A new 2441TH issue

[shannong]

I've been carefully monitoring this behavior, and I believe I've identified the problem specifically. I'm now very certain the issue is caused by noise/voltage spikes on the line coming from the HVAC system as a few others speculated.

 

I have a gas furnace. When it cycles on, first the circulation fan for combustion chamber turns on to flush any gases that might be present. Then the electric igniter begins sparking 3-5 times a second just before the gas valve is opened. Lastly, the blower turns on to begin circulating air through the ventilation system.

 

The Energy mode turns on/off when the electric igniter begins sparking. I have tested this about 15 times. Energy mode turned itself on/off three times during those tests. Each time was when the igniters were sparking. Energy mode has not turned itself on/off during any other test or scenario.

 

Since line noise is affecting the circuits of the thermostat, it seems a software fix won't be possible. Perhaps Smarthome can identify and provide an appropriate filter. Future devices produced would need to accommodate this known issue in the hardware.

 

For those who already own the thermostat, Insteon could provide an option to disable the Energy mode completely. It's really not a good feature to begin with. The idea is the Energy mode would be used each time you leave/return to save energy. However, the button for it is hidden behind the door of the device so it's not convenient to use since you need to press it multiple times a day after opening/closing the dorr. To be useful, the Energy button would need to be accessible on the outside like the Mode button which is very rarely used. The Mode button is probably used less than once a month yet is front accessible. Also, those buying the thermostat are obviously interested in home automation so they wouldn't be using the Energy option anyways but instead would have their home controller adjusting the set point.

 

It is my opinion that SmartHome has not acted responsibly or in the customers' best interest with these thermostats. The product reviews on their own site reference many problems with them including this issue. Forums on smarthome.com and other 3rd party sites for controllers (like this one) reference this problem with many saying they did contact tech support for Insteon about this issue. Several said they returned theirs due to the issues. There was even a SmartHome representative active on this thread (SteveL in Jan 2013) who was participating in the discussion and said he would get back with everyone but he did not. So there's seems little doubt that Smarthome is aware of this issue before now. My current thermostats were produced in Sept 2013 well after these issues surfaced.

 

I've sent almost the same message to Smarthome in reference to my active case and asked them to call me. I'll update those here with any new info.

[GreyFox]

Excellent, shannong!

 

However, I would expect that since there is no way in the 2441th to disable the energy saving mode, I would doubt that Universal-Devices is able to give us a way to disable it. However, Universal-Devices is able to give us a way to turn the energy saving mode on or off via the ISY. Hint, hint, Michel. And if they did that, we could programmatically have the ISY test to see if it is turned on, and if it is, to turn it off.

 

I'm not holding out hope that SmartHome will give us any relief.

[shannong]

However, I would expect that since there is no way in the 2441th to disable the energy saving mode, I would doubt that Universal-Devices is able to give us a way to disable it.

 

Currently the thermostat lacks a way to turn on or off Energy mode except through pushing the physical button so there's nothing ISY or anyone else can do to resolve this directly. You could monitor the temp change and change it back with the ISY, but I dislike that sort of workaround. You can also change the Energy variance to be a single degree rather than four to reduce the impact of this bug. Unfortunately, the variance can be set to zero.

 

It would be trivial for Smarthome to update the firmware to add a local or Insteon-accessible configuration option on the thermostat to disable the Energy option completely.

[GreyFox]

Shannong, if the ISY software can change the temperature setpoint, which is set on the 2441TH by pressing a button, and the ISY software can change the mode which is set on the 2441TH by pressing a button, why can't the ISY software toggle the energy saving mode, which (theoretically) is set by pressing a button?

[shannong]

Shannong, if the ISY software can change the temperature setpoint, which is set on the 2441TH by pressing a button, and the ISY software can change the mode which is set on the 2441TH by pressing a button, why can't the ISY software toggle the energy saving mode, which (theoretically) is set by pressing a button?

 

The analogy would be this similar question: If you have a modern car with Bluetooth on board and there's a button on the dash to tune the radio to preset Number 1, why can't you just use your phone to send a Bluetooth signal to the car to switch to the preset of button Number 1?

 

You just can't. There's no interface, programming or otherwise to make it work as it's currently designed. Is it technologically possible? Of course. But currently there aren't any facilities setup to make that happen.

 

Similarly, the 2441TH only allows Energy mode to be turned on/off locally with the button. [ unless you know the frequency of the line noise which causes it to change. ] The 2441TH does not send a "I just entered Energy mode" message either even though it send messages about other changes that occur locally. Just like it also only allows you to set the Energy variance to one but does not allow zero. That's just how they made it (unfortunately). The ISY team can't change that on their end.

 

One interesting note, Energy mode changes Heat and Cool set points at the same time when Energy mode is turned on/off. Those set point changes are sent out as individual Insteon messages. Detecting changes in Energy mode would be more easily noticed if looking for both changing at the same time.

[shannong]

Sadly, I haven't heard from anyone at Smartlabs, yet. I'm going to call them on Monday.

[GreyFox]

Shannong, would you please provide a wiring diagram of your workaround, showing both a before and an after. Thanks.

