs again, it triggers my events.
• I'm using the z-wave doorbell to announce the events, it works great!
• I use a plugin power meter on each of the 3 devices to record it's power patterns. 
• It's tricky, as many of the newer devices go into low power (0w or 1w) states many times during operation.

Ok, now that those 3 work, I've added one to my TV. Powering on the TV will trigger a switch to turn on all the AV equipment, and trigger a harmony macro to set the kids up on hulu to pick a show.

• In ISY options for this device, I see the following:
  • Trigger Threshold (Watts)
  • Holdoff (Secs)
  • Hysteresis (Watts)
• Unfortunately, the smarthome docs don't cover any of this, they only cover the calibration mode, and that doesn't result in working values in any of my test cases.
• I had thought the following, can you correct me where I'm wrong?
  • Trigger Threshold - once power usage reaches up to this level, the device immediately turns on. (or does it wait holdoff seconds first) ?
  • Hysteresis - once power usage drops to Trigger-Hysteresis and stays that way for holdoff seconds, the device turns off.

For my TV, I've found the following:

• It consumes ~12-15w power on standby (off) and that jumps up to 40w+ and sometimes 250w while operating. Power fluctuates a lot, but rarely does it go down to 20w or lower while on (and then only really briefly)
• When turned off, power usage drops to 0w for a fraction of a second, and then jumps back to ~12-15w usage.
• I had it on 24v/5v/1sec and that didn't work well. (trigger/hysterises, holdoff)
• I changed it to 20v/4v/16 sec and that still doesn't work well.

Here's the failure mode I see:

• Power on TV makes the linc go on immediately, and the switch powering my a/v gear all turns on. So far so good.
• If I make it past the first couple of minutes on turn on stage, then it's stable. Turning off the TV makes the linc go off, which triggers the switch to turn off all my a/v gear. good.
• Sometimes when I turn the TV off, and power usage drops from ~100-200w down to 15w, the linc turns off immediately, even when holdoff=10 sec or 15 sec. Any idea why? Shouldn't it wait?
• BUT during the first ~60 seconds of operation, sometimes the synchrolinc will turn off. The light goes red from 1 to 5 seconds. The power meter shows the TV as consuming 60-100w at that time and clearly not off. This false reading causes all my a/v equipment to get shut off, then quickly turned on again. It often loops through shutting it off and on again ~3 times then it stays on.
• Sometimes it works just fine during the first ~60 seconds of operation. The only thing I've noticed is that if I get a bad turnon with the on/off loops, the next one seems to work fine, and an hour later it seems to work fine. But if I wait 4-6 hours later, I'm back to the looping behavior. Almost as if some capacitor has drained in the linc and sent me back to purgatory.
• I have all events logged to a file in the ISY and am now 100% positive the TV is on and working when the linc says it's off - it's not some other remote or device or harmony turning off the tv.

I think I can work around these issues by using ISY as a true "holdoff" - forcing my programs to wait until power is out for ~15 seconds, without going on during that time, to then fire my TV off programs. BUT I'd rather use the device properly as intended!

 

Thanks for any help!!

 

Rob

[robl]

Still not solved, I replaced it with another unit, and it behaves identically. So my guess is that the units are not defective, my understanding of how they work is!

 

I still have a horrible loop of all my equipment going on and off again (as the synchrolinc falsely triggers my macros), but if I can make it past the first 10 minutes or so, the device behaves fairly well for the rest of the day. It resets again the next day to behave poorly.

 

I tried removing the device from ISY, doing a factory reset, going through calibration mode THEN adding to ISY, but the problem persists. Darn. Will look at the REST api next to see if I can read out current voltage and see if rPi can check it to see if it's a false reading or not. Terrible workaround though.

 

Any hints or suggestions would be most appreciated!

[larryllix]

Still not solved, I replaced it with another unit, and it behaves identically. So my guess is that the units are not defective, my understanding of how they work is!

