UPB weak signal

[dlmorgan999]

One thing I discovered fairly early on was that the SA Phase Coupler actually reduced the sig level to several switches.

That's an interesting discovery. I originally had an HAI coupler but replaced it with an SA one to try to make the signal stronger. I may swap it back and do some more testing. Today I also ordered a couple of the SA noise filters. I'll report back on if/how they help.

-- Dave

 

[webmtn]

Multiple topics have been touched upon here and a great suggestion was made by Ano. His suggestion is how we go about identifying circuits that attenuate UPB signals and also how we find UPB noise interferers. Let me comment on some of the other items in the thread.

1. When using Upstart, and you are seeing very low signals, then you can have intermittent operation of UPB devices. I don't like to see signals below 10 - 15. Having said that, I have one device in my house that consistently measures 7 lesters and it operates extremely reliably. Once in a blue moon, it will miss a signal, but it is very rare. I tell users of UPB if the signal gets below 10 -15, then they need to take steps to improve signal strength. There are multiple ways of attacking signal strength issues - phase couplers, phase repeaters, relocating PIMs, etc.

2. We have seen instances of UPS's attenuating UPB signals. Normally, this occurs when a UPB device was installed on the same circuit as the UPS and normally it needed to be a fairly large capacity UPS. Just a question on this - was your UPS installed in such a manner that it was electrically located very close to your breaker box? We have also seen AV eqpt, especially power conditioners, also attenuate UPB signals.

3. The SAI ZNF-10A filters will probably isolate the UPS. You might also try some of the Leviton X10 plug in filters if you have some laying about (they're only rated at 5A, so analyze your current requirements). We've had some success with these devices in filtering noise that interferes with UPB.

4. There are two types of phase couplers - inverting and non-inverting. SAI makes the inverting style and PCS / HAI offer the non-inverting style. SAI says that inverting phase couplers work better when the utility transformer is less than 200 feet from the home and they also say that you can put in multiple inverting phase couplers in homes with multiple breaker boxes. There is some evidence that installing multiple non-inverting phase couplers in the same home can be detrimental. If the transformer is more than 200 feet away, then a non-inverting phase coupler might work better. We've had very mixed results using phase couplers. In most cases, either style phase coupler will improve signals, but depending upon the installation, one might improve them more than the other. On occasions, we have had users try both styles and keep the one that works best for them. Sometimes, we have used phase couplers on hand that we'll use for this testing, so contact me if you want to consider this. I can't promise anything, but we can try. There was another comment on the thread about an inverting phase coupler actually lowering signal levels. We have seen a few instances of this also, but it seems that it impacted stronger signals and didn't affect lower signals much, if any. If it lowers strong signals some, but helps lower signals, then we'll accept that compromise.

5. A comment on the PCS SPR phase repeater. We have used these with great success and heartily recommend them for installations with large numbers of devices. They are quite expensive to use on smaller installations, but sometimes they might be very worthwhile. However, recognize that the repeater is really intended to work with Gen2 devices. The SPR will help Gen1 devices, but you won't be able to tell it using Upstart. (Just as a side note, the SPR needs to be enrolled in Upstart in order for Upstart to take advantage of it. On a couple of occasions, we have had dealers install repeaters into installations with Gen1 devices. We don't see any improvement using Upstart, but we see a definite improvement in the perception of how well the system is operating. There were changes made to the Gen2 core to make Gen2 devices work better with the SPR.

6. Another technique we have used with success when an intelligent controller is being used in the home deals with the location of the PIM. Oftentimes, the controller is located where it is convenient for the installer or the homeowner. We highly recommend locating the PIM device as close to the breaker box as possible. This reduces one layer of attenuation from the PIM location to the breaker box. Obviously, this does not help with the situation where you have a multibutton transmitter located a long distance from the breaker box trying to control a receiver on the opposite phase also a long distance from the breaker box. There are other ways of dealing with this situation.

Anyway, just some comments on means of dealing with low signal levels.

 

[dlmorgan999]

Anyway, just some comments on means of dealing with low signal levels.

This is excellent detailed information. Thanks very much!! I think my biggest issue is probably my main UPS. It is located in the same room as the PIM (but the UPS and the PIM are each on separate dedicated circuits). Both of them are only about 15-20 feet from the breaker panel. The main UPS is an APC 1500 (so somewhat large). I also have a smaller UPS in the same room and another small UPS in another room. I just ordered three ZNF10 noise filters and my plan is to use them to isolate the UPSs and then reevaluate.

I also have a couple of large TVs and a couple of heavy-duty receivers and I'm thinking these may be contributing to the problem. I may end up putting noise filters on those as well. I think I have an old X-10 noise filter lying around somewhere so I might try that on a lower power draw device.

