UPB severe problems started yesterday

[Chakara]

I say this hesitantly as it hasn't happened for about 8 months - but we used to get noise spikes that I am completely convinced came from outside the home. They would prevent the UPB from traversing phases in the house. Quite annoying, and unpredictable.

Now I knock on wood

-Chak

 

[ano]

UPB is transmitted as a series of precisely timed very fast rise-time pulses. Unlike X-10 that puts a amplitude modulated frequency on the powerline, UPB puts very large "noise" spikes. Although they don't look like high-frequency noise, they can be filtered out like high-frequency noise. Spike protectors aren't typically a problem because they just clamp down at voltages over 170V, which is the typical peak of a 120V RMS. Since UPB pulses occurs near the zero crossing, and they don't exceed 170V, they aren't clamped by spike protectors.

A UPB filter is a low-pass filter, meaning it not only filters out any noise spikes which could confuse UPB, it would also filter out UPB spikes, and therefore isolates UPB communications. In addition any device which would dampen out UPB pulses (or suck the signal) would also be isolated. So UPB filters not only block UPB signals, and noise that could bother UPB, but they also block potential UPB signal suckers.

PCS makes UPB filters up to about 20A at 240V, but any low-pass powerline filter should work as a pretty good UPB filter, and X-10 filters works as pretty good UPB filters as well.

 

[Basildane]

Thanks for all the good info everyone.

Just for the sake of learning, I did some more testing and I wanted to share what I found.

I have a typical split phase house, with a passive UPB coupler.
I disconnected the coupler.

Now the results are really interesting.
If I connect the PIM to phase A, I can hit the devices on phase A, but almost nothing on phase B.
If I connect the PIM to phase B, I can hit all the devices on both phases reasonably well.

This is with the UPS online.

If I turn the coupler back on, everything goes back to zero.

Leave the coupler on, and disconnect the UPS, and I get full signal to both phases again.
Its like the coupler is using the UPS to null the UPB pulses.

Again, this never happened until this week. We had 100% perfect operation for 2 years.

So right now, I have moved the PIM to phase B, cut off the coupler, and ALL devices (except the 2 that fried) are working reasonably well.

 

[IndyMike]

Thanks for all the good info everyone.

Just for the sake of learning, I did some more testing and I wanted to share what I found.

I have a typical split phase house, with a passive UPB coupler.
I disconnected the coupler.

Now the results are really interesting.
If I connect the PIM to phase A, I can hit the devices on phase A, but almost nothing on phase B.
If I connect the PIM to phase B, I can hit all the devices on both phases reasonably well.

This is with the UPS online.

If I turn the coupler back on, everything goes back to zero.

Leave the coupler on, and disconnect the UPS, and I get full signal to both phases again.
Its like the coupler is using the UPS to null the UPB pulses.

Again, this never happened until this week. We had 100% perfect operation for 2 years.

So right now, I have moved the PIM to phase B, cut off the coupler, and ALL devices (except the 2 that fried) are working reasonably well.

Hello Basildane,

What strikes me in the above is that adding your coupler (with the UPS online) appears to bring signals down to zero. It shouldn't be possible for the UPS to load your PIM through the coupler. The couplers are good, but they will exhibit some impedance - this impedance should allow some signal level at the PIM.

You mentioned that you have never noticed a noise indication through Upstart. I'm suspicious that this feature is not functioning properly for you and that your UPS is putting excessive noise on your powerline.

Here's the scenario:
1) Your UPS incorporates a EMC filter (Electromagnetic Compatibility) at it's input. Depending on how this filter was implemented it may or may not have been a signal absorber. Nonetheless, the filter prevented the UPS from placing noise on the Powerline.

2) Your electrical event exceeded the voltage rating on the UPS filter capacitor - it's not functioning now and your are getting noise at the UPS switching frequency on the Powerline.

3) Without the coupler installed, the UPS is putting noise on one phase of your powerline. When you switch the coupler back in, the noise is also coupled and interferes with your UPB.

Your other results experimenting without the coupler (phase A vs phase B) will depend on the noise frequency from the UPS and the location of the PIM relative to the UPS. If the noise source is between the PIM and your load panel it can easily swamp communications. Line length can be your friend here since the noise will attenuate as a function of distance from the noise source.

Best I can come up with at the moment,

IM

 

[priorityelectric]

I have a house with about 20 UPB devices. Everything has been perfect for years. I have an HAI phase coupler (split phase in my house).
Woke up yesterday and 15 of the devices were not responding to commands. They do operate normally locally. At the same time I found a circuit breaker tripped. I also have a Leviton whole-house surge suppressor.

So, I loaded Upstart and got zero response from most of the devices. To make a long story short, I cut off breakers until I got to my office UPS, which is an APC 2200 RM.
When I cut THAT off, the signal jumped back up to excellent on almost every device. (2 of them are damaged and will not reset).

The UPS seems to be sucking the UPB signal down to zero. Upstart says "zero" for noise. I've never seen any reading for noise. Does that feature even work?
Anyway, this UPS has been running for over 5 years. Longer than I've had the UPB equipment. And it went from perfect to unusable overnight.

My theory is that we had a spike that killed two of the UPB's, and damaged the UPS.
What to do? Isolate the UPS from the line? Would a "noise" filter work in this case?

Sounds like GEN 1 UPB or SA UPB and more than likely a SURGE.
I would replace the UPS if you still use it at all.
I would also consider GEN 2 UPB by PCS pulseworx
The reason no noise level is it is not noise it is attenuation aka signal suck.

 

[Basildane]

Update.

We still have unreliable UPB control. You never know what the status of the house is going to be.

Since my last post, I installed a 20 amp filter on the UPS. I get signal of 80 - 120 on most devices. Noise level is 0 - 2.