 

I still have a horrible loop of all my equipment going on and off again (as the synchrolinc falsely triggers my macros), but if I can make it past the first 10 minutes or so, the device behaves fairly well for the rest of the day. It resets again the next day to behave poorly.

 

I tried removing the device from ISY, doing a factory reset, going through calibration mode THEN adding to ISY, but the problem persists. Darn. Will look at the REST api next to see if I can read out current voltage and see if rPi can check it to see if it's a false reading or not. Terrible workaround though.

 

Any hints or suggestions would be most appreciated!

Did you factory reset the device before linking to ISY?

 

If not, do it, and then use ISY restore, to gain easy linking again.

[mwester]

Try plugging a filterlinc between the syncrolinc and the load.  Televisions use switching power supplies, which have some "interesting" characteristics which can confuse Insteon devices.

[lilyoyo1]

I have the LG oled tv and put hue lights behind it. The lights are one color when the tv is off and a dim white when the tv is on. I found when I first turned on the tv, it would do exactly what you are experiencing. To fix, I simply added a wait to my program instead of allowing it to trigger as soon as it happened. I left the default settings on the syncrolinc (outside of the calibration for the tv).

[stusviews]

A few years ago, I did what mwester suggested (post #4) from the beginning  The SynchroLinc has been flawless.

[robl]

Thanks, all! With an intermittent problem, it's hard to diagnose.

 

I tried a filterlinc, and initial indications were good, then the SynchroLinc went to off in the middle of a movie, about 1 hour in, when the TV was pulling >100W. Grr. The next morning, it went back to the old behavior of indicating on, then off, then on, then off, and cycling every 30 seconds or so while the TV was on the whole time. (which of course then turned off my roku, amps, etc. then on then off per my programs) Perhaps our 80 inch Visio puts out more parasitics than a normal TV? Perhaps a regular surge protector or something as a line conditioner may help with the spikes.

 

I've resorted back to the timer, but at 40 seconds, it's a bit of a pain. When my wife finishes watching TV and turns it off, she wants to know it's working, so she just stands there for up to a minute for the timer to expire and all to turn off. Hmm.

 

Btw, if anyone else goes down this route, one thing I found very helpful was to make a bunch of notify/email/logfile actions for the AV center. So if any of my SynchroLinc's triggers on/off or a related on/off module or light changes, it appends to a file. If any of my programs run, it calls a different notify and appends the program name and status to the file. If any of the ISY IR fields are received, it also appends the IR code and action to the file. Finally, my Harmony triggers IR39 and IR40 when the ISY device is turned on, and I used IR30-38 for debug for common harmony actions to see when they turn on/off. When those special IR codes are ready, I have a dummy program with the same name as the Harmony Activity trigger and it only appends to the log file.

 

Net result, I can debug all the complicated change of events, and even better see how long my delays need to be to get past most of the power cycling.

[larryllix]

You mention your macro loops turning things on and off. Do you have your synchrolink downstream from your switching so that the power is intermittent?

 

If not, it sounds like you don't have your power on level set to match your power draw levels properly and/or the hysteresis not set high enough to accomodate all the power levels the equipment can draw.

 

I didn't find mention of your actual settings you are using.

 

Sent from my SGH-I257M using Tapatalk

[robl]

@larryllix, thanks for checking! Yes, the synchrolinc is independent of the on/off controlling the other A/V equipment. The goal was for peeps to be able to turn on the TV and have the rest power on and select hulu, and then to power off the TV and have it all shutdown gracefully. Just not unexpectedly at random.

 

My TV uses 12-15w on standby, then 50-120w at startup. On a pure black screen, it has 74w, and max power usage can climb to 350-370w. On my power meter, I never see it drop down to 15w while it is actively running.

 

Current settings are trigger=20w, Hysteresis=4w, holdoff=20sec. Does this seem reasonable? I also had trigger=24w but it worked the same.

 

For the most part it works, except when it goes into "spooky mode". For example, I can be watching a movie, power meter says ~100w is being consumed, and synchrolinc will suddenly turn red. Blap go my macros. And then 5 sec - 35 sec later, synchrolinc turns green again, and blap all my other equipment turns on. During this whole process, the TV is always on and the power meter shows substantial usage. For this example, it occured just once during a 1.5 hour movie.

