Jump to content

Problem after power failure


ergodic

Recommended Posts

I have an ISY 99i/Pro running 2.7.6. The ISY is powered from the PLM. The rest of the network is on a UPS (connected through a filterlinc).

 

Anyway, some drunken halfwit ran his truck into a power pole out here in the canyons a few days ago. As a result we have had several power outages lasting minutes to hours as the power company tries to fix everything.

 

Each time the power comes back on, after the ISY comes back up, it can't communicate with any of the Insteon devices. It is talking to the network fine, and it appears to be running fine, but a query on anything just generates the "lost communication with..." dialog after it gives up trying. I click the ISY reboot button, the ISY restarts, and all is then well.

 

Is there a way to figure out what might be going on? Would connecting the ISY to a UPS-provided power supply fix this? I'm sort of stumped - I can't imagine that the ISY comes back online faster than the PLM initializes, so what could even be the reason for this?

Link to comment

Yeah, I've read that thread. It's hard to see how a backup p/s is a fix for this -- for one thing even a large UPS will run out of juice sooner or later. And this happens every time, whether the ISY's been off for 6 minutes or 6 hours.

 

The PLM is V72 - not that old - got it about 6 months ago when the last one died.

 

I'm going to try V2.7.9, I've been holding off to see if it achieves release status, but this problem is starting to cause social unrest around here.

Link to comment

Hi ergodic,

 

I am so very sorry to hear about these problems. To be honest, I think your PLM is dead (cannot communicate with the rest of your devices). Moving to 2.7.9 is not going to help.

 

Indications that the PLM is dead:

1. Going to Tools | Diagnostics | PLM Status/Info shows disconnected

2. The RX light remains lit on your ISY

 

With kind regards,

Michel

Yeah, I've read that thread. It's hard to see how a backup p/s is a fix for this -- for one thing even a large UPS will run out of juice sooner or later. And this happens every time, whether the ISY's been off for 6 minutes or 6 hours.

 

The PLM is V72 - not that old - got it about 6 months ago when the last one died.

 

I'm going to try V2.7.9, I've been holding off to see if it achieves release status, but this problem is starting to cause social unrest around here.

Link to comment

The PLM most definitely is not dead.

 

After a power fail I can manually reboot the ISY (not the PLM) and everything then works fine. It does not show disconnected at any time.

 

The only problem is with communications following a power failure - which seems to necessitate a manual restart of the ISY.

 

I wouldn't dispute that the problem might be with the PLM if you think that's likely - I'll take the issue up with SH in that case. The PLM isn't that old. OTOH, I've never tried re-plugging the PLM after a power fail instead of the ISY. But I wouldn't know what to do with the information no matter which way that test turns out.

 

I screwed up my courage and installed 2.7.9, but I haven't had a chance to test power-fail recovery yet.

Link to comment

Hi ergodic,

 

Would you be kind enough to try this procedure (basically simulating a power failure):

 

Unplug your PLM (which removes power from ISY) and then plug it back in. Does this work?

 

Also, after power failure, is it the case that you cannot communicate with ANY of your devices or just a certain few?

 

With kind regards,

Michel

Link to comment

Well Michael, as is often the case with Insteon the plot thickens....

 

Tonight I noted that the ISY was having trouble communicating with some devices (roughly about a third, though that's just a guess.) Queries would either work after several tries, or just fail completely. Some devices had no problem at all.

 

So... following your suggestion, I unplug the power strip to which the PLM (and thus the ISY) are powered, and then plug it back in. Presto! everything starts talking again and no problems. I can query everything and it all responds more or less instantly.

 

Any ideas? Perhaps a voltage spike during one of the power failures did some harm? But then how could it just start working properly again? (Are these things protected internally?)

 

Just FYI, the main power strip is one of those 6' wiremold affairs mounted on the garage wall: it has an access point, an EZX10RF, a FilterLinc (running into the cabling cabinet UPS), and a pigtail to a power strip also running inside the cabling cabinet that has only the PLM on it. The ISY is powered off the PLM cable. I've had no problems with this config up until this stuff started happening....

Link to comment

Hi ergodic,

 

Thanks so very much for the test. My premature conclusion is that one of those devices on the powerstrip is the cause for your troubles. So, the next time this happens I would unplug those devices (excluding the PLM) one at a time to see if there's any improvement. If none, then the problem is either the PLM or ISY (I tend to think it's the PLM but only because of past experiences).

