ToggleLinc Dimmer - Failed Writing

[cranrob]

Hi cranrob,

 

Thank you again for the responses. Unfortunately they always breed additional questions

 

 

• [*:1ml364e5]No X10 repeaters, couplers, or boosters - just X10 switches.

 

You indicated that you can communicate reliably to your x10 devices with the PLM. That's difficult to do without some sort of phase coupling (passive coupler, booster, or repeater). It is possible (unlikely) that all of your X10 devices are on the same phase. The problem here is that you moved the PLM to the opposite phase. That should have impacted your X10 communication unless you have a coupler.

Believe it or not, by coincidence, all my X10 switches are on the same phase. When I move the PLM to the other phase, communication with the X10 devices barely works

 

1) What make/model X10 switches do you have? If Smarthome, they may have a built-in X10 booster (rather evil for Insteon).

2) Did you install the X10 yourself? If not, it's possible there is a coupler, repeater, or booster "hidden" somewhere in your system.

 

They are the rectangular push-button switches - I do not have model number handy. I installed X10 myself.

 

I got the ACPs last week. I was trying this without them previously. Without them, I was getting the "Failed Writing" message that I started this thread with.

 

If I understand the sequence correctly:

1) You initially had your PLM on the phase opposite your 4 insteon unit install. Without phase coupling you could not communicated with the Insteon units.

2) You installed the ACPs (established phase coupling) and things began to function somewhat.

3) You moved your PLM to the same phase as the 4 Insteon unit install.

True - when I moved the PLM to the same phase I was getting basically the same performance as with the ACPs, but of course my X10 devices stopped working, so I moved it back

.

 

What we haven't established is whether the PLM can communicate with the 4 units, without the ACP's, when it's on the same phase. Like Lee, I am having trouble believing this is a simple signal absorption problem. If you pull the ACP's and can still communicate with your devices (no worse than before) we should start considering "other options".

 

Lee brought up a possibility in his last post that we haven't specifically addressed: Are all of the 4 Insteon units tied to the same Power/neutral feed? It's not at all unusual to have multiple Power feeds in a quad box. It's also not unusual for them to be on opposite phases.

There is a single black/white pair bringing power to the box. The remaining wires go to the fixtures.

[cranrob]

Last night, I was able to write links to all the devices by airgapping all but one device and doing a "write updates" to the non-airgapped one.

I am now seeing very interesting behavior when operating the switches themselves:

 

- After links were written, I was able to turn the scene on and off with the controller (KPL on/off buttons).

- After several on/off cycles, the front pots stopped responding

- I airgapped the front pot switch, remaining lights then functioned normally

- I un-airgapped the front pot switch, and all worked again (for several cycles).

- After several cycles, none of the lights (except the KPL load) were responding

- I airgapped the non-working switches, un-airgapped them, and they started working again.

 

So there seems to be a communication issue even between the switches in the switch box. Airgapping/un-airgapping seems to fix it temporarily.

[IndyMike]

1) What make/model X10 switches do you have? If Smarthome, they may have a built-in X10 booster (rather evil for Insteon).

2) Did you install the X10 yourself? If not, it's possible there is a coupler, repeater, or booster "hidden" somewhere in your system.

They are the rectangular push-button switches - I do not have model number handy. I installed X10 myself.

The Switches sound like "X10 Brand" push-button units. I'm not aware of Smarhome making units in this style but I'm 100% certain. It you get a chance, please do have a look at one of the units for a model.

If I understand the sequence correctly:

1) You initially had your PLM on the phase opposite your 4 insteon unit install. Without phase coupling you could not communicated with the Insteon units.

2) You installed the ACPs (established phase coupling) and things began to function somewhat.

3) You moved your PLM to the same phase as the 4 Insteon unit install.

True - when I moved the PLM to the same phase I was getting basically the same performance as with the ACPs, but of course my X10 devices stopped working, so I moved it back

.

OK, things are making sense for me again. I'll put the "Booster Theory" to bed.

Lee brought up a possibility in his last post that we haven't specifically addressed: Are all of the 4 Insteon units tied to the same Power/neutral feed? It's not at all unusual to have multiple Power feeds in a quad box. It's also not unusual for them to be on opposite phases.

There is a single black/white pair bringing power to the box. The remaining wires go to the fixtures.

Understood. With only one power/neutral input to the box we can eliminate a number of possible problems. I'm sorry this is painful but, if we can cross enough likely candidates off the list we'll hopefully back into a solution.

[IndyMike]

Last night, I was able to write links to all the devices by airgapping all but one device and doing a "write updates" to the non-airgapped one.

I am now seeing very interesting behavior when operating the switches themselves:

 

- After links were written, I was able to turn the scene on and off with the controller (KPL on/off buttons).

- After several on/off cycles, the front pots stopped responding

- I airgapped the front pot switch, remaining lights then functioned normally

- I un-airgapped the front pot switch, and all worked again (for several cycles).

- After several cycles, none of the lights (except the KPL load) were responding

- I airgapped the non-working switches, un-airgapped them, and they started working again.

 

So there seems to be a communication issue even between the switches in the switch box. Airgapping/un-airgapping seems to fix it temporarily.