 

Sometimes it also hits during initial power on. It will cycle on/off 4-6 times then stabilize. If I try it an hour later, first turn on is fine and I don't see these cycles. But next morning it might do it again. Note the TV stays on consuming power the whole time.

 

I plugged in a filterlinc between the TV and the synchrolyinc. It might be slightly better, but both problems remain.

 

So for now, I have an ISY macro that waits for synchrolinc to say it turned off, waits 30 seconds, then if no on event occurs, it sets a var=0 and triggers things off. The downsides is my wife will turn off the TV then wait a minute just to be sure it all works as expected.

 

Thanks for any help!

[stusviews]

Increase the trigger (best if you use the setup procedure as a starting point) and Increase the hysteresis value.

[larryllix]

I was just ging to post exactly the same as Stu advised.

 

Also, did you ever factory reset that beast? I would factory reset it and use admin console to restore it again.

 

I had some problems with mine and it still seems to lose it's settings or sensing? every year or so. Usually an unplug and plug back in again fixes it all but I am not sure these things were the best for stability.

 

Your measurements for 50 -120W at startup and 74W at blank seem like something is wrong. Perhaps the TV starts up without the backlighting on and then turns it on later???

 

Anyway, I would get your trigger up past the 50W minimum you metered and the hysteresis level down below it.

That puts your hysteresis at say 60W - 30W = 30W = below the min ON but above the max Idle.

 

For testing turn the timer to 0 seconds, test it for a while and then add a few seconds to it making it more stable.

 

Let us know how something like that works out.

[robl]

Thanks for the good ideas - waiting for a block of time where I can watch some movies, I'll try raising the trigger threshold and hysteresis then.

 

Another Q on the syncrolinc and noise filters - I tried using both on my washer and both on my dryer. Sometimes it seems the signal there is intermittent.

 

And ... the 10A noise filter on the dryer seemed to be DOA. I ordered another, and it worked for 10 minutes, then died. Dang drier! It's a gas dryer, and with my watt-meter, usage was ~230-350w with occasional spikes up to 600W. Seems like it would be fine for 10A (is that ~1200W at 120V AC?). Ok, so I removed the filter, and now 1 month later, the synchrolinc on the dryer is partly fried. The led turns on/off as needed, but the signal never makes it out. If I force a query from the ISY console, it will update, but never on it's own. I moved the device to my test area near ISY and it still doesn't get the signal.

 

So ... another idea is the 20A wired in filter from X10-pro. I bought 1, but it's wire-in only, not pluggable. Has anyone ever bought a plug base or hacked up a power cord to make this work in a plug-cord fashion? If I do this, i'd like to make it as safe as possible, and avoid just wire ties or simple solder. I'd hate to burn the place down.

 

Has anyone else done this or have any other ideas?

 

Thanks in advance! 

[stusviews]

There are any number of plastic and metal housings available for electric devices. Use a 3-wire extension cord cut in half (approx.). This device should be used to hold the wire in place both entering and leaving the enclosure.

 

In any case, use the instructions in the Owner's Manual to set the starting point for the SynchroLinc.

[Brian H]

The 10 Amp Smarthome FilterLinc has a internal soldered in slow blow 10 amp fuse in it. Maybe there are very short current spikes eventually blowing the fuse.

Another possibility is poor construction of the FilterLinc. There is a thread here on very poor construction. Like cold solder joints.

 

JV Digital Engineering makes an XTB-F15 15 amp filter. I asked if it was effective on the 135.61 KHz Insteon power line frequency. Jeff indicated it was.

 

ACT now out of the X10 line of devices. Use to make an ACT AF120. It was a 15 amp filter and had an external fuse if you ever managed to pop it.

You may still find some on auction sites.

[larryllix]

Thanks for the good ideas - waiting for a block of time where I can watch some movies, I'll try raising the trigger threshold and hysteresis then.