 

With kind regards,

Michel

Link to comment

OK. Had another power failure tonight. Things came up wonky - ISY couldn't talk to any of the devices. Just timed out after its retries and marked the device with "!".

 

I removed the other devices (EZX10 and AP) from the power strip which had no effect. I plugged those back in, then re-powered the ISY but not the PLM and that cured it.

 

The one test - an important one I think - I haven't tried is to re-power just the PLM and not the ISY, Something I currently can't do because the ISY is powered off it. I'll get some sort of DC adapter rigged in there for the ISY and then try a simulated power failure at the breaker box.

Link to comment

Hi ergodic,

 

I think our choices are pretty limited: it's either ISY or the PLM. After power failure, ISY and the PLM are powered up simultaneously and thus they should work immediately. Now, if the power circuitry in the PLM is defective, when they are both powered on, the PLM might not have enough juice to power up ISY while it's going through its start up sequence.

 

Of course, the best test would be to either replace the PLM or ISY and see if the problem is solved. If you do not have an extra PLM, please contact us and we'll send you a replacement (refurbished) ISY to test with.

 

With kind regards,

Michel

OK. Had another power failure tonight. Things came up wonky - ISY couldn't talk to any of the devices. Just timed out after its retries and marked the device with "!".

 

I removed the other devices (EZX10 and AP) from the power strip which had no effect. I plugged those back in, then re-powered the ISY but not the PLM and that cured it.

 

The one test - an important one I think - I haven't tried is to re-power just the PLM and not the ISY, Something I currently can't do because the ISY is powered off it. I'll get some sort of DC adapter rigged in there for the ISY and then try a simulated power failure at the breaker box.

Link to comment
  • 2 weeks later...

Well, I started having sporadic comm problems even without the power failures in the last week. Resetting the ISY and PLM usually clears it up for a while and then the nonsense starts again. The ISY will control a device fine from the console and then start having trouble with a query. I'll query it a few times and it will start responding again. And so forth.

 

With nothing else to try, I decided tonight to reprogram the PLM (not replacing it). The first two attempts failed after a short time (the second attempt I think gave -20000 as the error). So I power-cycled the PLM and ISY.

 

The next attempt failed about halfway through: the event log shows it was trying to program group 40 for a device and the next line is "Failed: Cannot write PLM control link for group 40". So, ever the optimist, I tried yet again. This time the ISY's PLM reprogram ticked along, setting group 0 and 40 for each device until it was done.

 

As I wasn't swapping out the PLM I didn't put the MS and RemoteLincs into linking mode (that's a huge undertaking here) - I assume that was OK?

 

I also noticed a few odd "UNKNOWN" lines in the event log while it was reprogramming the PLM: "[uNKNOWN] 02 0A" and "[uNKNOWN] 02 08" What's the deal with those?

 

Anyway, things are now working again for the moment. What does all this add up to? Is this symptomatic of a failing PLM, or something else?

Link to comment

Yet another tidbit: I discovered after the PLM reprogram that the ISY was not seeing device status changes. The devices were working and controllable from the ISY, but changes at the switches weren't being seen by the ISY.

 

So I did a "restore device" on one, and it then started working. I've gone through about a third of the devices in the ISY now one by one, and as I restore them they reconnect. (There's still a small right-click issue in the 2.7.9 GUI btw).

 

On a few of these, I see the "[uNKNOWN]..." message as it restores the device, and these are the ones that seem to take a while. For example, "[uNKNOWN] 02 5D". In that case there is a device "0F 02 5D" in the scene so I assume the address is truncated or corrupted somehow, but where would this be coming from?

 

If I figure out which devices those are and remove it and all it's scenes and reprogram it from scratch is that likely to remove the [uNKNOWN] stuff? And will that even help?

Link to comment
If I figure out which devices those are and remove it and all it's scenes and reprogram it from scratch is that likely to remove the [uNKNOWN] stuff? And will that even help?

 

My experience is that restore devices eliminates the unknowns. Of course, this assume good communication during the restore process. I recall, too, that one can restore the entire system. This is something that could be run overnight. It might be a better option that restoring individual devices, one at a time.

Link to comment

I didn't replace the PLM, I just selected "Restore PLM" for the one I have.

 

As best I can determine it from the event log, the ISY didn't reprogram - or attempt to reprogram - the devices.

 

Possibly it "knows" the PLM ID hasn't changed, or possibly there was an error of some sort during the process that didn't display but caused it to abort.

 

I don't know why reprogramming the device links should even be necessary if the PLM isn't changed but it seems to be in this case.

Link to comment

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...