I have a couple of devices that are sometimes as low as 3-4 on the device end. I have several more that are in the 6-7 range (in fact NONE of the 15 devices that are on the same phase are any higher than 9 on the device end). As you said though most of them work fine - I only have two or three devices that are intermittent. I also have one switch that's not being detected any more. It may have died but I figured I would wait until the noise filters are installed to determine if it needs to be replaced.

-- Dave

 

[ano]

O.K. A few things. As stated, with SA inverting phase couplers, you can add multiple ones to increase your signal, I use three. But this will only help when the "Other" phase is weak.

What you need is an isolator. These don't exist commercially as far as I know, but you can build one easy. It just so happens that I built one just yesterday. I didn't have a signal sucker, but rather a new Panasonic Inverter Microwave that generates so much noise, my nearby UPB lights fail to work.

O.K. the device. You will need a 100 uH +/-(Micro Henry) choke rated for the current of your load, and 10 mF (Micro Farad) AC capacitor rated for at least 180VAC. The caps are typically used to start motors or control the speed of ceiling fans, so they are easy to find. I got one at Home Depot in then ceiling fan area. The chokes are a bit tougher, since Radio Shack doesn't carry these anymore, but try eBay or Digikey. The value isn't critical, but it but be rated to handle more current than you will put through it. I used a 15 Amp model for my microwave. I built the entire thing in a plastic outlet box, with an outlet in front, and about 3 feet of heavy duty power cord. The choke goes in SERIES with the outlet on the hot side, and the capacitor goes across the outlet from hot to neutral. Oh, and a fuse is not a bad idea either. So you have just created a "low-pass" filter.

If you have no idea what I just said, as a technical friend to build one for you. Mine took maybe 20 minutes to build once I had the parts, and my microwave noise is completely gone.

 

[dlmorgan999]

What you need is an isolator. These don't exist commercially as far as I know, but you can build one easy.

I have a pretty good background in electronics and I understand what you are talking about. I also have a very good parts store about 10 minutes from my house so I can probably get anything I would need there. In my case though I have no noise on the line so I'm going to try the filters on the potential "sucking" sources first and see what happens. This will be a good tool for the toolbox though anyway so thanks for the info!

 

[Steve]

O.K. A few things. As stated, with SA inverting phase couplers, you can add multiple ones to increase your signal, I use three. But this will only help when the "Other" phase is weak.

What you need is an isolator. These don't exist commercially as far as I know, but you can build one easy. It just so happens that I built one just yesterday. I didn't have a signal sucker, but rather a new Panasonic Inverter Microwave that generates so much noise, my nearby UPB lights fail to work.

O.K. the device. You will need a 100 uH +/-(Micro Henry) choke rated for the current of your load, and 10 mF (Micro Farad) AC capacitor rated for at least 180VAC. The caps are typically used to start motors or control the speed of ceiling fans, so they are easy to find. I got one at Home Depot in then ceiling fan area. The chokes are a bit tougher, since Radio Shack doesn't carry these anymore, but try eBay or Digikey. The value isn't critical, but it but be rated to handle more current than you will put through it. I used a 15 Amp model for my microwave. I built the entire thing in a plastic outlet box, with an outlet in front, and about 3 feet of heavy duty power cord. The choke goes in SERIES with the outlet on the hot side, and the capacitor goes across the outlet from hot to neutral. Oh, and a fuse is not a bad idea either. So you have just created a "low-pass" filter.

If you have no idea what I just said, as a technical friend to build one for you. Mine took maybe 20 minutes to build once I had the parts, and my microwave noise is completely gone.

Boy I wish an isolator could be built for my 220V 40A cooktop but I was told there is nothing that can be done for that.

 

[mustangcoupe]

Boy I wish an isolator could be built for my 220V 40A cooktop but I was told there is nothing that can be done for that.

I'm sure one can be built Steve, but the inductors for the 40 Amps would be relatively large...

Someone else please confirm my information but I believe, you would need one inductor in series with each hot leg, and one 180 VAC cap from each hot leg to ground, and a 270 VAC cap across the 2 hot legs.

 

[ano]

Boy I wish an isolator could be built for my 220V 40A cooktop but I was told there is nothing that can be done for that.

I'm sure one can be built Steve, but the inductors for the 40 Amps would be relatively large...

Someone else please confirm my information but I believe, you would need one inductor in series with each hot leg, and one 180 VAC cap from each hot leg to ground, and a 270 VAC cap across the 2 hot legs.

Yes, one per leg would be best. The problem with the cooktop is its the heating elements that are drawing the power, but its the very low power ignitor that is probably making all the noise, but separating them would be hard.

You might want to try complaining to the cooktop manufacturer, but maybe you have.

 

[hucker]

New years is over and I'm the only one still up... I've been seeing some UPB strangeness lately and thought that I'd run some tests based on what I've learned in this thread.

