UPB Generation II Core

[mlester]

For several months now PCS has been rewiting the core receiving algorithms in order to attack the few noise sources that can effect UPB receiving. All comm can be effected by "noise" as it it defined for that particular physical layer method - it doesn't matter if its powerline, rf, hardwired, auditory, hand signing, morse code, etc.

UPB "noise" is anything that looks like UPB pulses to a UPB receiver. Most of this UPB type noise is already handled by the current core - but we decided good isn't good enough. We have learned so much over the last 2 years while weve been designing our Industrial powerline version of the core (much slower - too slow for residential) that some of this can be fairly simply applied to the residential core.

The main culprets we know of are old X10 dimmers set to FULL LOW (so low you cant see the bulb filiment glow) and Lutron devices Radio Ra/ Homeworks - something in their design/power supplies. This is all "steady state" noise. Absolutely regular, present the same in every 1/2 cycle. This is pretty easy to fix.

Then there are a few - very few "random" noise generators. Large DC brush motors. Very rare microwaves,and florescent ballasts. We know of only a couple models of each of these that make some noise. Almost ALL home devices don't make any UPB type noise.

Our goal is to attack the few noise effects there are and push UPB reliability as high as we possibly can. Any experience out there with specific noise sources will help us get it as good as we can. The recieving algorithms involve countless tradeoffs and small adjustments in different parameters, and are also very limited by the resources in the inexpensive microprocessors.

Honestly - we don't get very many calls about noise issues at PCS but the few we have gotten need to be addressed. As more and more UPB is installed by PCS and our licensses the more "fringe" issues emerge. That is normal and to be expected. I here a lot about the "rock solid" UPB comm - but our goal is to address the <1% that are issues, not the >99% that are not issues.

Any input that can make this rewrite come out as close to perfect will be appreciated.

 

[bfisher]

I hope I'm not opening a can of worms here... I have NO issues with any noise - I don't think I've ever had a missed UPB signal on any of my switches/outlets/modules. However, I think there are 3 more common issues that we (the DIYers) would really like addressed.

These are the reasons commonly given here in the forums for not using UPB (and trying Insteon or ZWave):
- delay with the switches (while the switch waits to see if you double tap, etc). Many of my switches don't use doubletaps and I'd love to be able to disable this delay so the response is almost instantenous. For the switches I want doubletaps - let me choose that option.
- price - $79 per switch is very expensive and forces people (myself included) to really think hard if I need to automate that particular switch. If you could get the price point down to $40-50 (or even less) - you would own the market. UPB is very reliable, but the price scares many DIY away.
- buzz - doesn't bother me in the least, but I've seen several people here complain about it.

Sorry - I know it wasn't the question you were asking. But I wanted to take this opportunity to share with you what I see as the bigger issues facing UPB in this market segment.

(I know we have 1 member here with a stove that causes noise problems. If he doesn't post in a day or so - I will dig up the old posts and make sure you have the info)

 

[WayneW]

For several months now PCS has been rewiting the core receiving algorithms in order to attack the few noise sources that can effect UPB receiving.

Welcome to CocoonTech, Marshall.

For anybody that didn't recognize the name, Marshall Lester is the CEO of PCS (Powerline Control Systems) and the inventor of UPB.

 

[toymaster458]

I had one major noise issue in my house that took out a whole phase of my house. I bought some under the cabinet lights from Lowes (I need to go back and get the make and model) and connected them to a couple SAI UML. Now according to the instructions that came with the lights they are dimmable so I so no issue using the UML in dimming mode. Boy was I wrong. At first it worked fine but as time went on things got worse. One of the UML would not respond to any UPB commands and the light always stayed on. The noise level on the phase started at about a 1 or 3 then by the time I was having major issues grew to about 15. The other UML was missing commands. It was not untill I changed the UML to non dimming that the noise went away.

 

[Steve]

Hi Marshall - thanks for dropping by CocconTech! I am the cooktop guy Bruce was referring to. I will paste some detail here for your convenience or you can PM me for additional details if you wish.