 

Another Q on the syncrolinc and noise filters - I tried using both on my washer and both on my dryer. Sometimes it seems the signal there is intermittent.

 

And ... the 10A noise filter on the dryer seemed to be DOA. I ordered another, and it worked for 10 minutes, then died. Dang drier! It's a gas dryer, and with my watt-meter, usage was ~230-350w with occasional spikes up to 600W. Seems like it would be fine for 10A (is that ~1200W at 120V AC?). Ok, so I removed the filter, and now 1 month later, the synchrolinc on the dryer is partly fried. The led turns on/off as needed, but the signal never makes it out. If I force a query from the ISY console, it will update, but never on it's own. I moved the device to my test area near ISY and it still doesn't get the signal.

 

So ... another idea is the 20A wired in filter from X10-pro. I bought 1, but it's wire-in only, not pluggable. Has anyone ever bought a plug base or hacked up a power cord to make this work in a plug-cord fashion? If I do this, i'd like to make it as safe as possible, and avoid just wire ties or simple solder. I'd hate to burn the place down.

 

Has anyone else done this or have any other ideas?

 

Thanks in advance! 

BTW: Your initially posted understanding of the hysteresis is incorrect.

 

eg.

If you want the device to turn on at 500 W, off at 400 W

Set the Trigger to 500 W

Set the Hysteresis at 100 W

AFAIK the timer is only for the On delay. ie. >500W for X seconds. I don't think it affects the Off time but could be wrong on that one. (Holdoff = hold off)

 

 

So if you set trigger and hysteresis both to 500W, it should turn off at 0W or never turn off at all.

[robl]

> JV Digital Engineering makes an XTB-F15 15 amp filter. I asked if it was effective on the 135.61 KHz Insteon power line frequency. Jeff indicated it was.

 

Thanks! I've not heard of this site before, and the filters seem to have a reasonable price. Has anyone used the "XTBM X10 Signal Meter" or Signal Analyzer to help track down and reduce noise? I'm a bit leary that it's all X10 and not Insteon, so tracking commands may not work, but perhaps it would help with noise. Ideally I could plug it in at various outputs, read the noise, and disconnect stuff and try to optimize the results. I've already spent some 20-30 hours unplugging everything, switching off circuit breakers, to try and diagnose, but it hasn't helped much. I now have a signal bridge and ~5 wireless repeaters in hot spots, but still find devices like the IR transmitter, which are powerline only and not dual band, still behave erratically.

 

To be fair, the house is odd, 1 normal 200A panel, 3 sub-panels, solar, a sewage grinder & alarm (big pit to collect stuff and shoot it uphill to the street), and 2 whole house fans. Outside is stucco, which inhibits signals, and it's ~7 split levels. Right now I've got almost everything behind an insteon power conditioner, except for the big stuff like the heater/whole house fans/sewage grinder/dryer as they pull in too much current.

 

> BTW: Your initially posted understanding of the hysteresis is incorrect.

 

Thanks - what you wrote on hysteresis my understanding as well, I must have garbed the previous post. Doh! Didn't know that holdoff is on time only. I'll check and experiment. ...

[Brian H]

I have an earlier hardware XTBM. It does show receiving an Insteon Signal but not very informative. It does show any power line noise and X10 signal strength.

I still have a few X10 Chime Modules, early Insteon Modules that allowed an X10 Address to be programmed into them and 16 button Palm Pads.

I have the XTB-IIR X10 coupler repeater that has over 20 volts X10 signals on the power lines. It knows Insteon and respects it. By not falsely decoding the end of an Insteon message as an X10 one. Causing the Insteon command to be corrupted.

[larryllix]

> BTW: Your initially posted understanding of the hysteresis is incorrect.

 

Thanks - what you wrote on hysteresis my understanding as well, I must have garbed the previous post. Doh! Didn't know that holdoff is on time only. I'll check and experiment. ...

ahhh. MY BAD!