My house is 3200sqft and 2 of the UPB devices are located in a shed 100 ft away (this distance did not appear to impact the tests). I have an APC 1400 SmartUPS. I have 46 UPB devices. I have an SAI ZPCI-W phase coupler.

I ran the network com test in UPStart first with my UPS plugged in and next with my UPS unplugged. The time difference between the two runs was a few minutes.

I found the following interesting things by looking at the data (I'm doing some rounding for ease of discussion - see the attachment for exact numbers):

1) With the UPS on, many of the devices are sub 10 on PIM and Device signal!
2) Unplugging the UPS improved my average signal at the PIM and the Device by about 20 units!
3) The 'same' phase signal at the PIM and the Device averaged about 10 units more signal than the 'other' phase.
4) My PIM is on the 'other' phase!


The big conclusion is that using the UPS located very close to PIM makes the signal levels unacceptable. Right now it is not easy to move the PIM to the same phase or to move the UPS around...

So to the folks that make UPStart: Doing this analysis was pretty tedious, it would be really cool if the software did the network com test and calculated metrics for the network and provided you with a way to compare one run to the next. You could very easily see if something changed. Ideally it would just flag metrics that changed significantly or cross some threshold so you wouldn't have to pour over all of it (unless you are a geek like me).

PS I attached a PDF of the spread sheet, if you want the excel file PM me...the forum software won't let me upload files that are of spread sheet format.

 

[electron]

I like that idea, great suggestion!

 

[dlmorgan999]

New years is over and I'm the only one still up... I've been seeing some UPB strangeness lately and thought that I'd run some tests based on what I've learned in this thread.

My house is 3200sqft and 2 of the UPB devices are located in a shed 100 ft away (this distance did not appear to impact the tests). I have an APC 1400 SmartUPS. I have 46 UPB devices. I have an SAI ZPCI-W phase coupler.

This is a very interesting post! My house is quite similar to yours - 3600sf, same phase coupler, very similar UPS. In my case the UPS and the PIM are about three feet apart although they are on different circuits. I've got two SAI ZNF10A-P noise filters coming tomorrow and I'm going to use one of them on the UPS to see what happens.

I also think that's a great idea regarding showing changes between runs. During my testing I did something similar to what you did and it *is* very tedious.

 

[Steve]

My house is also similar - around 3000feet and I have the same exact APC 1400 SmartUPS. My PIM (for 'normal' testing) is on a server on a different circuit than everything else. My Elk is on another, my UPS is in the office on yet another (but closest to panel). When I was having trouble with UPstart the latest time and tracked it down to that Outdoor motion light, I did pull the UPS, but what I found was that the UPS actually filtered some of the noise out, and the signal was better with it plugged in! But it was late at night, I was tired, and focused on that 1 device. I really need to do the same comprehensive test that hucker did.

 

[BraveSirRobbin]

Is it just me or could you replace the words "UPB" with "X-10" and think this was an old post?

UPB was supposed to be immune from these noise issues. I was actually thinking of switching to UPB, but after reading posts like this (and this isn't the first one of its type) I changed my mind.

Why should I switch when:

• I went through my home and diagnosed my X-10 signal sucks
• I already have my X-10 signal sucks blocked
• My Ocelot will not do UPB
• I will have to fork out big bucks to replace current X-10 switches
• I see no RF capability with UPB (PalmPads rendered useless)

Anyways, not to get this post to far off topic, but UPB seems like a hassle resonant of the X-10 days...

 

[dlmorgan999]

Anyways, not to get this post to far off topic, but UPB seems like a hassle resonant of the X-10 days...

To me it's a matter of degree. I've had plenty of bad experience with X-10 as well (I started using it in the late 70s when Radio Shack started selling it!). I had far more issues with X-10 than I have had with UPB. Even with the extremely low signal levels I'm seeing on some of my devices I only have one or two switches that are being stubborn.

I personally really like the extremely flexible capabilities as well as the Windows software for configuration (compare that to the PCS X-10 switches that all had to be programmed at the switch itself). I think the bottom line is that no PLC technology will ever be completely immune to issues like this but there is still a good place for them in the automation universe.

-- Dave

 

[hucker]

Is it just me or could you replace the words "UPB" with "X-10" and think this was an old post?

UPB was supposed to be immune from these noise issues. I was actually thinking of switching to UPB, but after reading posts like this (and this isn't the first one of its type) I changed my mind.

Why should I switch when:

• I went through my home and diagnosed my X-10 signal sucks
• I already have my X-10 signal sucks blocked
• My Ocelot will not do UPB
• I will have to fork out big bucks to replace current X-10 switches
• I see no RF capability with UPB (PalmPads rendered useless)

Anyways, not to get this post to far off topic, but UPB seems like a hassle resonant of the X-10 days...

I suppose that I should add. I have never had a switch not activate the load even at these very low signal levels. That is super important.