 

Now that's a handy piece of information. By operating the Scene from the KPL you've effectively eliminated the PLM and signal loading over the 25' to the panel. Nice work.

 

I understand that the TL's would stop responding to the scene until you airgapped them.

Question: After they stopped responding to the Scene, did you try operating the TL's manually (prior to airgapping)?

 

If they don't function manually, they may get getting spiked by the load. This is odd, since you are running incandescent loads, unless you have additional "protection" in your can fixtures.

 

You can eliminate the question of the fixtures by disconnecting the loads and "capping" the red output wire.

[ELA]

I have just reviewed your entire history once again to see what else could be in question.

You listed all the loads in one post and these add up to quite a large load. If all devices are on one circuit, and all in the same 4 gang box, that makes for a rather large bundle of wires.

Have you double checked your wire nutted connections for connection integrity?

 

When you did testing with all devices non air gapped, and only activated one load, we're the other loads left in the on or off state?

 

If you own a volt meter can you measure the voltage at the 4 gang , or one of the loads with all loads turned on?

 

One other interesting bIt of information that could help is to run the following test if you are willing.

 

Leave all load switches in the off state, so all loads off. All non air gapped.

With level 3 diagnostics on, poll each switch for a get engine command, repeatedly.

Wait about 5 seconds between each poll. Then post or review the results.

[cranrob]

Question: After they stopped responding to the Scene, did you try operating the TL's manually (prior to airgapping)?

Yes, they work manually after they stop responding to the scene.

 

You can eliminate the question of the fixtures by disconnecting the loads and "capping" the red output wire.

I will try that - actually it would be easier for me to just unscrew the bulbs rather than futzing with the switch wiring.

[cranrob]

Have you double checked your wire nutted connections for connection integrity?

Actually, I completely rewired the switches awhile back to eliminate the possibility of bad connections. Instead of using wire nuts, I used these this time: http://www.idealindustries.com/prodDetail.do?prodId=in-sure&div=0 - they take up much less room than wire nuts when dealing with large bundles of wires, and give positive confirmation of a good connection.

When you did testing with all devices non air gapped, and only activated one load, we're the other loads left in the on or off state?

I've tried both.

 

If you own a volt meter can you measure the voltage at the 4 gang , or one of the loads with all loads turned on?

 

One other interesting bIt of information that could help is to run the following test if you are willing.

 

Leave all load switches in the off state, so all loads off. All non air gapped.

With level 3 diagnostics on, poll each switch for a get engine command, repeatedly.

Wait about 5 seconds between each poll. Then post or review the results.

I can try these ideas. I have a voltmeter.

[cranrob]

I've done a little bit of more testing, and I've found that when a KPL or TL stops responding to the scene (whether from the ISY or the KPL), airgapping and then un-airgapping the device in question gets it working again for several on/off cycles.

[Michel Kohanim]

Hi cranrob,

 

To me, that would indicate a bad switch which needs to be replaced.

 

With kind regards,

Michel

[cranrob]

To me, that would indicate a bad switch which needs to be replaced.

I had a feeling that was the problem all along. Now, any advice for attempting to identify the bad unit(s)?

[Michel Kohanim]

cranrob,

 

Not really except for out of ordinary behavior such as something not turning on/off or blinking or staying on regardless of what you do.

 

With kind regards,

Michel

[IndyMike]

To me, that would indicate a bad switch which needs to be replaced.

I had a feeling that was the problem all along. Now, any advice for attempting to identify the bad unit(s)?

 

Hello cranrob,

 

That's what we're attempting to do with the tests we're asking for.

 

1) You've already eliminated the PLM by determining that the scene members exhibit problems when activated by the KPL.

 

2) By removing the loads from the units you will be removing the possibility of absorption or impulse spiking due to the attached loads. You had indicated that you could remove the bulbs from the fixtures and re-try testing. Removing the bulbs will eliminate any series wired protection in the cans. If there is EMI filtering, it will remain (this would be odd to find in a can). After looking at the push-in connects you are using, I understand why you do not wish to remove them. Doesn't appear to be a release mechanism.

 

3) All dimmers produce impulse noise - this is normally tolerated by both Insteon and X10 communication because the data transmission occurs at the powerline zero crossing, well away from the impulse. Sufficiently high impulse spikes can "upset" the micro in Insteon units. This normally renders them totally inoperative (manual operation included) requiring a air-gap to reset. Your 480W incandescent load will produce a fair amount of repetitive impulse noise. I would not expect it to upset the unit unless there were "additional circuity" in the cans or a bad connection to the unit (this is what ELA was alluding to).

 

Other thoughts:

1) At present, your PLM and X10 units are on Phase "B" (my designation) and operating well. Do you operate these units often enough to determine if they are operating 100%, or just most of the time? In other words, are you sure that your Phase "B" installation is performing well?

 

2) Your Insteon units are on Phase "A" and require frequent resets. It's possible that you have a noise source that periodically activates and interferes with communications on Phase "A". We would be looking for a 120V source since phase "B" appears OK. Insteon units utilize "collision avoidance" - If they see something in the 120 KHz range near the powerline zero crossing, they will interpret the powerline as "busy" and wait for it to clear prior to transmitting.