 

I didn't see the subtract in your OP = trigger-hysteresis, and I didn't see anybody correct that.

Carry on!

[robl]

Thanks for the feedback, everyone!

 

Bottom line = I think the synchrolinks have serious issues and reliability problems. Devices that seemed to work fine changed over a couple months (and at different times with no other big changes to the house) to start giving false positives. Unplugging them seemed to help for a bit, and then some number of days later, it'd happen again. My wife started getting really cranky hearing "dishwasher finished" when we hadn't run it for a few days, or "clothes washer done" in the middle of a load. I tried many workarounds, such as having an off trigger a program and if it stayed off for at least 5 minutes, then trigger the action. That helped places where the holdoff wasn't long enough. But it didn't help with false triggers. Another program that looked for a false on of < 1 minute and then ignored it seemed to work, but then I found the washing machine really did go to 0 watts < 1 minute before the end of a cycle and then spin up again. So that stopped working. Bleh.

 

I bought a XTBM, measured signal strength from every outlet on every circuit of the house. I drilled holes behind the washer so I could put an extension cord up near the ceiling with a repeater on it. I tried a "noise wall" where I put a noise filter between the synchrolinc and the appliance AND between the syncrolinc and the rest of the system, and nestled a signal repeater in with it. I isolated just about everything in the house that was a signal sucker with ~10 different noise filters, and measured the signal improvements with XTBM.

 

So ... 40 hours down the drain, a bunch of $$, a few new holes in the house here and there ... and still problems with this device.

 

My recommendation is to stay away from Synchrolinc - maybe try it again once it becomes dual band.

 

The XTBM experience is worth another thread ... Jeff and JV digital engineering was great to work with!!! And his noise filters worked better than insteon's. For example, for my amazon alexa switching power supply, it was on a circuit that had very low signal strength, and the power supply was enough to push the signal strength too low for XTBM to measure. Insteon's noise filter didn't help, but his did. My dryer FRIED 2 insteon noise filters, but his 15A noise filter is still running fine on it. The XTBM is a great appliance to snapshot signal strength. I was able to see that the powerline coupler was causing too many issues reducing strength, which lines had very low signals, and played with moving the ISY to various locations and finally found one that reaches all of the house without a need for repeaters. The drawbacks of XTBM is that it's not very insteon aware. It doesn't understand the wireless repeaters, using them makes the circuit look like it's full of noise. So ... after optimizing the place with no repeaters, I'll ad a couple back in on the low signal areas for completeness.

 

XTBM also helped a lot to show me how the old single-band devices, like the IR transmitter, were signal suckers and causing issues. All my work to make the IR transmitters work better was for naught, they were limitations to the transmitter or the source device not being able to process signals quickly enough. 

[larryllix]

Over a few years I have also had problems with reliability of the SynchroLinc.

 

I am monitoring my dryer but I was detecting the dryer inside light bulb at 6W so I had it set to about 5W trigger with 2-3W of hysteresis. Probably not high enough for it to reliably detect the difference between the bulb On and the dryer completely Off. I was trying to do triple duty with notifications of the dryer door left open to avoid cold air leaks in the winter. 

 

After about four occasions  of factory resetting the beast, I finally gave up and went to just detecting the motor running  Much less touchy level detection for the SynchroroLinc at about 120W trigger and half that for hysteresis.

 

This was about 2 months ago now, so we'll see if it changes the reliability of the SynchroLinc. My guess is the dryer motor starting spikes just make the firmware blowup and misbehave. Setting should not affect the reliability then.

 

I have had this experience with a few Insteon devices now, and now know anything can happen. I am still convinced my OnOff plug-in module caused scenes to trigger in my house every few random nights for about a year. I always thought it was one of my programs I had long forgotten. Simply unplugging and plugging the module back in, resolved that one ...fingers crossed!!!! 

[larryllix]

Has this device just disappeared with another defective SH design?

 

My SynchroLinc has finally crapped completely out after two years of flaky service, needing to be power rest every few months.

