X-10 Modules not Responding after Smart Meter Install

[GMAN]

Well I was working around the house this morning ( inside as it's 20 degrees outside) and wham the lights went out. I heard noise outside and went out to find the worker swapping out the meter.

Aside from cleaning off some grease?  on the painted panel below it,  so far so good. I don't have a huge setup, but Remotes, Macros, etc.
seem to be working.

Or do problems have yet to arise?

Of course I had to go around and reset the usual things and for measure I reloaded the interface.

I'm in So-Cal, (Edison) so if any impromptu notes are being compiled about the meters being installed let it be known and I will post any info I can find on it.   G.

[Noam]

Would something in-line between the meter and the breaker box be able to filter out the noise easier? Like a step-down transformer, but not actually stepping down?

Yes, a filter between the meter and the distribution panel could block the signal.  One of my customers had a PZZ01 installed.  As I recall, it helped, but didn't entirely solve the problem.   An in-line filter would typically have to handle 200 amps, and might be pretty large.  Think of a 2-phase version of the XPF that could handle 10 times the current.

But as I said in a prior post, I think the utility companies want this signal to intrude into our homes because the eventual intent is for them to be able to shut down our major electrical appliances during peak demand so they do not need to provide on-line reserve generating capacity.

Jeff

As I understand it, the PZZ01 sits around the neutral wire, and inductively senses the signals to block, then injects the opposite signal to the powerline to cancel it out.
I'm thinking more along the lines of disconnecting the lines coming into the panel, hooking them to this "black box," and then connecting the "black box" to the breaker box. Whatever is inside that box would isolate the power lines from the other side.
Doesn't a standard step-down transformer isolate the power? Isn't that why PLC signals don't pass beyond the pole transformer?

If the stepping-down is done by the ratio of windings on one side to the other, then wouldn't an equal number of windings on both sides keep the voltage the same, yet still isolate the noise?

[Noam]

Well I was working around the house this morning ( inside as it's 20 degrees outside) and wham the lights went out. I heard noise outside and went out to find the worker swapping out the meter.

What brand and model of meter did they install?

[JeffVolp]

As I understand it, the PZZ01 sits around the neutral wire, and inductively senses the signals to block, then injects the opposite signal to the powerline to cancel it out.
I'm thinking more along the lines of disconnecting the lines coming into the panel, hooking them to this "black box," and then connecting the "black box" to the breaker box. Whatever is inside that box would isolate the power lines from the other side.
Doesn't a standard step-down transformer isolate the power? Isn't that why PLC signals don't pass beyond the pole transformer?
If the stepping-down is done by the ratio of windings on one side to the other, then wouldn't an equal number of windings on both sides keep the voltage the same, yet still isolate the noise?

I did discuss an in-line filter above.  A transformer would also provide the isolation, but it would be even larger and more expensive than a 200A in-line filter.  Think about the 60Hz transformers that were in old tube-type TV sets.  They were rated about 300 watts.  A transformer for an entire home would have to handle 48,000 watts (240V x 200A).  That would be enormous and EXPENSIVE.  Even an in-line filter would have huge inductors to handle the 200A.

While the concept behind the PZZ01 may be viable here, it probably doesn't work well because the LonWorks transceiver is very powerful.

Test results from the TSS notch filters has not been that promising, but that may be due to line inductance.  Even a few feet of wire between the filters and the distribution panel can have enough inductance to reduce their effectiveness.  I hope the customer can re-run the test with the filters located directly adjacent to the panel to cancel out as much line inductance as possible.

Jeff

[dhouston]

But as I said in a prior post, I think the utility companies want this signal to intrude into our homes because the eventual intent is for them to be able to shut down our major electrical appliances during peak demand so they do not need to provide on-line reserve generating capacity.

As more and more utilities move to solar and wind supplies, they also encounter another problem which is somewhat the opposite of reducing peak demand.

• http://www.nytimes.com/2011/11/05/business/energy-environment/as-wind-energy-use-grows-utilities-seek-to-stabilize-power-grid.html?_r=2&scp=1&sq=taming%20wind%20wald&st=cse

[Noam]

A transformer for an entire home would have to handle 48,000 watts (240V x 200A).  That would be enormous and EXPENSIVE.  Even an in-line filter would have huge inductors to handle the 200A.

I never said it was a GOOD idea, or a cheap one.
I was just wondering if it would work (in theory, at least).

[GMAN]

Meter picture

[dhouston]

There is nothing to stop it from coming into the house.  He read .5V in his lab, which is a long run from the distribution panel.  The X10 signal was .3V barefoot, and 2V with the XTB-IIR in service.

Your mention of the Duke engineer's lab joggled a few dormant brain cells and I recalled this...

• http://www.smartmeters.com/the-news/142-duke-energy-unveils-technology-centre.html

If the Duke engineer you are communicating with has his lab there, it's only 2-3 miles from me. If so, I can help should he need a real-world test site.

[JeffVolp]

If the Duke engineer you are communicating with has his lab there, it's only 2-3 miles from me. If so, I can help should he need a real-world test site.

Actually, that is my customer who has been doing all the testing.  The 75KHz notch has helped, but has not entirely solved the problem.  More "BareBones" PCBs should be here tomorrow night to add a second notch for the 86KHz.  Duke offered to pull the smart meter off his home, which may be the final solution.

Jeff

[dhouston]

The Duke engineer told me thet the problem was NOT with the smart meters.  They have actually removed some of them and the X10 problems persisted.  He said the PLC was coming from the data aggregator (communications node) out at the transformer.  That's what talks to the meters and then relays via cell back to Duke.  Removing the meter alone does not fix the problem.  They have to turn the aggregator off too...

I quoted the above from the Smart Meter Survey, moving it here to avoid cluttering that thread with comments.

It might be helpful to know the Aggregator make/model/specs to see whether reports of problems with non-Echelon smart meters might be related to the same Aggregator.

It would also be helpful to know whether any logging application logged X10 commands at the time of the random ONs (in the above case, JeffreyB's Ocelot did NOT log any).

The Echelon Smart Meters are used by every customer of the largest electric utility in Italy without apparent X10 issues.

[JeffVolp]

I have some additional information from the customer I am working with in Cincinnati.  He saw both the 75KHz and the 86KHz frequencies.  One appears to be from the data aggregator, and the other from the smart meter.  He is using two notch filters now that should be attenuating both frequencies.  However, the tuning may not be optimal because it can be effected by line inductance.  I have parts coming in to build another version that should be more immune to line inductance.

The last report I received from him was that there had been no X10 failures for 16 hours since the installation of the notch filters.  I think we are approaching a viable solution.

Jeff

[Noam]

I have some additional information from the customer I am working with in Cincinnati.  He saw both the 75KHz and the 86KHz frequencies.  One appears to be from the data aggregator, and the other from the smart meter.  He is using two notch filters now that should be attenuating both frequencies.  However, the tuning may not be optimal because it can be effected by line inductance.  I have parts coming in to build another version that should be more immune to line inductance.

The last report I received from him was that there had been no X10 failures for 16 hours since the installation of the notch filters.  I think we are approaching a viable solution.

Jeff

How are the "notch filters" installed?

[JeffVolp]

How are the "notch filters" installed?

It is a simple plug-in module a little smaller than the X10 XPPF.  To reduce the SmartMeter signals, plug one into each phase as close to the panel as possible.

Jeff

[Noam]

That sounds simple enough. 

[JeffreyB]

I just went out to the transformer and got the info off of the "data aggregator":

Ambient
Model X-3100-GB1

Jeff