Hard wired communication link?

[GAR]

110804-1042 EDT

Why are there not TED models with hard wiring from the MTU to the RDU or Gateway to eliminate the unreliability of the power line communication path? It seems to me it would have been self-evident that when the 5000 series was created that a hard wire link would have been a better system design choice.

.

[rotus8]

The biggest reason is that code says you may not have low voltage wiring going from inside an AC panel to outside the panel, without the wiring being rated for the full AC requirement.

[GAR]

110804-1249 EDT

rotus8:

That is not an adequate reason to use power line communication. More than one way could be used to get the data from inside the panel to outside.

.

[rotus8]

You should go to work for TED and show them how to do it.

[GAR]

110806-2228 EDT

rotus8:

One way to do this is to make the MTU fit in a knockout so that on the outside of the main panel is a connector for the signal cable. There also are other ways to get the data out.

From the number of posts with with communication problems it seems that getting rid of the power line communication link would be a good idea. For my 1000 system I run an extension cord from the point where the MTU connects to the bus to where I have the display and data collection computer. So the display is connected into the extension cord. The computer elsewhere. I still have some dropout problems. Usually not too long.

.

[TEDSupport6]

Our research and development department is constantly working on ideas for improving TED for the future, particularly when it comes to MTU/Gateway communication. Once those ideas are proved effective, they are moved into production. Thank you for your feedback!

[jrwalte]

It isn't easy, but you can technically make a 'hardwire link' between the MTU and Gateway now.

You purchase an in-line filter and install it on a new breaker to isolate the breaker from the rest of your home. You then run a new electrical line from the isolated breaker to a new outlet that is in reach of your Ethernet network. The MTU/Gateway are now segregated on a filtered, dedicated electrical line.

[cloudix]

Why not use WIFI or Bluetooth as another option? forget the wires if you don't really need them.

[GAR]

110812-1218 EDT

jrwalte:

That is essentially what I do presently, except I do not have the in-line filter. I use an extension cord.

My point of this thread is that power line communication is a poor way to build a reliable communication system. The reliability is still poor when going thru the gyrations of testing and filtering. Whereas, with an isolated hard wired communication path the reliability would be extremely high.

I have customers sending large amounts of data (100s of millions of bytes) at 115 kbaud, with no error correction, and no errors over distances of hundreds of feet. I can subject the interconnect cable to a common mode voltage of 1000 V RMS at 60 Hz with no data errors. My path has optical isolation at both ends. I do not suggest that the TED system needs my kind of isolation, just that it use a hard wire connection. Probably with simple optical isolation at the MTU. High baud rate capability is not needed. A hundred bytes per second are probably adequate. So roughly 1200 baud would do.

.

[tazer]

Why not use WIFI or Bluetooth as another option? forget the wires if you don't really need them.

Wireless does not work well from inside a metal enclosure, such as an electrical panel.

In my experience, PLC has worked pretty well. The only issue I had was when I upgraded to Insteon home control devices and the TED's data flow caused some interference. This was easily resolved by using an X10 filter (about 8 bucks, ebay) at the main panel and a dedicated circuit. This achieves a hardwired connection between MTU and gateway.

Let's say TED did offer a specific hardwired solution... You'd still have to run a wire from the electrical panel to your gateway. I don't see how that offers any real advantage over the filtered-dedicated-circuit approach. In either approach you're having to run a new wire from panel to gateway. The filtered-dedicated-circuit approach is superior in that the same wire provides both power and communication.  The only disadvantage is an additional thing (the filter) crammed into the space of your electrical panel.

[GAR]

1108123-0750 EDT

tazer:

The best signal to noise ratio and good bandwidth is going to be achieved with a separate hard wired or fiber communication path.

RF is vulnerable to microwave oven interference and other RF sources.

Power line carrier is a very noisy low bandwidth, and high attenuation path.

In most homes there is not too much problem to install some CAT-5 cable or other similar wire.

The TED system appears to be a one way path from the MTU to the RDU or Gateway. Thus, unless you can improve the signal to noise ratio (filters help this) the only way to reduce errors is to add redundancy to the data. This is won't solve the problem of total dropouts.

.

[rotus8]

The Gateway <-> MTU connection is bi-directional.

[GAR]

110813-1008 EDT

Even if the MTU communication path is bi-directional is it being used to request retransmission when errors are detected? Further is there any means employed to detect errors? The evidence from problems does not indicate any error correction. How many bytes of data are transmitted each second?

.

[jrwalte]

Except most if not all state codes do not allow you to run low-voltage CAT5 into your service panel, and many of us do not have exposed service panels (they're flush in a wall) so we can't just attach a metal box to the external side of the service panel and install the MTU and the CAT5 into the new metal box.

TED has to think of all installations this product needs to work in, and so far no better option over PLC has been mentioned.

[tazer]

The TED system appears to be a one way path from the MTU to the RDU or Gateway. Thus, unless you can improve the signal to noise ratio (filters help this) the only way to reduce errors is to add redundancy to the data. This is won't solve the problem of total dropouts.

Have you tried using an in-panel filter and a dedicated circuit yet? Your other post mentioned using an extension cord but without a filter. The filter really is the key to having the dedicated circuit.