Homeplug HL106, how well do they work?

P.S.: Our required data rates will be _very_ slow, under 100bps like in the old telex machine days. And only occasional bursts once in a blue moon. So this would be like sending morse code via a HDTV channel.

The datarate is moot. With GFI's it's the carrier (4-20MHz) that counts. I have found that even X-10's 120KHz carrier gets blocked by my GFI's. Surge protected power strips will also block the signal. As for talking across the 2 phases a repeater works well but I didn't see one on their website. Passive coupling helps. Any capacitive loads, like switching supplies, on the power line tend to sink the signal. Art

You have, I assume, a good reason to not use wireless?

Yup: $$$$$ :-(

For some reason anything becomes super-expensive the millisecond you want data to be piped into a LAN and from there onto the web. The only option is to roll your own which is cost prohibitive when the quantities are too low. Intentional radiator cert and all that fun stuff. Then there's all those international regs to deal with. For example, we need a sustained 1000ft indoor to outdoor range regardless of building materials. Pricey 900MHz systems can do that in the US but not internationally.

So, yesterday I tested a pair. The results were not too encouraging: Obviously they do not like surge protectors. Which are installed in many places here. Through one the connection worked, sort of, on and off. Through two of them (one at each location) there was no reliable connection possible.

Anyone know of lower bandwidth devices that can work with much smaller signal levels? All we need is 100bps (yes, as in bits per second), not Internet speeds. But it has to be reliable.

Some related info that may help. Many moons ago I had similar problems with my X10 stuff which uses 120KHz for the carrier freq. El cheapo surge protectors which used a triad of capacitors effectively shorted the carrier out but the ones which used transorbs worked much better. GFI's blocked the carrier due to the coils in series with each wire. Bypassing the coils with a small cap fixed that but of course wasn't to code and blah, blah, blah. I forget the value of the caps but I added them externally - black line to black load & white line to white load. HTH. Art

I wonder how on earth you are going to communicate 1000 ft over a power line communication system, especially in the USA ?

In Europe in which a single low voltage (230 V) distribution transformer could serve customers in an area of at least 1 km², Some ISPs especially in Germany used Mainnet and similar systems to connect to the customers with some success (a few mbit/s total capacity).

In the US, with low voltage (110 V) distribution transformers ("pigs") in nearly every other pole, the powerline communication area is going to be very short. This effectively killed the ISP power line attempts in the US, since the signal would have to ve fed into the medium voltage line, then couplers across every MV/LV transformers would be needed, in order to reach the customers on the low voltage side. Such devices are expensive (high isolation voltages needed) and of course needed to be installed by the power company.

Thus, in order to communicate from one LV network to a device on a different LV network serviced by a different distribution transformer, would require bridging the two transformers with suitable couplers, in order to use the MV line for intertransformer communication.

Of course, these couplers would need to be installed by the power company. Thus, in practice, how would you be able to communicate for up to 300 m in the US using power line technology ?

We don't need to hop any transformers because it will always be on the same premises. Properties in the US, even residential ones, can be huge. There are many homes that sit on 5-acre or 10-acre lots. Wealthy people have much of that fully landscaped. There are granny flats, guest houses, pool houses and so on. All on the same power transformer.

capacitors

You rarely find transzorbs, mostly MOVs. But either way is a problem because both have a tremendous capacitance. My surge protectors are a bit on the extreme, each has a whole fistful of MOVs in there. A dozen or so each.

and

Sure, I'd know how to fix it. Little inductors would do the trick, for example. But this has to be a plug-and-play product. We don't want to hunt down and exchange all possible interfering devices at each customer.

Looks like RF then.

capacitors

and

Right. I've gotten my X-10 system to work. Several times. :-( Over the years, you add or replace some device or other, and then the system goes blooey. You track it down, add an L or C or 2, and everything works perfectly, again. The last time it was a replacement of an electric toothbrush that screwed up the system. Grumble.

Ed