[shannong]

No diagrams. No surge suppressor work around to consider.

[larryllix]

I disconnected the original line from the furnace and pulled it out of the housing and connected one pigtail with the RJ-45 to it so it could plug into the surge suppressor. I used the 22AWG wires to make the pigtail because the thermostat wires are solid core 20AWG and won't fit into an RJ-45 mod. I then created a 3 ft pig tail also with an RJ-45 mod on one end to connect to the surge suppressor. I ran the loose end of this one into the furnace to connect to the screw terminals on my unit. Viola!

 

No warranties expressed or implied. Try at your own risk. YMMV.

 

I wonder if donut-style ferrite beads would help that by just running the lines all through a unit. It could eliminate any common-mode spikes and noise.

[shannong]

I wonder if donut-style ferrite beads would help that by just running the lines all through a unit. It could eliminate any common-mode spikes and noise.

 

Maybe.

 

The issue seems to be voltage spikes. For a ferrite bead to be effective, it needs to be made with the right material, wound the right number of times, etc to have an impact on the frequency that you're dealing with. No real way for me to approach that. I'm not an EE and don't have any diagnostic tools. I find using a surge suppressor an easier approach than buying a bunch of ferrite beads trying to find one that might work and/or finding the right number of them to connect inline with each other.

[larryllix]

I was thinking about getting one of the 2441ZTH remote/wireless units just to have a temperature sensor. Other than the usual Insteon lack of actual information (to even find out if it sends analogue data) I did notice this in the troubleshooting chart section of the User's manual.

----------

Problem:

When an AC adapter is used, the temperature and humidity settings are changing rapidly.

--------

Probable cause:

Noise from the AC adapter is affecting the Wireless Thermostat’s electronics.

----------

Fix:

Be sure that your AC adapter is outputting between 5 to 7.5 volts.

Change to a better quality AC adapter.

-----------

 

Now this unit is an independent, normally 3V battery operated unit and yet they have problems with AC ripple demanding you run a 5-7.5V walwart with it? This has to tell us something about the sensitivity of the electronics to power supply noise.

 

This is assuming the basic power supply is a similar design as the 2441TH unit. I suspect it would be.

[shannong]

 

Now this unit is an independent, normally 3V battery operated unit and yet they have problems with AC ripple demanding you run a 5-7.5V walwart with it? This has to tell us something about the sensitivity of the electronics to power supply noise.

 

This is assuming the basic power supply is a similar design as the 2441TH unit. I suspect it would be.

 

 

Not an encouraging signal for sure. They both seem to have power issues.

 

I not sure how much would be the same. It's my understanding that HVAC systems are based on 24V whereas this device is somewhere between 5V-7.5V. It also can switch from AC to DC and back.

[larryllix]

I not sure how much would be the same. It's my understanding that HVAC systems are based on 24V whereas this device is somewhere between 5V-7.5V. It also can switch from AC to DC and back.

 

I would love the see the printed circuit card for these things. Years ago it was reported a circuit board would be designed and built. Then a after refining, a tech would cut out components until the device stopped working. Putting that component back in the tech would continue until no more parts could be cut out and it would still function. Who know what process they used for this design.

 

Sensitive logic inputs typically "pull themselves up". Good engineering practice ties them to the plus power line or better yet through a higher value resistor. If that part is removed the prt functions the same except that the logic signal input usually functions the same. being connected to a foil line on the circuit board it becomes a small antenna and picks up stray emf/rf noise nearby creating guess what? Logic changes or even worse internal transistor insulation breakdown that acts ups in too many ways to mention. Now this device transmits RF so we do have an environment.

 

Strange while I was typing this I realised analogous situation: I picked up a Mr. Christmas Light sequencer last year for $20. This fall, full of piss and vinegar, I hook the thing up only to find out after it worked fine for about two hours, I have to cycle the power off and on about 10-20 times to make the thing start up to play its tunes. hmmmmm... guess what? After a night's worth of tracing circuit board foils, I find an empty spot where they forgot to install the capacitor that acts as a cold boot reset circuit to the CPU chip. I grab some old capacitor I found and solder it in. Works like a charm, everytime. Not a trace of previous solder there. Now I reassembled the thing. When I put the circuit board back into the case the capacitor I installed stops the circuit card from sliding down into it's plastic slot mount in the case. These jerks did exactly what I stated above and cut out the prototype's capacitor and it worked just fine (they didn't forget) on that unit, so the, (probably Chinese) kids manufacture 1,000,000 of them without the capacitor. Then it hits the retail stores for about $159 and users return them. I pick the thing up at WallyMart for $20 wondering why. Most users would have just thrown them out for the $20, a year later, and without a receipt.

http://www.mrchristmas.com/THE-LIGHTS-A ... P1839.aspx

 

My suspicions intensify with this SmartHome item. Too bad.

[GreyFox]

Shannong, if I understand you correctly you have two 4 conductor 22 AWG cables connected to an RJ45 plug on one end, and 5 of the total 8 conductors connected to the terminals in the furnace. You then make another 8 conductor patch cable of 22 AWG with one end terminated into an RJ45 plug and the other end gets connected (at least the 5 conductors that match the furnace side) to the wires leading to your thermostat. Am I correct?

