Fanlinc / 6 button switch strange issue

[bretta]

Not sure if this is an ISY issue or an Insteon hardware issue.

I have 8 ceiling fans, all running fanlincs controlled by 6 button switches.  I set each of them up with 5 scenes, light on/off, Fan low, med, high and off.

All work well.

I had a hardware issue on one fan, where when selecting high, the fan motor would sound like it was about to explode.  An internal issue in the fanlinc was screwing with the power output to the motor.

I just replaced that fanlinc with a new one.   I used the "Replace" feature in ISY admin console.   It didn't work well at all. In fact, ISY continually crashed each time I tried to do it.   In the end, I deleted the original fanlinc from the console, added the new fanlinc as a new device and then added it to the existing scenes.

Now I have an interesting issue.   Everything works exactly as designed, EXCEPT....  If the fan is running on Medium, with the "fan med" scene active, and I hit the "fan high" button, the fan stops.  Every other combination of buttons works as expected.  Fan low to fan med, or fan high, changes the fan speed accordingly, for example.  The only time it is an issue is if it is already running on medium, and I want to speed it up to high.    The green LED goes to red on the fanlinc and the fan motor stops.  Pressing high again turns the fan on high.    The LED for the High button on the switch illuminates as expected, but the fan motor shuts off.  Going from any other condition, to fan high, works as expected. It's only going from medium to high that there is an issue.

All scene attributes are identical throughout the scenes, for this fan, and compared to all the other fans that work fine.   Buttons are all set to toggle.   I even tried deleting the entire scene and rebuilding it from scratch.  No change.

Screen shot of the scene for high speed.

Any wizards out there have a thought on this one?  
 

[bretta]

Not a fix, but a work around.

I just discovered that by going into Advanced Properties and setting Device Communication to "2 Retries" will get the fan onto high speed.

Watching the LED on the fanlinc, pushing the button still turns the fan off, the original problem, but then the ISY sends the 2nd "retry" and that turns the fan on high as desired.  Perhaps a 2 second delay in the overall reaction but achieves the desired result.

Still curious as to thoughts on the actual issue.

Thanks.   Brett

[jec6613]

If it's not getting the first command, it's due to network noise almost always.  Something on your electrical is throwing noise and causing the troubles.

I have that issue for one switch, and for the life of me I can't figure out why just that one since the other one of the 3-way pair works 100%, but if I increased the retries it works fine.

[oberkc]

While this is not necessarily consistent with your clues, I would certainly look at the scene definition a little closer.  (I assume you are on v4.X software, so I am having a little trouble remembering the details of the admin page, compared to v5.0 software.)

Please confirm: 

- that the fanlinc button "kit east fan high" is defined as a "controller" in the scene.  (I cannot tell for certain from your screen shot.)  

- that the fanlinc button "kit east fan high" attributes is full on when tapped.  Is it possible that this button is commanded to turn to zero when toggled ON?

- that there is no program that is triggered by this "kit east fan high" button which might explain this odd behavior

I am also curious what happens when you command this scene to turn on from the admin console versus the button.  Do you get the same results?

Yes, there might be comms issues, but I am having trouble understanding why a comms issue would turn the fan OFF if scene definitions are correct.  

You could try showing a list of the links in this device and comparing it to the ISY records.  If they don't match, try restoring it.  

 

[bretta]

17 hours ago, jec6613 said:

If it's not getting the first command, it's due to network noise almost always.  Something on your electrical is throwing noise and causing the troubles.

That's not likely in this situation.  Every other command gets through correctly.   
There are 16 Insteon components in that room also, which helps with transmission.

[larryllix]

Possibly a bad fan capacitor arcing through, and disrupting the initial Insteon signal response.

[bretta]

22 hours ago, oberkc said:

While this is not necessarily consistent with your clues, I would certainly look at the scene definition a little closer.  (I assume you are on v4.X software, so I am having a little trouble remembering the details of the admin page, compared to v5.0 software.)

Please confirm: 

- that the fanlinc button "kit east fan high" is defined as a "controller" in the scene.  (I cannot tell for certain from your screen shot.)  

- that the fanlinc button "kit east fan high" attributes is full on when tapped.  Is it possible that this button is commanded to turn to zero when toggled ON?

- that there is no program that is triggered by this "kit east fan high" button which might explain this odd behavior

I am also curious what happens when you command this scene to turn on from the admin console versus the button.  Do you get the same results?

Yes, there might be comms issues, but I am having trouble understanding why a comms issue would turn the fan OFF if scene definitions are correct.  

You could try showing a list of the links in this device and comparing it to the ISY records.  If they don't match, try restoring it.  

 

Couldn't work out how to split the quote of your post, so answers in the same order as your questions/comments.

"kit east fan high" is the controller in the scene.  Confirmed.

Attribute for that button is fan on high speed. Confirmed

No programs even remotely related to any fans in the house.  Confirmed

Didn't think of trying the scene from the console.  Just did.  Instant response.  1 click of "On" in the scene, fan came on high.

Exactly my point.  If it were comms issues, it wouldn't react.   Instead, it turns off.  Plus the fact it only happens from Med to High, no other combination of actions has an issue.   With over 300 Insteon devices, I've had my fair share of comms and cross comm issues.  This doesn't fit any of those.

Not sure if there's another way to check links, but when looking at the fan motor in the admin console, it shows membership as a responder to the 4 buttons on this switch, nothing else.  

Thanks for your thoughts.  I'm open to any others you may have after reading my responses.
Brett

[bretta]

9 hours ago, larryllix said:

Possibly a bad fan capacitor arcing through, and disrupting the initial Insteon signal response.

That's an interesting thought.  In theory, the capacitor would see a larger surge on high speed startup, but the fan doesn't even have the chance to try.  As soon as you hit the button on the switch, the LED turns red.    A second push of the button and the fan starts up.  

This fan has so much torque that, from a dead stop, it twists the body of the fan assembly with an audible clunk before motion is transferred to the blades. A small amount of slack in the components, I assume.   
To me, it would draw a lot more power starting from a stop, than increasing from medium speed, and that is when the issue occurs.

[larryllix]

41 minutes ago, bretta said:

That's an interesting thought.  In theory, the capacitor would see a larger surge on high speed startup, but the fan doesn't even have the chance to try.  As soon as you hit the button on the switch, the LED turns red.    A second push of the button and the fan starts up.  

This fan has so much torque that, from a dead stop, it twists the body of the fan assembly with an audible clunk before motion is transferred to the blades. A small amount of slack in the components, I assume.   
To me, it would draw a lot more power starting from a stop, than increasing from medium speed, and that is when the issue occurs.

The only hole in my idea is that on high, there isn't likely to be a capacitor in series. However on your theory the fan on high would generate a much larger voltage spike starting from a standstill. 

This may take some spike suppression across the motor to kill it before it destroys the fanLinc also. At least an experimental install to help find the problem. I would try a 130vac MOV of 10,000 joule rating or more and twist it into the fan motor to fanLinc connections, leaving it hanging in the air for the trial. Large heat shrink over the whole live part and leads later, if permanent install works.

[oberkc]

2 hours ago, bretta said:

Not sure if there's another way to check links

There is.  I recall it available from the right-click on the device>>>show links.  This reads the links directly from the device.  Once shown, you would have an option to "compare" to the ISY records.  They should match.