 

I wonder if this is the same PSU cap problem that has plagued the PLMs. Since there is no replacement around I may need to attempt to get out my huge magnifying glass and soldering iron once more.

[robl]

Yes, the SynchroLinc is a terrible product. Good concept, horrible hardware. 3/6 of mine failed within 6 months. A couple more are still in use, but flaky. And ... no failures in my other 100+ insteon devices.

 

For triggering events like the wash is done or the dishwasher has finished, the constant false positives are a complete fail. And having to baby some devices by putting a $30 filter before it and another $30 filter after it is a fiscal nightmare.

 

A 3rd problem is it's the old single band (AC device) only. Signals aren't as reliable, and my XTBM sensor finds these old single band devices tend to put a lot more noise on the network hindering other insteon devices. A fail/fail.

 

As an alternative, I tried a small z-wave power puck. I'm only using on my huge TV at this point, but it hasn't had any false positives and didn't need the 2 filters. It only reports power usage in watts (corrected), so getting the same hysteresis isn't too hard with an ISY program. "if power < 25 watts, sleep X seconds, then invoke event". I like this mode much, much better than the Synchrolinc attempt to do everything. This way I can report instantaneous power usage in logs and easily profile the device, instead of having to second guess SynchroLinc and monitor it with a power meter. Much easier and simpler.

 

I've read that some z-wave power pucks can't handle large usage devices, like some washing machines, so you might want to check around if you go that route. Replacing the other devices is way down on my TODO list, as my wife still has bad memories of hearing "washing machine finished" and "dishwasher done" in the middle of the night (eventually hid in a program) and sometimes 6 times and hour when they weren't even being used ...

 

Rob

[larryllix]

Yes, the SynchroLinc is a terrible product. Good concept, horrible hardware. 3/6 of mine failed within 6 months. A couple more are still in use, but flaky. And ... no failures in my other 100+ insteon devices.

 

For triggering events like the wash is done or the dishwasher has finished, the constant false positives are a complete fail. And having to baby some devices by putting a $30 filter before it and another $30 filter after it is a fiscal nightmare.

 

A 3rd problem is it's the old single band (AC device) only. Signals aren't as reliable, and my XTBM sensor finds these old single band devices tend to put a lot more noise on the network hindering other insteon devices. A fail/fail.

 

As an alternative, I tried a small z-wave power puck. I'm only using on my huge TV at this point, but it hasn't had any false positives and didn't need the 2 filters. It only reports voltage, so getting the same hysteresis isn't too hard with an ISY program. "if power < 25 watts, sleep X seconds, then invoke event". I like this mode much, much better than the Synchrolinc attempt to do everything. This way I can report instantaneous power usage in logs and easily profile the device, instead of having to second guess SynchroLinc and monitor it with a power meter. Much easier and simpler.

 

I've read that some z-wave power pucks can't handle large usage devices, like some washing machines, so you might want to check around if you go that route. Replacing the other devices is way down on my TODO list, as my wife still has bad memories of hearing "washing machine finished" and "dishwasher done" in the middle of the night (eventually hid in a program) and sometimes 6 times and hour when they weren't even being used ...

 

Rob

Thanks for that information.I have never had a false positive with my SynchroLinc. I used an extreme hysteresis by checking what the SychroLinc thought was full load and lightest load.

However mine just failed to report off many times and I always figured it was too sensitive to detect the dryer internal lamp at about 6 watts.

 

You mentioned detecting voltage only then continued to mention logic using power figures in you ISY logic. Was this a mistake in expression?

[robl]

You mentioned detecting voltage only then continued to mention logic using power figures in you ISY logic. Was this a mistake in expression?

 

That was meant for the z-wave power puck - it reports current power usage to ISY. Since it can change frequently, that also requires a program in ISY to make a condition on the power value and to add a time delay before triggering. Then this solution easily replaces the Synchrolinc. 

[robl]

OK.  You stated " It only reports voltage" in your post #22.  That wouldn't detect appliance load.

 

Good catch, typo, fixed above! Thanks, I understand the confusion now.