[shannong]

Never mind. Discard the surge suppressor work around. I was doing further testing tonight and discovered I couldn't turn on the AC. Being winter and all it was last on the list.

 

When 'Cooling' would turn on, nothing would happen and sometimes the tsat would blank out/reboot. I tried removing the yellow wire from the surge suppressor and going direct with it. It worked fine but when I tested again with Heat the problem returned where the Energy mode would engage when the electric igniters kicked off.

[shannong]

Smartlabs is telling me that using an external 24V power supply separate from the furnace will alleviate the issue. I'm doubtful of that because MWareman had previously posted about that, but their going to send me one free to try.

 

I'll update when I receive and test.

[Brian H]

larryllix and anyone else that would like to see internal and external photos of many of the Insteon devices that use RF.

They are in the FCC Database. Along with test results for compliance to FCC requirements.

https://apps.fcc.gov/oetcf/eas/reports/ ... Search.cfm

In the Details choice after finding the device you are interested in.

 

Grantee Code:SBP

Product Code: 2441T for the thermostat. 2441ZT for the wireless thermostat. 2441V for the older dongle that plugged into the Venstar T1700,T1800 and T1900

 

If you just input the Grantee:SBP and search. It will find all of the devices in the database that are from Smartlabs. Including the hardware version 2 2443 Access Points, 2413 series PLMs, LED Light Bulbs, etc.

[larryllix]

larryllix and anyone else that would like to see internal and external photos of many of the Insteon devices that use RF.

Very nice! Thanks!

 

Is there some special permission required to see the documents? I cannot get access to them, only a listing showing their existence.

[Brian H]

I don't believe any special permissions are needed.

I pick Detail for the choice and then accept the warning that the FCC is not responsible if the vendors file have a virus.

My version of IE give me a page has mixed secure and insecure contents. I say no to display secure parts only.

Then the files are listed for that ID number.

I then click on the choice and let them open. I save them after they load. With a new name as all of the are called the same Get Appliacation Attachment.pdf

[larryllix]

I pick Detail for the choice and then accept the warning that the FCC is not responsible if the vendors file have a virus.

My version of IE give me a page has mixed secure and insecure contents. I say no to display secure parts only.

Then the files are listed for that ID number.

I then click on the choice and let them open. I save them after they load. With a new name as all of the are called the same Get Appliacation Attachment.pdf

I can get to the file list page with the lists of documents. I get no warnings or input requests and the list has no hyperlinks for the document text lines, just plain text.

 

FCC > FCC E-filing > EAS > List Exhibits Page FCC Site Map

[Brian H]

Are you looking at the Details or Summery list?

 

I do remember someone having some problems with getting the files, but I don't remember what the fix was.

[larryllix]

Are you looking at the Details or Summery list?

 

Bingo! I thought it was one link labelled "Details Summary".

 

Turns out it is two links "Details" and "Summary"

 

Thanks!

[Brian H]

Glad you found them.

I have found some very informative.

[larryllix]

 

Maybe.

 

The issue seems to be voltage spikes. For a ferrite bead to be effective, it needs to be made with the right material, wound the right number of times, etc to have an impact on the frequency that you're dealing with. No real way for me to approach that. I'm not an EE and don't have any diagnostic tools. I find using a surge suppressor an easier approach than buying a bunch of ferrite beads trying to find one that might work and/or finding the right number of them to connect inline with each other.

Usually these things are not that "engineered". Since you think this I noise spikes coming for the furnace circuitry connected to the thermostat and it is easy to wire the wires into the thermostat I would try running each wire through a 1/4-1/2" steel hex nut and just see what happens. If it works somewhat try to land better ferrite cores to replace the hex nuts. Ferrite cores are not tuned but present impedance to signals that change faster. Since the signals are low frequency 60Hzand slow changing heat and A/C switching signals little effect will be seen on them. It's not rocket science and for a few bucks you may luck-out and have a solution.

http://www.digikey.ca/product-search/en ... ng/3408554

 

I would bet a HVAC supplier could even sell you a few commonly used by installers for other thermostats.

[shannong]

Usually these things are not that "engineered". Since you think this I noise spikes coming for the furnace circuitry connected to the thermostat and it is easy to wire the wires into the thermostat I would try running each wire through a 1/4-1/2" steel hex nut and just see what happens. If it works somewhat try to land better ferrite cores to replace the hex nuts. Ferrite cores are not tuned but present impedance to signals that change faster. Since the signals are low frequency 60Hzand slow changing heat and A/C switching signals little effect will be seen on them. It's not rocket science and for a few bucks you may luck-out and have a solution.

http://www.digikey.ca/product-search/en ... ng/3408554

 

 

I was thinking more of the wrapped core kind that are placed inline rather than just the beads where wires run though and confused the two in my head. I probably could harvest a bead that encases the wires by cannibalizing one from a keyboard or probably ten other things I have lying around the house. The first thing I'm going to try tonight is using an external 24VAC power supply.