I have spent COUNTLESS hours running Upstart scans while turning on and off every appliance and electrical device in the house. Made no difference.
And to answer one previous post, all my devices are new, Gen 2, HAI devices.

What makes it really difficult is I can sit for an hour watching my laptop connected to the PIM, and everything works fine. Then, when I'm not looking, random commands fail.

One minute I get 80 on a signal to a room, then it is 2. Then it is 80 again.

Any new suggestions on troubleshooting? I really don't like the thought of ripping out 25 switching and starting over with something else.

 

[mikejrc]

as many others have suggested you should discard your UPS
i know it cost you a lot of money and it would be hard to replace but its causing you more harm than good now. They dont last forever and as someone stated, over time when they get hit their capabilities diminish and die eventually.

I dont mean to sound negative or rude so i apologize, this is one of the reasons i didnt choose UPB for lighting control, im an electrician and like someone mentioned, electricity is very unpredictable and subject to so many variables, from generation to distribution to grounding to whatever is out there, that marriage between AC and sending signals at the same time is a very difficult one, it may work for some but not all the time and not for everyone

I have worked in big projects (casinos) with a lot of dimmer panels and lighting controls and have never seen UPB implemented, Crestron and Lutron most of the time.

My theory is that your UPS took a big hit, one that was so fast and hard that it didnt even have time to log it.

 

[electron]

were you turning on / off appliances while analyzing your UPB network, or flipping breakers? I personally would turn every breaker off, and turn them on 1 by 1. I have had bad computer power supplies and a surge strips (must have gotten hit by a surge) cause problems for me before (back in the X10 days). Definitely worth a shot if you haven't tried this yet. I would also call the manufacturers of your UPB devices, and see if they are willing to help (they usually are).

 

[Basildane]

as many others have suggested you should discard your UPS
i know it cost you a lot of money and it would be hard to replace but its causing you more harm than good now. They dont last forever and as someone stated, over time when they get hit their capabilities diminish and die eventually.

I dont mean to sound negative or rude so i apologize, this is one of the reasons i didnt choose UPB for lighting control, im an electrician and like someone mentioned, electricity is very unpredictable and subject to so many variables, from generation to distribution to grounding to whatever is out there, that marriage between AC and sending signals at the same time is a very difficult one, it may work for some but not all the time and not for everyone

I have worked in big projects (casinos) with a lot of dimmer panels and lighting controls and have never seen UPB implemented, Crestron and Lutron most of the time.

My theory is that your UPS took a big hit, one that was so fast and hard that it didnt even have time to log it.

Ya, but as I said, I put in a filter in front of the ups and my levels are back up to 80 - 120.
I ran a test without the ups this morning and the levels were up another 10 or so above that, and that's it - not much.
I don't think that is the issue.

At this point I'm not going to drop another $700 on a UPS that may or may not change anything.
I'm looking for another possibility.

 

[Basildane]

were you turning on / off appliances while analyzing your UPB network, or flipping breakers? I personally would turn every breaker off, and turn them on 1 by 1. I have had bad computer power supplies and a surge strips (must have gotten hit by a surge) cause problems for me before (back in the X10 days). Definitely worth a shot if you haven't tried this yet. I would also call the manufacturers of your UPB devices, and see if they are willing to help (they usually are).

I did call HAI months ago, no help.

I am turning off breakers and testing, yes, but the problem is so random, it may go all day without a glitch.

I woke up this morning and 1/2 of things that should have happened, did not happen. Then I plug in a laptop and scan everything and its "excellent" readings across the board.

 

[cornutt]

Are you taking your measurements from the same place where your controller PIM is plugged in? If you are measuring from another place in the house, your measurements may not reflect what the controller is seeing.

One concern I have is that it sounds like your house got hit by a pretty good spike, and you may have some arcing occurring somewhere. A place to start would be to go to your panel and make sure all of the breakers are pushed down firmly onto the bus. Feel of them and make sure none of them are getting hot. If you are experienced with wiring, do this (or hire an electrician to do it): take the front off of the panel, make sure that all of the circuits are secure on the breaker terminal screws, and make sure that all of the neutrals and grounds are secure on their busbars.

Go around the house, unplugging and re-plugging things one at a time. If you find an outlet that is scorched, cracked, or doesn't hold the plug firmly, replace it. If you have a light switch somewhere that is arcing, replace it.

You might want to set up you laptop (again, plugged in at the same place as the controller) with Upstart and leave it running with data logging for a day or so until you see failures happening. Then you can look in the log and see which commands failed. That might help you correlate with something happening in the house at the times of day when the failures occur.

 

[mikejrc]

I strongly agree with cornutt
as I stated earlier its seems like you got hit pretty hard
you should start from the panel then working to each device at at time

 

[tmbrown97]

In UpStart you can set it to run and take measurements and recurring periods on its own... you could set it to take 1 reading of the entire house every 15 minutes for a week - then go back and look at when your signal was at its lowest and try to narrow down to a time of day, then think about what may have been happening (doing dishes, running a motor, etc). It'll export to CSV and you have to import it and manipulate it from there.

 

[Basildane]

Thanks for all the good advice.

A remarkable thing happened last night. I just happened to be standing next to the system when it failed.
I had my laptop and PCS backup pim ready to go.

I fired up Upstart and got excellent signal. No communication from HAI.
So I connected the laptop directly to the HAI pim. Dead.
For confirmation, I hooked up the PCS pim, perfect.
Back to the HAI - dead.

So, the HAI pim is defective.

 

[cornutt]

Ah, there you go. Whatever kind of spike hit your house must have fried it.
You might want to double check your controller and make sure all of your zones
and outputs are working. I took a lightning hit last spring and it killed the
output that was connected to the thermostat bus. I had to send the controller
board in for repair.