I do have a cooktop that pumps out quite a bit of noise. Here is a quote from another post containing the model (bolded) information:

I'm sure it depends on the model. When we redid the kitchen we splurged on the cook top and got a Dacor with fancy schmancy controls (MET304). I'm guessing the electronics behind the controls it what's causing the problems. I guess my only hope is that someone builds a 220V 40A filter that blocks out UPB freq. I'm praying Dacor has a good answer for me. Heck, with the amount that stupid thing cost, they should build me a filter! I suspect though that since UPB is typically put in 'high end' homes by pros (due to cost) and I'm guessing a lot of these homes have high end appliances, etc, then we will start to see more problems and eventually filters. I just happen to be the poor (broke) guy who can't use x10 and happen to have a few high end appliances (that are still financed).

I was working with Brad from Simply Automated on this. From my understanding, SAI actually purchased some of the cooktop components to test and built filters to suppress the noise. He also indicated their long term strategy was to build logic into the firmware to 'ignore' the noise. I do not have the filters in place at this time. Last I spoke to Brad in March they were just about ready. Here is some detail from an email sent to Brad that outlines my test and documenting the noise. I believe the thought was that the cooktop used triacs which induced this noise:

I just completed some testing that was very consistent, I will outline it below. Setup info - there are 4 1/2 burners that I called A, B, C, D and D+. They started at upper left and circled around to lower left like this:


A      B
D,D+  C
 
Burner D comes on inside by default and then a separate button turns the outer portion on which I called +, so D is inner only and D+ is inner plus outer.


I ran the test with the PIM on both phases of the power line - plugged into outlet on Phase A then in different outlet on Phase B. The test was very close between the phases, so I am only going to show it once and consider it the same for each phase. Each burner has a slide control that goes from 1 (least power) to 10 (full power). I tested each burner on 1, 5 and 10. In reality I found that 1 thru 9 behaved pretty much the same. Here are the results.


 
Burner A  Noise 
On 1          16   
5              16   
10              0     
             
Burner B             
On 1          19   
5              20   
10              0     
             
Burner C             
On 1          12   
5              12   
10              0     
             
Burner D             
On 1          10   
5              10   
10              0     
             
Burner D+             
On 1          26   
5              26   
10            26   

Essentially each burner when run individually has Severe ( > 10 ) noise. This is the noise reading from the red vertical bar with reading under it, not numbers on the left.  The really funny thing was that when each burner was run full tilt, the noise stopped. The noise was pretty much the same on settings 1-9. Also of note is the Burner D, the combo burner, put out abnormally high amount of noise when run in combo mode (D+, both inside and outside on).

I also ran a combination of all burners on, which had similar results to just Burner D+, noise of around 26 on all settings except 10. With all burners on and running full tilt (10), noise was again at zero (0). I think you would see similar results for just various combinations.

I actually stumbled upon this because my wife happened to be cooking when I was setting up some switches. The noise effectively killed all use of UPStart. Phantom switches (yellow boxes) would show up all over the place and I would not be able to get UPstart to see a switch in setup mode. The good news is that outside of setup, it appears UPB is robust enough to power thru the noise and is only seldom affected when the cooktop is on. Recently, when replacing some switches I think I noticed some noise when another appliciance, a GE Advantium, was on but I would have to do more complete testing - but it was nowhere near the levels of the cooktop. Other noise issues I heard alot about at EHX was using CFL's, especially dimmable ones.

If I may, while we have your ear, I'd like to go over some other issues beyond noise. Bruce pretty much mentioned the top 3 reasons (Delay, Price, Buzz) that people shy away from UPB. It appears that UPB's (and PCS in particular) strategy is to market to the professional installer market and as such prices stay high to give dealers/installers a large profit margin. If that remains your strategy, I wish you luck, but In My Humble Opinion I think you are losing out on a large market and huge potential. UPB has a lot of strengths that appeal to all market segments including DIY and especially advanced DIY. I can tell you I personally see dozens (and can imagine hundreds if not thousands) of potential customers choosing other technologies, especially Insteon largely because of the above reasons. It is again my opinion that Insteon would be making a huge dent (already has to an extent) and really kicking you in the pants if it were not for their own problems. I guess my point is if you are targeting strictly the Pro market, then keep up the good work, but you have a much bigger market on the verge of exploding if you can address the 3 mentioned items above. Another thing to consider product wise is a light/dimmer bar on the switch - people seem to really like those. Also, the LED comes up from time to time. Different manufactures have different colors, and in fact, some manufacturers even have different colors in different lines. That makes it very difficult to use devices from different lines or manufacturers. Also, the current trend seem to be using a Blue LED. Now, if you could incorporate a multicolor LED (not just 3) that was software controllable, that would be hot! and could be set to match anything.