 

 

If the "bulb removal test" doesn't produce positive results, I would ask that you remove you problem TL from the box and wire to an appliance cord (I think Lee inquired about this previously):

1) Test first on phase B (near the PLM), and if OK, then on Phase A (near the ACP). Perform device link table reads or restores.

2) If the TL doesn't operate near the PLM, we'll call it defective. If it doesn't operate on Phase A (opposite phase), well need to hunt for noisemakers.

3) If it operates in both locations, you have a problem in your quad box.

[LeeG]

cranrob

 

What are the positions of the ToggleLinc Dimmers and KeypadLinc Dimmer (?) in the 4 gang box and the loads on each device

 

1, 2, 3, 4

 

1 - type device - load wattage

2 - type device - load wattage

3 - type device - load wattage

4 - type device - load wattage

 

The dimmers 1 & 4 have to be derated 200 watts, dimmers 2 & 3 have to be derated 400 watts

[cranrob]

cranrob

 

What are the positions of the ToggleLinc Dimmers and KeypadLinc Dimmer (?) in the 4 gang box and the loads on each device

 

1, 2, 3, 4

 

1 - type device - load wattage

2 - type device - load wattage

3 - type device - load wattage

4 - type device - load wattage

 

The dimmers 1 & 4 have to be derated 200 watts, dimmers 2 & 3 have to be derated 400 watts

 

Interesting! I didn't see that originally in the documentation, but now I do. I'm definitely over the rated wattage for devices 3&4 the way they are positioned:

 

1: KPL - Rear Pots - 240W

2: TL - Rear Chandelier - 160W

3: TL - Front Chandelier - 480W

4: TL - Front Pots - 420W

 

I'm going to remove bulbs to get wattage way down and see how things are. If that fixes it, I will replace TLs with the 1000W SwitchLincs.

 

Fingers crossed.

[IndyMike]

 

Interesting! I didn't see that originally in the documentation, but now I do. I'm definitely over the rated wattage for devices 3&4 the way they are positioned:

 

1: KPL - Rear Pots - 240W

2: TL - Rear Chandelier - 160W

3: TL - Front Chandelier - 480W

4: TL - Front Pots - 420W

 

I'm going to remove bulbs to get wattage way down and see how things are. If that fixes it, I will replace TLs with the 1000W SwitchLincs.

 

Fingers crossed.

 

Lee and ELA are exactly correct that your install is above the derating guidelines. You probably don't want to leave this as your final configuration. Unfortunately, I don't believe this is the "cause" of your communication problems.

 

Smartlabs established the derating based on the heat flux surrounding the switches and worst case conditions:

1) Insulated wall

2) Plastic or other non-conductive J-Box

3) Sealed wire entrances

4) Maximum (100%) on level on all dimmers.

 

Even with the above conditions, it would take time for the ambient temperature of the J-Box to reach a level that was detrimental to the dimmers.

 

Heat is generated by the Triac output in the dimmer. The heat generated is roughly proportional to the power delivered to your lamps. If you program your units to operate at 50% output level, they will generate roughly 50% heat.

 

From what I can see of your logs you are operating the dimmers at the following pre-set levels:

1: KPL - Rear Pots - 240W: Programmed level 7F (50%)

2: TL - Rear Chandelier - 160W: Programmed level 7F (50%)

3: TL - Front Chandelier - 480W: Programmed level FF (100%)

4: TL - Front Pots - 420W: Programmed level 7F (50%)

 

For testing purposes, the only unit that I would question would be 3) TL with the 480W load at 100%. This might be a problem if you left the units in this condition for a period of time. The other 3 units should be OK at these levels. Unfortunately, you are seeing problems on all units.

 

Again for testing purposes, you could reduce the on level of unit 3 to 50%. If you want to go the extra mile, pull the face plate to allow convection cooling (assuming you don't have young children).

 

I do want you to pull the bulbs on your Can lights as a loading test. I do not want this to be confused with a wattage problem at the switch. In the end, I think you will need to remove the units and attach them to an appliance cord for testing as I described earlier.

 

Final comment: Take a good look at the instructions for the 1000W dimmers. Installing these in a multi-gang box will require you to remove fins and further derate the units. If you adhere strictly to the guidlines, I don't think you can make this install work with the loads you have. I would rather you install a relay unit in place of unit 3) (which is running at 100%) since these produce no waste heat and require no derating.

 

IM

[cranrob]

Thanks IndyMike,

 

I'm actually running the front chandelier at 50% as well - you probably saw a log where I reset the TL and didn't completely configure it.

 

I don't have SwichLincs, so I wasn't aware of the requirement to remove fins. Thanks for the heads-up.

 

I will do the test you recommend soon.

 

Thanks again.

[IndyMike]

cranrob,

 

Not sure if you saw CopyRon's post here 15 Devices Won't Communicate

 

He had a X10 TM751 transceiver that was transmitting and preventing Insteon communication. These are rather nasty because they do not have a powerline receiver and hence no collision avoidance. They blindly place X10 on the powerline without regard to existing traffic.

 

Have any installed?