And on a final topic, please take a look at look at this post and this one. These have to do with power quality/surges. Last week an apparently relatively minor (based on the fact that there is no other discernable damage) power glitch took out 10 of my 15 switches. I happened to be researching whole house surge protectors at the time and became very concerned when this happened. What scares me is that 65% of power surges come from inside the home - A/C, dishwasher, vacuum, garbage disposal, hair dryer, etc - practically anything with a motor. A whole house potector at the service entrance still leaves the switches unprotected against these internal surges. My only conclusion was that perhaps the switches should have an onboard MOV or some other type of protection mechanism. Do you have any thoughts on this and how to protect our significant investment in UPB devices?

Thanks again for stopping by. I look forward to any improvements in UPB and hope to see some changes that result in a larger DIY market.

 

[mlester]

This is to Steve, "Last week an apparently relatively minor (based on the fact that there is no other discernable damage) power glitch took out 10 of my 15 switches."

No normal surge should ever "take out" any UPB switch, let alone 10. Electrically they are pretty bulletproof. Are these PCS switches? If so send them to me for analysis. If not, who's are they? Then send them to the manufacturer and make them call me on this.

mlester

 

[mlester]

About this - - "delay with the switches (while the switch waits to see if you double tap, etc). Many of my switches don't use doubletaps and I'd love to be able to disable this delay so the response is almost instantenous. For the switches I want doubletaps - let me choose that option."

We can do this. Is it really worth it? We were testing this yesterday since Bill Olson mentioned it and it seemed to us like the delay was barely noticable - but if it's bothersome to people we can make it a setting - just like the post above suggests.

Any other requests for changes?

mlester

 

[Mike]

To give further feedback to Steve's comment, I wound up installing Insteon myself. Those three issues (price (twice that of insteon, although when I was looking it was not as big of a gap), delay (I am fairly certain it would drive my wife nuts) and the buzz (wife acceptance again)) were the primary factors in not selecting UPB. Some of the features (multi button switches) and the software looked real promising though.

In my case, had the latter two issues not been an issue, I would have looked at it further (although my next review would have been the value provided by UPB to justify the higher cost).

Hope this helps.

 

[Steve]

This is to Steve, "Last week an apparently relatively minor (based on the fact that there is no other discernable damage) power glitch took out 10 of my 15 switches."

No normal surge should ever "take out" any UPB switch, let alone 10. Electrically they are pretty bulletproof. Are these PCS switches? If so send them to me for analysis. If not, who's are they? Then send them to the manufacturer and make them call me on this.

mlester

They were Simply Automated US2-40's. They are on their way back to Martin at Automated Outlet. I believe he is going to look at them and send them to SAI.

Martin: Could you also please get whatever feedback there is to Marshall based on his request...

 

[toymaster458]

The delay is a BIG issue. My wife beleaves when you turn on a light it should turn on NOW! not in a second. This delay gives the optical illusion that UPB is slow.

As far as request I would like to see light bars on masters and slave switches. Outdoor lamp and appliance modules, a switch designed just for fans, High, Med, Low, Off buttons only with a led next to each button showing the current status!!

Another nice feature would be for the switches to repond with their status after the receive a link command not just when it is controlled locally.

 

[rocco]

The delay is a BIG issue. My wife beleaves when you turn on a light it should turn on NOW! not in a second.

My wife, as well.

Her point was that the old $3 switch turned on instantly, so she would prefer those.

 

[mlester]

Is this delay in going on in PCS switches - or only in Simply Automated switches?

Is this noticable when:

1) the softstart ramping is turned on? (default)

2) When the switch is set to instant on on one tap? Or to non-dimming which is the same effect?

3) Using the double-tap as the instant on? (which is the default)

If the softstart is set "on" then there is a noticable delay as the ramping goes up and the light bulb heats up. But I don't think that is what you all are talking about.

mlester

 

[toymaster458]

On a SAI Switch even if I have it set to non-dimmable and Snap! it has a delay because it is waiting to find out if you want a double tap or go into setup with 5 taps.

 

[toymaster458]

Taking a quick peek at the PCS product line another feature request would be optional Blue LED and mutli-color back lights on the Controllers so each button can be a different color with Blue as an option also. Another feature would be having the Wall Controllers as a load controlling switch also.

 

[elcano]

Another request I would like to add are screw-on fixture dimmers. For installing just between the lamp and the light bulb.